Static Zombie

Puzzle Hunt 123

The Microsoft Puzzle Hunt, which I've been working on for about 15 long months, is now over. This event broke me. It is probably the last puzzle event I will ever run with a volunteer committee.

We tried some big, risky things in this event, and I'm very happy about that. I'd rather fail spectacularly for trying something different than do something safe that doesn't push the envelope. At least one of our innovations-- timed puzzles that teams were encouraged to solve as a team as an in-conference-room event-- was a resounding success. This concept was born from my feeling that out-of-conference-room events represent a tremendous amount of overhead for something that a small percentage of players ever see. From a cost/benefit perspective, they're a horrible investment. I wanted to find a way to create special moments the entire team could partake in. Timed puzzles, specifically constructed to be conducive to group solves, were a great low-cost, high-impact solution, and they seem to have been universally adored.

Some experiments work, some don't. In retrospect, it's clear how different decisions would have made the event better. I take the blame for all the problems that didn't get corrected. There was no one leader-- the hunt was essentially run by committee. That doesn't excuse me from responsibility for poor design or execution. We had a chance during the event to correct the biggest problem-- players being blocked from accessing more puzzles-- and I pushed the wrong priorities. Instead of looking at the evidence that teams just didn't want to use our existing release valve of moving from the Competitive to the Recreational division, I stood by it under the belief that any change of course at 3 AM would represent a breach of trust to the Competitive teams that had moved beyond the blockage, and to the teams that had already switched to Recreational to get around it. I still believe that to be true, but breaching that trust and unblocking players may have been the lesser evil.

I feel deeply disappointed that, after 15 months of planning, the event we ran was not the event people wanted to play. I grossly misjudged what people wanted from Puzzle Hunt. Competition is deeply ingrained in the DNA of its players, and they accepted enormous amounts of frustration rather than give that up. Some people on the organizing committee thought that might happen, but I didn't believe it. I was wrong. I accept the blame. I deeply apologize to all the players whose fun was compromised as a result. I also feel terrible for all the puzzle authors whose work got less exposure because of it.

The event was created when two teams, each planning a Hunt, ran out of steam on their own and merged (the events merged; almost all of my original team simply bailed). That was reflected in many ways in the event, and usually not for the better. Elements conflicted with each other. Problems compounded each other. And mostly, the creators were just tired and ready to be done. It frustrated me to be the front man for an event that I didn't entirely believe in, and it depresses me to feel so defeated by the experience. I don't intend to put myself in that position again.

City Chase Seattle

Every muscle in my body aches.

Yesterday I participated in City Chase, an "urban adventure" competition running in multiple cities around the world this year. Teams of two have 6 hours to complete 10 challenges (from a menu of 14 choices) located all around the city.

When my teammate and I arrived and met up with a pair of friends-- with whom we traveled the entire day-- we quickly realized as we surveyed the crowd of young, athletic competitors that we had signed up for a very different event from the rest of them. Many of the other teams, with their camel packs and lycra, were clearly there for the "race" aspect of the event. When the perky hosts got on stage to lead the group in a series of warm-up exercises and received about a 95% participation rate, I had an acute feeling of culture shock. These people were serious. We were there to have fun. And to be fair, so were they. Our definitions were just a little different.

The event kicked off with a tiny scavenger hunt as a way to stagger teams out from the start-- answer some trivia, find a couple of goofy things (a stranger the same height as you, a live animal, etc), that kind of thing. Then you received your list of "ChasePoints" and could begin planning your own route for the day, restricted to travel by foot or public transportation. The winner of the event did the smart thing and immediately ran themselves in the opposite direction from the closest ChasePoints, thereby avoiding crowds and experiencing no wait times. The key word there was "ran". Our foursome was on a strict no-running plan, so we want with the path of least resistance and hit the closest sites first.

Here's the rundown of what we did:

Shinteki: Decathlon 4

Because you (yes you, Wei-Hwa) demanded it, a recap of the fourth Shinteki Decathlon, which the fiancee and I played in a couple of weekends ago (on different teams).

This was unquestionably the easiest Decathlon yet. Briny Deep solved all regular clues without taking any hints, missing only two bonus clues. Which isn't to say the event was easy. I think the difficulty level was just right, and most of the clues were solid. In the past, bonus puzzles were hidden inside each other puzzle. While intriguing in concept, in practice they were often very difficult to find and, due to the constraints such a scheme places on their construction, sometimes not very good. The new system dispensed with the hide and seek and gave us a booklet of bonus clues outright, leaving it to us to figure out which ones linked to which main clues and how. This worked out much better, giving teams something tangible to chew on between main clues. Definitely a keeper going forward.

As always, a big thanks to Brent, Linda, Martin, and the entire JPT crew for running the event.

The theme this time was Child's Play, and all of the clues hewed nicely to that theme.

Shintekimon: Teams faced off in rounds of Shintekimon using the traditional Shinteki Palm devices. After naming their Shintekimon, teams could battle each other by using the Palm's beaming ability. The result was anywhere from 0 to 5 rounds of battle, with wins, losses, and ties reported for each round. When we thought we were ready, we could fight the reigning champion, Superstar. Defeating him opened the gate for the team to move on to the next clue. Shintekimon could be renamed at any time, and the more battles you fought, the faster you got hints. We were just converging on the preponderance of Rs and Ss in the various champion names when the hint dropped revealing were were playing Rock Paper Scissors by comparing the names of the two competitors and ignoring all letters except R(ock), P(aper), and S(cissors). Another few minutes and I'm confident we would have hit it on our own. We spent a little too much time blindly battling and not enough analyzing our data. Or too much time analyzing our data and not enough battling and earning hints. Take your pick. A nicely-conceived puzzle that leveraged the presence of all teams and got everyone interacting.

Nursery Rhymes: A long climb up a hill to a scenic view, thus continuing the Shinteki tradition of getting teams sweaty at the start of the event for maximum van ambience. Along the way we encountered reworded nursery rhymes we had to recognize, and at the top we got a double crostic to make sense of them. Entirely straightforward, it would have been nice to have some kind of twist here to spice up the puzzle. Other than recognizing the rhymes, of course, which were all pretty obvious.

Connect Four: A travel Connect Four set with letters written on both sides of the checkers. Each checker indicated which column it belonged to, and the board itself had word separators. A rather clever use of Connect Four, turning a game about dropping checkers in columns into a two-sided drop-quote. Noticing that the blank-on-one-side checkers had only once place they could be gave us our start point, and we made pretty short work of the grid by solving from the bottom up on one side and using the back as a sanity check. The next leap-- recognizing the grid itself as a calendar-- was satisfying, making good use of the seven-column grid size and neatly explaining the blank leading/trailing checkers. A solid puzzle and a fun one to solve as a group.

Red Light, Green Light: To obtain the next clue we played a quick game of Red Light Green Light with a human traffic light-- some quick, childish fun. The clue itself was a set of eight cards, each with a set of eight transinserted (scrambled, along with an extra letter) items. Unscrambling the items identified the extra letters, which themselves formed transinserted members of a ninth set. A classic, recursive puzzle form, and one that is marvelously suited for team solving (unlike, say, a cryptogram or sudoku). I'm a big fan of puzzles composed of self-contained micro-puzzles for that reason. Lay out all the cards and everyone on the team can contribute, calling out answers and filling in the blanks.

Gashlycrumb Tinies: At the Winchester Mystery House we received a copy of Edward Gorey's The Gashlycrumb Tinies and a puzzle derived from it, a set of word balloons containing strings of letters and numbers that looked like cryptic crossword enumerations. No coincidence, that, because each balloon was the result of running a word through a cryptic-style transformation suggested by the manner in which one of the children in the book died. This part of the puzzle was nifty, fun, and a good group solve. The next step-- taking the resulting sequence of children and reading them as binary based on their gender, felt arbitrary and an unnecessary extra layer. The hint for it was on page one of the book, which I looked at in the very beginning and then completely forgot about by the time we needed it. Even had we been staring at the hint the entire time, the puzzle would have been more satisfying without the shift from cryptics to binary. This feeling was exacerbated by the choice of message generated by the first step-- a series of children's names from the book-- which suggested a form of recursion. In fact it was completely arbitrary and could have been many different letter sequences, but the apparent signal of names from the book kept us from searching the book for other clues and rediscovering the binary hint.

Jenga: A set of Jenga blocks and an algorithm for manipulating them. Once we'd completed the algorithm, the blocks themselves formed letters when seen from two of the four sides (since the letters, YAHOO, are left-right symmetrical). Less a puzzle than an exercise in following directions. Fun to work through together, but just meh overall because of the extremely low difficulty.

See 'n' Say: At a farm-themed park we faced a giant See 'N' Say depicting eight farm animals. Answering an animal trivia question allowed us to set the initial position of the dial and see where it wound up pointing and what animal sound it made (which weren't the same). Once we'd collected all eight data points-- sets of two locations at 45 degree increments on a circle-- we immediately knew we had eight semaphore letters. But the data gave us garbage. Fortunately the trusty Palm device recognized the garbage as a partial answer and told us to use all the data. We'd been using only the endpoint of each spin and the sound it produced, but the start point also gave us information-- an orientation. Viewing each semaphore pair with the start point of the spin as "up" gave us the real answer. We knocked this one out in very little time, and it was quite cute, but simple. Shame they weren't able to hack real See 'N' Says, which would have been teh awes0me.

Coloring Books: A great idea marred by its form factor. We received a small, staple-bound booklet with grids of numbers on the back pages and photos of childrens' book covers-- titles digitally removed-- on the front pages. Since this was conveniently distributed at a library, we went inside to use the internet to help identify the books we didn't recognize. The key was that all of the books had a color in their title, thus mapping their page number in the booklet to a color. The number grids were therefore paint-by-numbers (in the original sense!) artwork. Correctly identifying the art and reading the first letters in order gave the final answer. The big problem with this puzzle is that, distributed in book form, it was highly serial. With a team of four, serial is bad. I'd be surprised if any team didn't tear the book apart, if not for the identification step, then certainly for the coloring step. But the grid pages weren't numbered, and until you solve the grids you don't know that the order is important, so taking the book apart effectively destroyed a key bit of data. Plus, when something is presented as nicely as the booklet was, there's a natural resistance to disassembling it. Coloring books. I get it. Very clever. Now figure out how to distribute the puzzle in a way that helps solvers instead of hindering them. Aside from that gripe, the process was quite fun and I enjoyed this puzzle.

Monopoly: Located at San Jose's Monopoly in the Park, the world's largest Monopoly board, this puzzle involved solving crossword-style clues whose answers consisted of two word phrases-- a color matching a property group, and a four or six letter word. The second word was then broken into bigrams and the bigrams mapped to each property in the color group. Finally, a bigram sequence served as a guide to how to fill in an 11x11 grid, pointing us back to the giant Monopoly board to extract a final answer. Everything flowed pretty naturally despite the excessive flavor text. Our team really nailed this one, which always feels good. Neat location to visit, although I have to say that the world's largest Monopoly board isn't really all that large.

Candy: A bunch of different kinds of candy, each modified into its own puzzle that contributed an answer to a meta puzzle on a Tic-Tac-Toe grid. The nature of the meta required some pretty unsatisfying answers to the micro-puzzles, however, which severely detracted from the solving experience. And some of the puzzles were just... bad. One puzzle, for instance, consisted of a bag of M & Ms with a sticker on it showing a 4-digit number. Some of the M & Ms were normal. Others said "arhsall" and others said "athers". Ok, we get it, it's hinting us toward Eminem. Then what? Like every other team, we counted the candies in each color and tried to make sense of them, but we already had the answer-- Eminem. Just index each digit of the sticker into Eminem to get the answer, NEEM. Of course! Neem! Everyone knows neem is a large, semi-evergreen tree of the East Indies, right? Ugh. Neither the answer nor the path to getting there felt good, and some of the other candy puzzles suffered from similar problems. They were just too micro and terse. The meta came together nicely, but the cost of admission was rather high.

Finally, a note to the Shinteki crew. We always love the Shinteki schwag we get after each event, but after the genius of the clipboards from Decathlon 3, the Jenga sets this time out were a disappointment destined to gather dust somewhere. Perhaps more gear for the stylish Shinteki player in the future-- sunglasses, bucket hats, utility belts, insulated bags-- stuff we might use at future events. Or a giant SHINTEKI dark chocolate bar. Because who doesn't need more dark chocolate?

Midnight Madness: Back to Basics

Last weekend was the latest Bay area Game, Midnight Madness: Back to Basics, and possibly the final one to be run by Snout now that team captain Curtis is moving to Portland (although a Portland-based Game would be many kinds of awesome). As with their last Game, Hogwarts and the Draconian Prophecy, Snout hit the ball out of the park on theme and story. Midnight Madness was a cheesetastic Disney film from 1980, notably mainly for the screen debuts of Michael J. Fox and Pee-Wee Herman and the scene-chewing performances of virtually everyone else in the cast. There's little to recommend the film otherwise, except that the plot revolves around a puzzle-filled road rally and inspired Joe Belfiore to create the first incarnation of what we now call The Game at Stanford, and later again in Seattle.

This Game followed the basic plot of the film, and as with The Apprentice: Zorg which aped The Fifth Element, this proved to be a tremendous amount of fun. The route of the Game echoed that of the movie as much as possible (given that the former took place in the Bay area and the latter was set in Los Angeles), and many of the clues themselves took their cues from the film. It was fun to know that our next stop should be a mini-golf course, a diner, or a brewery, and sure enough, we wound up at one. The one-to-one mapping of Game to film has multiple side effects. It creates a narrative without one being explicitly laid out within the Game itself. It increases the payoff to some clues, their alignment with the movie increasing the sense of elegance and craftmanship of the overall event. It centers the player, giving them a sense of progress and advancement. Briny Deep has already decided to follow this model for our next Game, whenever that might be (and we know the film that will form our template).

It's unfortunate, then, that so many of the clues themselves were disappointments in one way or another. Many felt arbitrary. Some flat-out misled us unfairly. At least one was broken. If the Game's artistic program scored a 10, its technical merit only rated half that. There were few brilliant aha moments, no clues that felt revelatory, no intriguing handouts or manipulatives, and nothing that felt truly fresh. Snout used completely standard, off-the-shelf puzzle forms more than once. The clues often felt like afterthoughts, rushed together because something was needed rather than crafted for their own sake.

The game began with a delightful a cappella rendition of the Midnight Madness theme song, but instead of a tear-open-the-clue high-energy start, Snout opted for a Midnight Madness pub quiz. Teams were called at random-- some getting called multiple times before other teams got called at all (which, while "fair" in a mathematically pure random-is-random sense, was not a great experience for teams waiting to be called). No team waited too long, and the gap was unlikely to mean much overall, but it was kind of a downer to be all geared up and ready to go only to stall out and have to wait our turn to answer a question correctly and earn the starting clue.

Start clue: The opening clue, just as in the movie, was a card with a few cryptic lines and a row of numbers at the bottom. The text was straightforward wordplay, and the numbers a simple decimal-to-hex-to-calculator-spelling conversion (if I remember right, the card read "249973 ==> 773d5", nicely suggesting what to do). We were gone in no time. We liked this clue-- it was easy, everyone contributed to cracking it, it mapped directly to the corresponding movie clue, and gave us good positive energy to lead off with. All of which got sapped at the next location.

Binoculars: The idea for part one of this clue was terrific. At this point in the film, teams went to an observatory and looked through the telescope to find the next clue. A bratty kid was using the telescope before them, however, to spy on women as they got undressed. This location was atop a hill with a panoramic 360 view for miles. Forewarned to bring binoculars (thanks to an eagle-eyed teammate who saw the information hidden in the Captain's Meeting presentation), we were able to use them to find two female silhouettes and accompanying data posted in the windows of far-off buildings. The problem was, nobody could find the third. And the GC members staffing the location didn't seem to know anything about the clue. I specifically asked one of them if we could see everything we needed to see from that spot at the top of the hill, and she said we could. I later found out that the third was only visible from a location below and to the side of the hilltop. The way the clue was set up, with a box (bearing a combination lock) at the summit, there was no reason to think we had to venture off the hilltop. Time passed. Team after team arrived, and none left. If I'm GC, at this point I make some kind of announcement about the general vicinity of the third data set. Maybe a 60-120 degree arc to narrow it down for teams. Perhaps a nudge, at least, that we'd have to leave the hilltop. But GC remained mum and allowed teams to collect there, shivering in the cold, frustration mounting. Finally, sunlight gone, they distributed tubes (simulated telescopes) with the data embedded.

About that data. Each set consisted of three equations, one atop another, along the lines of X-X-X, (X^X)/X, X*X+X, and so forth. Each set had a total of nine exes. We got excited at the idea of replacing each X with a different digit from 1 to 9, so that each equation solved to the same value. But that was wrong. Instead, we were supposed to replace the X with a single digit-- the same digit for every X in every data set. Then we were supposed to solve each equation and sum the results within each set. That would give us the correct values to use on the combination lock. There was nothing to indicate what the correct value of X was, or that we needed to sum the equations. With no way to confirm either the value or the approach, the puzzle was essentially intractable. It could have been solved with minor changes to the notation they used, adding a horizontal line below each stack of three equations to suggest a sum. Instead, most teams needed guidance from GC to hit upon the right approach. By the time we left this site, we were testy and disheartened. We didn't understand why GC hadn't provided help on the hilltop when NO teams were able to make progress, and we were crushed when the puzzle itself proved so arbitrary and unsatisfying.

Pianos: The film brought teams to a piano museum where the clue was the Pabst Blue Ribbon jingle painted on a tiny piano. We arrived at a GC member's home filled with pianos and were handed a bundle of strips on which musical scores were inscribed. Immediately Andrew, our resident musical prodigy, perked up as the rest of the team shrunk back. But as he played one of the scores and looked at us quizzically, the rest of the team brightened as we realized it was a commerical jingle. And so we set to, Andrew playing the music and the rest of us identifying the products. We made short work of it and were puzzling over what to do next, when Andrew noted that each of the scores had a mistake. Aha! I'd already sorted the music alphabetically by product, so it was quick work to copy the wrong notes onto a blank staff in that order and identify the Klondike bar jingle. What would we do for a Klondike bar? Apparently, we'd hop around like kangaroos while singing an incredibly bad version of the Friends theme. This was a great clue for us-- we destroyed it in record time, leaving well ahead of all other teams, thanks entirely to Andrew's musical ability. I shudder to think about what this clue would have been like for teams without musical aptitude. But for us, this was a fun, high-energy clue that tied in to the movie beautifully.

Brewery Nonograms: This was just a clue drop at a brewery, but even so the location was a little wonky-- instead of finding it behind the brewery as advertised, we instead found it in the alley beside the brewery. A small detail, perhaps, but when you're told to find the clue behind the brewery, you expect to find it behind the brewery. I was expecting some kind of block assembly puzzle (in the film, the (very lame) clue was on the side of cartons of beer, revealed as a forklift moved them into place), but instead we got a trio of completely standard Paint By Numbers puzzles. We divided and conquered. I got about halfway through one and knew it was going to resolve to a NULL symbol. When another puzzle solved to a CARD, I put them together to make CARDINAL. Then I looked at the partially-solved last puzzle and saw it was a coffee cup. "Is there a CARDINAL COFFEE in the area?" Sure enough. This clue worked perfectly well, but was nothing special. We were shocked to get standard nonograms, and explicitly opted to solve them by hand even though plugging them into a solver might have been faster. The rebus aspect seemed out of place, since the visual rebus in the movie came much later in the story.

Melons: Another great thematic fit. At this point in the film, teams are sent to a diner and told to look between the giant melons. A large-breasted waitress wore a necklace with a HUG ME charm, which anagrammed into HUGE M and sent teams to a minigolf course. when we arrived at the diner, we saw a large-breasted woman at the back of the restaurant. Upon closer examination [ahem], we saw she wore a necklace that said "HOT METER", which anagrammed into THE METRO. Tucked inside copies of The Metro newspaper in the diner's vestibule were a completely standard word search puzzle which, when all words were found, provided a message in the grid's unused letters. Once again, an off-the-shelf puzzle form with no twists. To their credit, however, the content of the puzzle was both thematic and fun-- a list of dozens of euphemisms for "breasts". There was much mirth in the van as we solved, with cries like, "I can't find PAWPATTIES!" Nevertheless, it was disappointing to find no hidden layer or extra depth to the puzzle. We also heard that at least one team found the Metro puzzles without ever going further into the diner to find the necklace, which is a shame.

Hitchhiking: At this point in the film the protagonists separate, and two of them hitch a ride with an extremely slow-moving elderly couple. Upon arriving at our next destination, we were met by a convertible driven by a pair of GC members dressed as old people. They invited us to go for a ride with them, and once two of us got in, proceeded to drive around the parking lot VERY slowly, while the rest of the team walked alongside the car. The two of them rambled on and on in that stereotypical old person way, getting tripped up on certain words that we needed to fill in for them. Totally fun and thematic way to gather the data, and the GC actors were terrific. Shame about the puzzle. One of the fifteen words in the list was SCRABBLE, and the narrative made a point of mentioning how RATTLESNAKE hit multiple triple word score spaces. So we immediately tried to reconstruct a Scrabble game with these words. But a little analysis showed that the letter distribution was completely wrong, and the first word in the list was too long to be an opening Scrabble play. Even so, the Scrabble vibe was strong enough that we kept looking for a way to make the puzzle Scrabble-related. No luck. The puzzle was much simpler and weaker. Completely unclued, we were supposed to notice that the first letter of each word appeared somewhere in the following word. Aligning the repeated letters in a single column revealed a message spelled in the next column. Huh? How exactly were we supposed to notice that? There was no context, nothing to guide us to that observation amid so many other potentially interesting properties of the words individually or the list as a whole. The first letters of the words weren't unusual-- there was nothing noteworthy about the first letter of EMBEZZLED reappearing in the next word. Start the list with ZERO, and populate the rest of the list with XYLOPHONE, QUESTION, JOURNAL, and the like. Make me notice the repeated letters. They certainly didn't pop from words like RATTLESNAKE, SCRABBLE, ALOE, and PITCHFORK. The Scrabble puzzle we invented as we solved seemed far more interesting than the puzzle we actually had.

Minigolf: Another location that tracked perfectly to the film, in which teams had to play through a minigolf course to discover a message hidden on the drawbridge on the 14th hole. Merely skipping to the end or browsing through the course wasn't enough to get the clue. So too for this clue. Each hole had a picture on it which, thanks to the iPhone, we gathered quickly and translated into a list of words. But then what? Nothing leapt out at us, so-- mindful of the corresponding clue in the movie-- we went back to the course. Two holes stood out. In one hole, as the ball passed underneath the lighthouse a recorded voice shouted "Fore!" In the other, upon entering the windmill a recorded voice said, "How about a game of air hockey after this round of golf?" Both seemed reasonable in context, but a trip to the air hockey tables still seemed in order. Eureka-- taped to the side of the table was a solving grid. But none of our words seemed to fit-- each was smaller than their corresponding grid row. We had to be missing something. What if there usually wasn't any recording at the windmill at all, and instead of changing an existing recording GC had added it? That suggested that they did the same thing at the lighthouse, which meant "Fore!" was important. Bingo. Each of the words in our list could be prepended with FORE to form a new word that fit the grid. This was a terrific puzzle from start to finish. We loved that the snack bar was open and we could grab some food. The fact that, as in the movie, you had to play through the course to get the information you needed was fantastic. The insights were very satisfying. My only criticism would be that once the place got more crowded, it would be very hard for teams to get the info from the air hockey tables without giving it away to other teams, and having that aha spoiled for us would have been a bummer. This was my favorite clue in the Game.

Radio Station: The next clue in the movie came from going to LAX and tuning in to the AM radio station that normally provides airport information. The times we live in make it impractical to put any clues near a major airport, so a train station filled in. Incongruously, the pointer to the clue was hidden on a lone Obama '08 sign on the lawn in front of the station. We might never have found it without calling GC, and I'm not sure why they chose that form, which was so unlike how we found clues in the rest of the Game. Regardless, we dutifully tuned our radio to the far end of the FM dial and identified a series of song pairs playing simultaneously in the left and right channels. The on-air bumper made a point of saying "It's Midnight Madness-- as in the movie, not the band," so we ignored the bands believing they didn't matter. Wrong! Every team we talked to were likewise mislead by this. Fortunately we called GC to verify our data and specifically asked for confirmation that the artists were irrelevant, so we didn't spend too long looking at the wrong data. Since the songs were presented in pairs, we knew we needed to combine info from the left song with info from the right. The proper way to do so was arbitrary and unclued. For each pair, we had to notice that one syllable of the song title on one side was the same as one syllable of the artist from the other (eg, Adam SANdler and SANta Claus is Comin' to Town). Even when someone suggested it, it sounded wrong to me because it was so arbitrary and messy. The other half of the data-- the other artist and song title-- was completely unused. The overlapping syllables weren't in consistent places, such as the last syllable of the left title and the first syllable of the right artist-- they were random. The whole effect was deeply unsatisfying, not so much a puzzle as "guess what we're thinking."

Hare Krishnas: In the movie, the next clue was disguised as the literature distributed by Hare Krishnas in the airport. Here, a couple of GC members costumed as Hare Krishnas pressed their literature on us as well. We later found out that only three teams were given this clue (the rest were skipped over it), which was probably a good thing-- it required a high level of attention to detail which wasn't easy to apply at that time of night. We received multiple copies of a religious screed full of typos. Close examination revealed that the copies weren't identical-- while some typos were shared, others were not. We had to find all the unique typos and highlight their locations on a master sheet. Those highlights formed a very good rendition of the Greyhound logo-- our next stop. This was a grind-- once we knew what we had to do, it took quite a while to actually do it. On the bright side, it lent itself well to parallelization and cooperation, so it was at least a good team puzzle. But shorter would have been better.LOLCats: At the Greyhound station we found a stack of LOLCat photos with edit marks in the margins. Obeying the edit marks allowed us to extract certain letters from the LOLCat text to get our next destination. I say "we", but I checked out on this puzzle and grabbed a few Zs while other pirates huddled in the back and forced an answer out of the LOLCats.

Pinball City: In the movie, Michael J. Fox plays a Star Fire videogame until he "beats" the game (which wasn't really possible), triggering a custom video telling them where the finish line was (also not possible). The house of a GC member stood in for Pinball City. No pinball machines, but three computers were set up running Star Fire via MAME. The ROM had been hacked to produce some incongruous sound effects under certain conditions. We needed to identify the videogames those sounds came from and, by observing the scores when those sounds got triggered, put them in the proper order and enter their initials into the high score screen. A for faithfulness to the film (although achieving a certain score would have been more accurate and, frankly, more fun), but much lower marks for the clue itself. Again, this felt arbitrary, and a long way to go for "name these three videogames". What if nobody on the team recognized them? Worse, entering the correct answer triggered a video that everyone in the room could see. This puzzle was only solved by a couple of teams-- everyone else just rode the solvers' coattails and eavesdropped on the video. Blech. Our team saw the video when another team solved the puzzle, but some of us felt dirty about leaving the site without having "earned" it. We had a little internal debate about it, but ultimately we decided to stick around until we figured out the right approach and solution ourselves. Making solvers wear headsets and providing key info through audio would have been one way around this problem, although there was no real way to prevent players from seeing the correct letters get entered into the high score board. Ultimately, the free ride was a better solution than some kind of turn-taking system would have been, but redesigning the puzzle to remove the problem would have been even better.

Hissy Fit: To reflect Michael J. Fox's character jumping out of his brother's Jeep and running away when he felt unwanted, we had to send our whiniest team member away, then try to entice him back via a cell phone game of Mastermind. We could only talk in 4 word sentences, and our teammate's response was dictated by the number of "correct" words we used. The magic phrase was "Jeff, you are special." You'd think that with an almost infinite domain space it would be exceptionally hard to zero in on the right words to say, but we locked on the "Jeff, you are" within about 5 minutes. Some fun playing around with filling in that fourth blank ensued, until someone hit on the right word. Amazingly, three teams-- none within earshot-- solved this puzzle within about 5 seconds of each other. We've had Mastermind puzzles before, but this was a fun twist.

Don't Get Hammered: This was a perfectly good puzzle wrapped in a frustrating form factor. Each of six inflated balls had about 14 pieces of data on them. Only two players from each team were allowed on the field at once, to gather the data or bat the balls toward the sidelines so teammates could read them. Meanwhile, GC members wielded inflatable hammers; when tagged, a player had to leave the field and tag in a teammate. Sounds chaotic and fun in theory, but was more chaotic and frustrating in practice. For starters, most of the balls were quickly punctured and deflated. There was a huge amount of data to gather, and strategy only got you so far amidst the chaos. Once we had the data, we completely blew the analysis phase by using the Post-Its GC provided instead of doing the smart thing and going directly to Excel, which is what we ultimately converted to. Once the spreadsheet was fired up, sorting the data into sets and putting each set in the right order fell out quickly-- hooray for the iPhone! The puzzle would certainly have been too simple had we just been given all the data, but this particular method of gathering the data was, I think, just a little too wild for my taste.

So where does that leave us? Overall the clues were disappointing-- there was too much unclued arbitrariness, too many instances where, in the course of solving, we created a more interesting puzzle than what we were given. There were too many opportunities for teams to skip their own ahas and get spoiled by the progress of other teams. On the other hand, the tight binding to the film made Midnight Madness: Back to Basics a lot of fun and solidified my belief in that model of Game structure. Snout has a lot of talent in acting, performance, and theatrics that was showcased quite well in this Game, and I'm glad I got the opportunity to play.

Shinteki Decathlon III

Another year, another Shinteki. This was the third Decathlon so I won't belabor the format-- see my report on the first Decathlon for details.

As usual, the organizers continued to tweak the format to good effect. This year we were told in advance when certain sites would close, so we could budget our time accordingly if we wanted to make it to all the clues. Partial answers now awarded partial credit, and some hint prices decayed over time so that the longer you worked on something, the cheaper it became to get unblocked.

As in the last Decathlon, each puzzle had a hidden bonus puzzle associated with it. Some bonuses were given in plain sight, others were hidden at the clue location, others had to be ferreted out by careful investigation, and still others were merely implied by the clue data. Each bonus was 15 points (regular puzzles were worth 100), and gave teams something extra to do during drives or if they found themselves idle at a clue site. I thought these were a great idea, and they were executed much better than last time, but I'd have liked to have known in advance if a bonus was obtainable only at the clue site (as at sites 4 and 7) so that a) we could decide to look for it before leaving, and b) we wouldn't waste time searching for it later. Notes for Decathlon 4.

I thought this event was solid from every angle, but especially in its theming. There was no story, but the overall theme of "time, space, and multiple dimensions" came through in virtually every clue, often in clever ways. As always it ran smoothly, the puzzles were fair and entertaining, and the overall vibe was mellow and social. And sending six of the 24 teams (three of which finished in the top 5), Seattle was there to Represent.

Sprint: "String Theory": Three teams were tied together by colored ropes, and everyone had to work together to untangle themselves. We might have finished sooner, but one team in our group was short-handed and so a fourth member was supplied by the organizers and instructed to just stand there and not help. Having one person anchored in place is handicap enough, but when that person is six foot fourteen, comedy ensues.

Knowledge: A stack of Trivial Pursuit-like cards. The questions on each card resolved to either a synonym of antonym of one of the Decathlon events (KNOWLEDGE, AIM, etc). The Shinteki symbol appeared on the back of each card, with each color of the symbol matching the color of a question category. Coloring in only those sections matching the colors of questions resolving to synonyms, and leaving the antonyms blank, created (highly stylized) letters spelling the answer. The trivia fell easily and we found the bonus in no time. We realized each card had two groups of answers quickly, but the puzzle answer eluded us. We took longer than we should have because when we first tried the correct approach we didn't see them as letters, so we spent a lot of time trying to make the data into binary values.

Teamwork: A set of four mini-puzzles at Golden Gate Park. Three were easy: a set of digital clocks with their states inverted, a word puzzle with a spiral grid and overlapping answers, and a table holding objects whose cross-sections form letters. The stereocube-- four stereograms taped to each side of a central square, was what gave us trouble. The images popped easily for me, and gathering the data-- each stereogram showed a 4x4 grid with circles at four different heights, spelling the word DOWN-- was simple. Interpreting it, on the other hand, was a problem. It was presented as a cube, which suggested to us that the four sides would work together somehow. We wanted to project the depth information back into the cube, but nothing we tried made any sense. Once we finally caved and took a hint to tell us to treat each side separately, we were able to see that if each side was a physical object and you looked down on it from above, the projections formed another set of letters. We were essentially doing the right thing, but we were only looking at the combination of all the data instead of each side separately. By the time we got to the final puzzle, a series of clue pairs to phrases with POINT, LINE, PLANE, SPACE, or TIME in them (representing 0-4 dimensions) our time for the entire site was nearing expiration. We tried creating a 5x5 alphabet grid and using the values as coordinates, but when that didn't work we took hints rather than waste more time. I'm not sure if we'd have hit upon the right approach-- base 5-- on our own, since base 5 is rarely used in puzzles (but it's totally fair game). So we finished this site wondering when we were going to bring our game, because we'd clearly left it somewhere else thusfar.

Enigma: Volunteers grilled hot dogs for everyone at a park while we solved a contraption made of PVC pipes holding various colored marbles. By picking up the object and rotating it freely, we could see the marbles pass by small holes in the pipes. Each marble was locked into a letter-shaped subsection of the maze-like device, and we needed to figure out those letters. This was really just a two-person puzzle, with one person rotating the maze and another inserting a pen into a hole to trap marbles as they rolled past and record the data. But while two people worked the gizmo, the other two could eat. Andrew and Dave developed a system and got irritated when Jeff and I tried to help, so I broke out TEA and started plugging in letters as they got found. The order of the letters was given to us by a string of balloons nearby, so once we had 5 letters we fuzzed the other 3. We might have been done sooner, but the orange and red balls got confused under the gray skies and our data was corrupt until we sorted that out. I didn't love this clue. Physical puzzles are always good, but this one suffered from two main problems-- it wasn't possible for all of us to work on it at once, and there wasn't much of a puzzle. We knew from the get-go what we needed to do, we just had to go through the process and do it. That's fine when the process is fun or interesting, but this was more tedious than entertaining. Something bigger that required two people to manipulate, one to trap marbles in holes, and a fourth to record the data would have been more successful, I think.

Classic: A packet of four mini-puzzles and a bag of interlocking plastic cubes. Each mini was a three-dimensional variation on a classic puzzle form: a maze, paint-by-numbers, minesweeper, and crossword. Solving each was fun in itself, and then we needed to recreate each solution's shape with the cubes and fit them together to determine how to use the leftover pieces to make a final shape to complete the cube. That final shape, when viewed from three different angles, formed three different letters (HAT). Very Gödel Escher Bach. I really enjoyed this clue. The minis were fun individually, the cubes were fun to play with (and we each took home a set!), and the finish was both elegant and thematic. Our team crushed this one, and finally seemed to be gathering some momentum.

Orienteering: Located on the top of a hill with a stunning 360-degree view of the Bay area, this was a great use of the environment. This hilltop featured a number of ~8 foot concrete circles in the ground, each one crosshatched into a 3x3 grid. Seven of them were numbered by GC, and then further annotated with many numbers in tri-colored chalk. Each number was oriented toward one of the four cardinal directions of the grid. The puzzle was called Conferencing, and given grid on each circle it wasn't hard to leap to phone-spell. We were decoding in minutes. At each circle, you had to stand on each side of the grid and look at only those numbers that were right-side-up from that vantage point. Each orientation contained an unbroken sequence from 1 up. The location of the number indicated a phone key, and the color indicated a letter from that key. Each circle therefore yielded a four-word clue, like CRUSTACEAN FRIED INTO PATTIES (CRABCAKES). Highlighted letters in the answers spelled BUTTONS. This, to us, seemed like the final answer, but it was just a partial. We were stumped about where to go next. Fortunately, the next hint became free after a few minutes and got us on the right track. Each answer contained the letters from a phone key (CRABCAKES). Using different letters from the same keys, we could form a different word which was the final answer. This last step was certainly tightly related to the rest of the puzzle, but since the entire puzzle involved phone buttons the partial of BUTTONS felt more like a final answer than a clue. I'd have liked it much better if the clue had been something like FIND BUTTONS or USE BUTTONS. That said, this was a terrific puzzle and an excellent example of how to incorporate the features of a location into puzzle design. We rocked it.

Wild Card: Masters of Space and Time. More an activity than a puzzle, this site had teams split in half. The masters of time needed to count to 100 seconds without the help of a chronometer. The masters of space had to walk to a target blindfolded. The former was pretty easy for Andrew, our man with rhythm, who used Stars and Stripes Forever as his mental metronome. The latter would have been close to impossible if the target hadn't been set up on the edge of a patch of dry grass, which let us "feel" our way to the right place. Meh. It was what it was, and I was happy to move on after each half of our team got a perfect score on our third tries.

AIM: Though we were given a copy of the game Laser Battle with this clue, we wound up not needing it. Cards depicted a game board configuration of mirrors and an arrow showing where the laser was firing. We needed to add a mirror somewhere to create a 10-bounce sequence ending at the indicated target. The locations of these mirrors mapped to letters on a final grid, spelling FLAGS OVER X. There was only one X on the letter grid, and each board's laser path crossed it. Treating the laser beam at that spot as semaphore gave us our final answer. We solved this in an hour at Red Robin, including eating time. Since the puzzle was so solveable without the game, I think including the game formed more of a distraction than an asset and made the whole thing seem less elegant.

Manipulation: Construct a polyhedron from a set of numbered squares and triangles. Put on the supplied 3-D glasses and see a path connecting the faces. Read Braille on the tabs connecting the faces in the order of the path. Done. This puzzle was something of a let-down. There was nothing about the 3D in this puzzle that couldn't have been done with a stereogram (which we'd already seen earlier in the event), so it felt anticlimactic. If you're going to use 3-D glasses, the payoff should be better. The path was visible even without glasses, so the 3-D effect added nothing to the puzzle. This was another case of being told exactly what to do and then just needing to plug through it, with very little creative thinking required. This would have been a great opportunity to hand us 3-D glasses and tell us to review our past puzzles, where we could have discovered 3-D data popping out all over the place that we hadn't been equipped to see before. The puzzle just wanted some better reason for using 3-D.

Endurance: Two lists of clues for 9-letter answers, two fill-in grids. The answers in each list were in alphabetical order, an essential aid in disambiguating possibilities. Once the grids were filled, nine words remained. Flavor text suggested these words would go "between dimensions" and connect the two grids, and sure enough at nine key intersection points you could bridge the grids with the leftovers-- some running from grid A to B, others from B to A-- and read the letters in the third positions of the resulting dimensional bridge. A solid, thematic puzzle, you say? Absolutely, but it gets better. All of the answers that fit horizontally had exactly one D. All the answers that fit vertically had exactly two Ds. And the nine answers bridging the grids had exactly three Ds. A brilliant bit of highly constrained puzzle design, to not only find enough nine-letter words with the correct properties but to arrange them so the grid could be completed unambiguously even if the solver didn't know about the 1-D / 2-D / 3-D property. Our time management fell apart here-- as we neared the end of the event, we should have started taking hints to ensure we finished in time. Instead, we were so focused on solving that the idea didn't occur to us until it was too late to capitalize on the information. We finished about 30 minutes after time expired. If we'd taken the hints right away, we might have solved it in time, which would have given us the points we needed to take first place. D'oh! After all the spatial puzzles, I was thrilled to finally get a meaty word puzzle squarely in my wheelhouse and only wish it had come a little earlier in the event so we'd have had the satisfaction of finishing before the deadline. Despite its familiarity-- it was essentially a standard fill-in puzzle with an impressive construction constraint and a final twist-- this was probably my favorite puzzle of the event.

Kudos to Just Passing Through for another terrific event. It's been years since Jackpot, folks-- aren't you itching to run another full-length Game? We'll be the first to sign up.

PiratesBATH

On June 9-10, the gf and I went to sunny CA for PiratesBATH. Although she'd played in past Puzzle Hunts, this was the gf's first Game. She didn't play with Briny Deep, however, but with The Bonny Wenches, a new team comprised of various lady friends of Briny Deep. I'll not comment on the Wenches' experience, since I wasn't in their van, except to say that at least two of them, including the gf, had enough fun that they're now talking about playing in the next Shinteki event in August.

This Game distinguished itself from others in three main ways. First, most of the main puzzles were contributed by the teams themselves, each of whom was invited to submit one for inclusion. The advantage for doing so, aside from having an opportunity to impress fellow teams, was that when you encountered your own puzzle you'd get to skip ahead to the next clue immediately. Second, clues were provided not via live phone support from GC or a PDA, but an ultra-low-tech scratch-off and envelope system wherein teams purchased pre-canned hints for points. Third, there was no overnight leg; instead, teams roughed it at a campground (we were told in advance to bring tents and sleeping bags).

Things have been so crazy busy since returning from the Game that it's taken me this long to find time to write about it. So please forgive me for resorting to bullet lists.

What I liked

P&A Magazine Issue 9

Nobody seems to talk about P&A Magazine anywhere, so... consider this the official thread for issue #9, which was released last weekend. Please be sure to preface any spoilers with a warning.

The gf and I are starting to work our way through it, and while we've solved puzzles 2-5, 7, and 8, we're currently stuck on 1 and 6. We've got the center hive filled in for #1, but haven't extracted anything from it. We have all the answers for #6 and have done the next obvious thing, but can't figure out where to go from there. We haven't yet started on the others.

We've stared at our data for #6 for quite a while and have no inspiration-- anyone have a nudge to give?

Shelby Logan's Trial

Last Sunday night I landed in Seattle around 11:30 PM after a weekend of sleep deprivation, and seven hours later I headed back to the airport for a day jaunt to Las Vegas to testify in a civil case against one of the organizers of Shelby Logan's Run. For some reason the idea of flying from San Jose to Seattle and then to Las Vegas seven hours later seemed better to me than taking an extra bag with me to No More Secrets and flying directly to Vegas from San Jose. It actually worked out fine, and I can report that the desert is far more allergy-friendly than California.

Shelby Logan's Run was a Game run in Las Vegas in 2002 by some Microsofties. It was, in many senses, the Game to end all Games. While this event had puzzles, the focus was on over-the-top experiences (and where better to offer them than Vegas?). In the course of the event some or all of us camped in a dry lake bed during a torrential thunderstorm; powerboated and scuba dived on Lake Mead; shimmied up a rock chimney; captured, cared for, and ultimately scanned a living rat; fired a semi-automatic weapon; drove ATVs across sand dunes in the black of night; rode a free-fall ride atop the Stratosphere tower; performed a song in drag at a gay bar; got pierced ears; explored an abandoned prison by flashlight; and more. It was my first Game, and no Game since has delivered anything close.

The Game ended prematurely when one player fell thirty feet down a mine shaft and became paralyzed from the neck down. A clue sent teams to a site where there were multiple abandoned mines, and in plain, unencrypted text told teams to enter a specific number and no others. This player entered the wrong mine (without a flashlight, I believe), and fell. A very real tragedy.

Inevitably, perhaps, lawsuits followed. I don't know who exactly sued-- the player, his family, or his insurance company-- but all organizers of the event were named in the suit, and all but one settled out of court. The last holdout finally got to trial, and I was asked to testify for the defendant which I was only too happy to do.

Every player signed a waiver when they sent in their fee to participate. A scary waiver. It explicitly called out that players might be called upon to perform strenuous activities (and listed many examples), with possible consequences including death. I remember talking about that waiver with my teammates before signing it-- it was hardcore. I don't know how that waiver holds up under Nevada law, but it wasn't vague and it wasn't perfunctory. I took notice.

I was in the van when the unfortunate player's team arrived, and the defense wanted me to testify as to their behavior and to provide the jury with a first-time player's perspective about the Game. I agreed for many reasons, the most important of which being philosophical-- people in our society don't take enough responsibility for their own actions. Were there things the organizers could have done to prevent the accident? Yes. But ultimately, the tragedy was the man's own fault. Americans don't like saying that. We like pointing fingers and finding someone else to blame. But every single player signed that waiver. They knew the event involved operating on very little or no sleep. They knew physical activity was involved.

Earlier in the event I drove an ATV at night and opened up the throttle a bit-- until I hit the next dune. I sailed over the crest and my headlight illuminated... nothing. I had absolutely no idea where the ground was. I could have been catapulting into an abyss for all I knew. It was terrifying. Not in the casual sense the word is commonly used, either-- I mean heart-stopping, pit-of-my-stomach, images-of-snapping-my-neck raw terror. When my wheels touched down, I immediately eased up on the throttle and took a safer, more sedate pace. I took personal responsibility for my own safety. Nobody told me how fast to go. That was up to me. I chose the level with which I was comfortable.

Players were given specific, explicit instructions about where to go at the mine site. What happened was terrible and tragic, but ultimately someone didn't follow instructions, went somewhere he'd been told not to go, did so alone and entered a dark tunnel without a flashlight. People in our society need to accept more responsibility for their own actions, even when those actions are tragically wrong. And in this case, I didn't believe the event organizers should be held responsible.

Philosophically, I wish that all the organizers had gone to trial instead of settling. I understand the desire to just have it all be over with, though, and not wanting to endure the stress or risk of a trial. The one defendant who went to trial was mainly responsible for programming the hand-held electronic device used throughout the Game, which had nothing to do with that particular clue site. My understanding is that, while the plaintiff attacked the waiver, the defense strategy had nothing to do with it but rather that the defendant simply had no part in planning, organizing, or executing that particular clue or clue location. Yesterday I found out that the jury returned a verdict that the defendant did not act negligently, which I assume means he's off the hook.

I understand the plaintiffs are still going after the owners of the mine, and there I think they have a much stronger case. Why on earth wasn't that mine shaft sealed? It seems so obvious. And so, while I think the plaintiff bears responsibility for what happened, the mine company unquestionably shares in it. I hope the plaintiff has better luck going after those deeper, and more culpable, pockets.

No More Secrets

I spent last weekend in the SF Bay area for No More Secrets, a Game ably run by first-time hosts Coed Astronomy. You may recall that at about the same time last year, I went down there for the Paparazzi Game and discovered that I was allergic to California. A few months later, however, I was fine during Hogwarts, so I hoped it was just a fluke.

No such luck. The reaction wasn't as brutal this time, but by midnight my nose was running freely and I was not a happy camper. I went through three travel-packs of tissues by the end of the Game. This has me very nervous about PiratesBATH three weekends from now. Are there any allergy remedies that work? I tried some pills (Claritin, maybe?) during Paparazzi, but they had no discernable effect.

Allergies aside, No More Secrets was a well-run, entertaining Game. It was particularly eventful for our team's captain, Jeff, who collaborated with the organizers to incorporate a marriage proposal into the Game's introduction. All the captains were asked to line up in a particular order, and then in unison reveal a page from their starting packets. Each page depicted a giant Scrabble tile which collectively spelled "JESSICA WILL YOU MARRY ME" with Jeff holding the "ME". It took Jessica a few moments to realize she was the one being asked, but she quickly said yes as Jeff got on one knee to present the ring.

Coed Astronomy took our Pit Stop idea from The Mooncurser's Handbook and tweaked it, replacing an assortment of activities with a single linear track of paper puzzles. I'd call the approach a success. It accomplished GC's main goal of slowing down leading teams while still providing them with something fun to do, and as a player it was often welcome to have a place to hunker down for a while and collaborate on puzzles in a setting more conducive to group-solving than a van. The final pit stop was, I think, an unfortunate choice, creating a sense of anticlimax. You really want the event to end with a triumphant surge of van-based puzzling, giving teams the sense of closing in on a finish. Instead, teams wound up staying at one location for hours, first doing a clever and fun series of puzzles (see below) and then killing time at the pit stop before moving on to the wrap party. Briny Deep was inexplicably on fire at this point, burning through the pit stop puzzles with criminal intent while esconced in comfy chairs, so in that sense we didn't mind...

I wasn't crazy about the answer submission system, which had teams phoning in the answer, receiving a new code word in response, and entering that code word into a laptop app to receive instructions on where to go next. It enabled GC to tweak the route and timing on the fly, so I can understand why they chose that approach, but the two-tiered system felt cumbersome. I suspect I'd have objected less if the app was on a Palm instead of a more-awkward-to-carry laptop.

The event's theming and story did nothing for me. The narrative was more of a distraction than a feature since it had no bearing on anything we were actually doing and wasn't particularly compelling. Narrative is hard-- I've only seen two games, The Apprentice: Zorg and Hogwarts, really nail it-- so that's not a knock against Coed Astronomy.

As in so many Bay Area Games, most of the locations were just clue drops with no correlation between the site and the puzzle, either thematically or mechanically. There are some teams who claim not to enjoy conference room-based puzzle events, but I'd argue that if the locations don't matter I'd rather just stay in a conference room. What's the point of sending me from one place to another if I don't need to interact with the location somehow, either in discovering the clue in the first place or solving it once I've got it? Why not just let me sit in a comfortable environment and solve a serialized set of puzzles instead? For Mooncurser's, most of our puzzles were tied in some way to their location. Players bowled at the bowling alley, watched a movie at a movie theater (and got a science fiction movie / theater - themed puzzle at the Galaxy 12 theater), traversed the corn maze, used statues as decoders at a sculpture garden, assembled a sign post at a sign post (well, they would have it they hadn't been skipped over it for time), played food Boggle after lunch, floated down a river to decode nautical flags, solved a song titles puzzle at the Experience Music Project... the locations mattered. Sometimes they can't. Sometimes you just need to get people from point A to point B, and the drive is too long so you need to put a clue somewhere between them to break up the drive. I get that, and that's fine. But most of the locations in a Game should, IMHO, fall into one of two categories: someplace cool or unusual and therefore interesting in its own right, or someplace thematically or mechanically linked to the puzzle found there. Otherwise, what's the point?

GC was friendly and helpful throughout, and we had no glitches either in puzzles or in logistics. We had a great time, and greatly appreciate all the effort that Coed Astronomy put into planning and running the event. I know how much work it is. Thanks!

A clue-by-clue rundown after the jump...

Punch Cards: A series of numbered punch cards with crossword clues. Each answer contained a directional word (UP, LEFT, SOUTH, etc). Tracing the path of the directions on each card yielded letters, and reading the letters in numerical order of their cards provided the answer. A nice first clue-- what to do was immediately apparent, and the form factor made it easy to parallelize and get everyone involved.

Triangles: The basic idea of this puzzle was similar to one I created for Mooncurser's (which was in turn adapted from a puzzle designed by my friend Aaron Weissblum), so I was initially quite psyched to tackle it. A bunch of triangular tiles were divided into thirds, with each third containing a terse text clue (UNIT OF WEIGHT, ANIMAL SOUND, BEAN). The triangles assembled into a bigger triangle by matching edges whose clues are homonyms/homophones of each other. Unfortunately, we matched CLOCK SOUND and BIRD as CUCKOO (instead of TICK and... CARDINAL, I believe), which caused us to have a couple of tiles that we matched only loosely or not at all and just shrugged off, figuring we had enough correctly assembled to get us started on the next step. Then by the time we got into the next step and got a mess of lousy data, we'd forgotten that we had some fudged connections. So we wound up blowing a good chunk of time on this puzzle unnecessarily. I didn't like the fact that once the thing was assembled, the clues on the outer edges didn't need to be solved in the same way that the inner clues were-- there was a completely different encoding system that kicked into effect at that point. That felt inelegant and cost us a bunch of time. Sigh...

Rock Wall: Our team went into the main entrance of the gym, where a climbing wall wth lots of interesting stuff marked off in tape awaited. The tapes were all different colors, and some were numbered and labeled with jazz and rock singers' names. As little girls had some kind of activity class around us, a bunch of grown men in pirate outfits hunkered down to try to figure out what was going on. Meanwhile, no other teams were there. We took photos of the wall and copied down as much data as we could, but feeling increasingly uncomfortable about being in the way of the kids, we exited and called GC. Turns out we were supposed to be around the corner, where a side entrance brought us to a room full of climbing walls. I got into a harness and climbed one of them, but Dann was much faster at it so we just let him do all the rest (my fingers were locked into claws after just the one climb). Each climbing path had a circle of pegs with words on them at the top, each circle corresponding to a rock we were given. Once we collected all the data we made short work of the puzzle, which involved finding the pair of words in each circle that could be anagrammed into the matching rock (AMETHYST, GRANITE, etc) and then reading those two words in the circle as a semaphore letter. I was amazed at my teammates' knowledge of rocks-- left to my own devices, I'd have had to resort to Googling a list of rock types and searching for likely anagram candidates.

Chess: At the Computer Museum we picked up an enormous chess-like board and a page of images grouped into sets. We quickly realized each image represented a six-letter word that could be found in a 2x3 section of the board, and that each section contained a chess piece on one of the squares thereby giving us a letter. So we got the initial message in short order. And even though three of us independently at some point said, "Hey everyone, 2x3 means Braille" we couldn't make the necessary leap of merging all the 2x3 grids from each image set. There were too many other approaches that seemed possible for us to converge on the right one. The puzzle would have been much stronger had it internally clued the merging step. There was also no reason for the words to be arranged in different patterns within the 2x3 grids-- some were simple loops, others were in rows, others were in S-shapes, etc. That seemed like signal, but it wasn't. All the words should have just been simple loops. We burned a lot of time here before getting nudged onto the right track by GC. This puzzle could have been really nice with a little more refinement.

No Morse Egrets: An audio clip gave crossword-style clues which solved to phonetic approximations of various team names (including ours: TIMBER KNEE DEEP!). The timing on this drive-and-solve worked out perfectly for us, since we solved it just as we pulled into the destination town, and everyone on the team enjoyed pitching in. A good group puzzle.

Pipes: A walk through a lovely covered bridge produced a word association maze. The hardest thing about solving this was doing so outdoors, where gusts of breeze threatened to blow away the pieces. I would have enjoyed a larger, more complex version of this puzzle, but word association puzzles are smack in my wheelhouse.

Phone Mastermind: The puzzle itself was entirely pedestrian and straightforward-- playing it by phone didn't alter the basic Mastermind process. Not sure why they bothered with this clue at all. The location, on the other hand, was superb. We loved the blue whale skeleton outside the marine center, and the view of the ocean was pleasant enough that we took a few minutes to just enjoy it.

Numbers: It's always satisfying to be the guy who cracks a puzzle, so when we sat down to look over the list of clues that came with this crossword grid and I realized that they made sense if you put a number in front of them, I was pretty happy. How we managed to not think of treating the values in the table as X and Y coordinates boggles my mind. Unlike the chess puzzle, though, this one was totally our fault. A nice puzzle.

Lego: Another cool location, out on the beach with a shipwreck at the end of the pier. The puzzle was also a great concept-- given a small set of Lego pieces, match each overhead schematic to the correct side view, assemble the pieces properly, and view from the front to see a letter. Except they picked a word that was too backsolvable. Once we had HIP????, we immediately called in HIPSTER. All this puzzle needed was a more carefully-chosen answer. Assembling the Legos was fun, but if a team can short-circuit a puzzle, they will. We'd rather have needed to assemble every letter.

Bank Heist: A huge disappointment. This clue had been built up over the course of the preceding clues. First we received a blueprint of the "bank" we needed to break into. Then we received a couple of transparencies that overlayed the blueprint with additional information about security measures we'd need to overcome. All good, and it seemed to be setting up a nifty scenario. But the reality was lame. Talking our way past the guard required no particular approach-- there was no puzzle to it, it was just basically wheedling until he decided to let us in. Once inside, we had to fish something out of a jacuzzi, which just wasn't very hard or interesting. Finally, the "motion sensor" we had to avoid tripping was a person with a squirt-gun who would fire if our operative's belt of bells made a sound. Turns out you can move pretty quickly in such a belt without jingling. The whole thing was an enormous let-down after the multi-stage build-up. There was no puzzle content, and the physical challenges weren't challenging. There was also a flow problem, in that many teams arrived more or less concurrently, which didn't help matters.

Vowels: This puzzle at the candy shop was one of my favorites (but then, I love word puzzles). Pairs of clues solved to words where all the consonants were the same, but the vowels might not be (e.g. ARENA/URINE). Treat AEIOU as a 5-bit sequence and generate a 5-bit value for each pair of words by turning a bit on if the corresponding vowel appears in the second word of the pair and not in the same position in the first word. If the vowel doesn't appear, or appears in the same place in both words, the bit is off. That gives you values in the 1-26 range, and therefore letters, giving new words. These words can also have their vowels changed, and so on, recursing to a final answer. A solid, fun paper puzzle.

Game of Life: A nice example of interacting with the environment. At this building, a large exterior window looked into an office where, on the far wall, computers were running identical programs. The top of the screen had pairs of monochrome pixel patterns arranged in three pairs, with colons between each pair. The rightmost pattern changed every second. Below those was a scrolling display of 5x5 monochrome patterns, followed by a set of additional patterns in other colors. We also received a PC "sequencer" app that showed us a single 5x5 pattern and three empty 5x5 grids. The grids were interactive-- we could toggle cells on and off, and a VERIFY button would presumably tell us if we had toggled everything correctly.

We'd been forewarned to know about cellular automata, and so we immediately recognized this as a Game of Life puzzle. I left my teammates to get busy while I used the bathroom. Some time later when I emerged, my team had gathered all the images and entered them into Excel, but hadn't figured out what to do with them. They'd tried evolving the first image in the app according to the rules of Life, but that didn't work. There were no obvious relationships among the patterns. I asked them to get me up to speed on what they knew and what they tried. The images at the top of the displays were obviously a clock, so we had the numbers 0-9. When they told me that the clock patterns were 4x4, I immediately knew what we had to do (although it really made no difference-- they could have been 5x5 and the puzzle would have been the same-- but for some reason the fact that they were smaller helped me jump to the right conclusion). There are only 4 rules in Life, which can be summarized as "If a [living | dead] cell has [x] living neighbors, that cell becomes [living | dead] in the next iteration" where x can have multiple values. The 4x4 number patterns were our primer. We had a sequence of 10 patterns, 0-9, and we had to use them to figure out the rules for a customized Game of Life. Once we had them, we could then correctly evolve the single pattern provided in our sequencer app. Once we did, it gave as a sequence of 26 5x5 patterns: an alphabet. This let us translate the sequence from the office into an email address. When we sent an email to that address, a message flashed on the screen in the office with our name and a bunch of colored symbols similar to the ones at the end of the office message. These symbols corresponded to the body of our message. We sent the alphabet next, discovered the symbols were essentially phone-spell (each symbol mapped to letters on the same phone key), and got the final answer.

I thought this was a super puzzle. There was a nice "aha" moment in figuring out what to do, and then executing it was fun and satisfying. I think it's hilarious that practically every team sent the alphabet as their message, which is undoubtedly the WORST possible choice, since any quick-thinking competing team could see that and short-circuit the entire puzzle. Better would be something like THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG, which gives you the whole alphabet but isn't as guessable. In practice I don't think any teams caught on to that and short-circuited, but the point remains.

Gibberish Songs: We powered through the next clue, a list of song lyrics turned into gibberish by replacing each word with a word that rhymed. A veteran of You Don't Know Jack's "gibberish questions", I made short work of this by simply reciting each word without the opening consonant. The whole thing couldn't have taken us more than five minutes to solve. That kind of success is fun to have, but I can't imagine it's what GC expected. A 5-minute puzzle just isn't worth the effort to create, test, and deploy.

Red lights: A large 5x5 grid of blinking red lights was positioned on the Stanford campus. Like everyone else we recorded the patterns and returned to our van. Life was still on our brain, but try as we might we couldn't get the patterns to evolve into anything interesting. We called in to GC to verify that Life was not the correct approach, just so we could get it out of our system and open our minds to other possibilities.

This was a puzzle that just should not have worked. Once Life was eliminated, there were no other clues as to what to do. All we had were a sequence of 5x5 patterns, and it was time to play "guess what the designers were thinking." But I did, in fact, hit on the right solution. The very fact that the data was so sparse limited the realm of possibilities. We were meant to XOR each pattern into the next one in the sequence, producing letters. Amazingly, however, three different people in the van all did the XOR wrong and got garbage. I could chalk it up to being tired, but we were sloppy. Once GC confirmed that XOR was the right idea, we got more serious about it and turned the task over to Excel. Excel is your friend. Excel doesn't make stupid mistakes. Always trust Excel.

DNA Runaround: Oh. My. God. Hands down, the worst puzzle in the event. This puzzle had a confluence of problems: my allergies were in high gear, our team was falling asleep, some of the data we had to gather was fuzzy, the puzzle had too many layers, it relied on a bit of data (that there's a DNA "start" codon) that many teams didn't have, the final step was very weak... a perfect storm of suck. A wide array of objects had to be identified according to the element (earth, air, fire, water) they could represent. These elements then mapped to the GTAC nucleotides of DNA, forming DNA codons. Then we had to leap there being a second categorization system for all the objects, this time based on card suits, which ultmately mapped to Braille. This was far too obtuse a puzzle for the time of night at which it was delivered. The final Braille step was extremely thin and unsatisfying, and the best thing I can say about the puzzle was that we got Pop Rocks, Swedish fish, Red Hots, and Airheads to snack on. I'm certain that James enjoyed this puzzle more than anyone else on Briny Deep, because he slept through it in the van.

Charades: NOW we're talking! Or not, since it's Charades. This was a perfect clue for the time of night, and I would have liked it to have lasted even longer. Each team got split in half, with one half "imprisoned" in a glass-enclosed, mostly sound-proof room and the other half left outside the room to "rescue" them. Each side had a seed word which they charaded to the other, and letters from each correctly-guessed word filled in blanks in the next one in the list. Ok, the mechanics of the puzzle aren't really the point-- the important thing is that we played charades back and forth between the two halves of the team until we got a final answer. We hear the Burninators hated this clue, which pretty much summarizes the difference between us and them [update: Wei-Hwa informs me that the Burninators in fact quite liked this clue, so it turns out this in no way illustrates any difference between us. =)]. Sure, there was no real puzzle here, but after the brain-sapping slog of the previous clue this was an ideal change of pace. I would have happily played charades for an hour here, but it only took us a few minutes to finish the chain.

Sodokube: A big metal cube, each side a 4x4 grid. Some cells had magnetic letters permanently affixed (six different colors, one color per side). We also received a supply of letter magnets in corresponding colors, enough to exactly fill the gaps in the cube. Inspection proved that each color had exactly the same set of letters, and that no letter repeated within each color. This was a sudoku cube. Once assembled, a path on the magnets traced out the answer word (MULTIDIMENSIONALITY). The hefty cube is a nifty artifact, and it's always nice to see a different take on a standard puzzle type. Ironically, we had an almost identical puzzle in Mooncurser's but cut it because, at the time, we thought it was too hard. It's also a suboptimal Game puzzle, in that a cube is difficult for multiple people to work with at a time. We wound up transcribing the data into Excel and splitting into two teams, one working with the spreadsheet and one with the cube itself. Kudos for the breakfast, however-- the quiches were lovely, and I particularly enjoyed the cheese-and-basil sandwiches.

Scrabble runaround: We started with a set of Scrabble tiles and a clue. At each stop, we got a new tile and had to anagram the set into an answer to the new clue. That answer became the basis for a very simple paper-puzzle that yielded the combination to a lock. Photos directed us to a nearby location where we had to unlock a mailbox containing the next tile/clue set. Lather, rinse, repeat too many times. This probably seemed like a good idea on paper, but in practice it went on a little too long and was only a 1-2 person puzzle. The ratio of walking around to doing fun stuff (solving puzzles) was too high. We wound up having 2 of us doing the walking and solving while the rest of the team followed in the van. The puzzles weren't meaty enough for multiple solvers to dig in. Tweaked a bit-- meatier puzzles, LONGER distances between mailboxes so that teams drove from point to point and solved during the drive-- this could have been a fun sequence, but instead it fell flat and wore out its welcome.

Sound maze: I slept through this puzzle-- possibly the first time I've ever napped during a clue. I understand it involved an audio tour of a park, and using the sounds to navigate from place to place. Other than that, I know little.

Double words: This was a very clever puzzle. We received two copies of the same puzzle grid, but while the puzzles on each sheet were identical, the sheets themselves weren't-- each had a different-colored note conveying essentially the same information. We weren't sure what to make of that-- it was clearly not just "here's two copies of the puzzle so you can work on it more easily", but beyond that nothing jumped out. The right edge of both sheets faded away, as if the copier was out of toner (and in fact the notes on each sheet mentioned such a toner problem). The puzzle itself consisted of crossword clues and boxes (most empty, but some already filled in) for the answers, with lines connecting some boxes to others in neighboring answers, indicating that the same letter was to be filled in for all such connected boxes (is there a name for this kind of puzzle?). Except it wasn't working. We filled in all the boxes, but some letters didn't carry over as expected and when we tried to extract a final answer we got parts that made sense and parts that were garbage. And then someone noticed that ?????LA?????? could be answered as either CHOCOLATE , as we'd done, or VANILLA , and it would still fit the clue. That's when the penny dropped, and we realized the same was true for EVERY clue-- there were two possible answers that fit. The faded right edge hiid the fact that the two different answers had different lengths. Brilliant! One set of answers went into one grid, the other set into the second, and presto-- a sensible answer. This was a great idea nicely executed, with a terrific "aha" moment.

My Voice is My Passport, Verify Me: To gain entrance to the next clue site, we basically had to play Karaoke Revolution and about a dozen correct notes. Our task was made a little harder by our inability to recognize our song, which (we later discovered) was Puff the Magic Dragon. So we were just trying to match the notes by feel, which is a lot harder than singing a song you know. Still, a few minutes of effort got us in the door.

Dragonquest: A very nicely conceived and executed puzzle in which parts of the Silicon Valley Microsoft offices were recreated in a medieval text adventure, and everything in the physical world had a virtual analog. We needed to open a locked box with colored buttons (a chest with colored gems) by determining the correct sequence to push. This involved talking to a portrait in the game that asked questions about pictures in the real world; finding half a parchment under the real toilet and the other half in the virtual outhouse; and so forth. While the rest of my team wandered the physical site, I explored the virtual one and called in when I found something of interest. That parallelization worked nicely. I wonder why they chose a horrid Scott Adams-style parser (GO N instead of just N) instead of a friendly Infocom-style parser. The parser cost one team at least an hour when their attempts to TOUCH RED met with "Don't touch that", where TOUCH RED GEM was required. Ugh. Happily, that didn't bite us. This was a fresh, innovative puzzle that worked well and I really enjoyed.

A Puzzlemaker's Dilemma

In the past ten years or so, I've participated in a LOT of puzzle events: 9 MS Puzzle Hunts, 5 Puzzle Days, 3 Puzzle Safaris, 1 Iron Puzzler, 6 full-length Games, 4 half-day Games, 1 SNAP and a couple other local walking hunts, at least 7 other treasure hunts at various places, and probably some things I'm forgetting. That body of experience has given me a certain adeptness at recognizing how a puzzle will deliver its final answer. Once upon a time, realizing that a message was in Braille or Semaphore was a deeply rewarding insight. Now it's second-nature. Any matrix with a 2x3 aspect ratio suggests Braille now, and any angular configuration screams Semaphore. Likewise with all the other standard encoding systems-- Morse, binary, ASCII, etc. It's old hat, and recognizing these schemes is second nature to most players of similar experience unless they're craftily obfuscated. Shinteki: Decathlon II had a particularly fresh encoding of 5-bit binary, mapping colors in national flags to the corresponding Olympic rings, and making that connection was satisfying. But that's the exception more than the rule these days.

This represents a big dilemma for puzzle event designers. The Game and Puzzle Hunt communities both have a core group of dedicated, experienced teams and a large field of less experienced ones. How do you create a satisfying event for both groups? How do you use the familiar language of puzzle encodings in a fresh enough way that it's challenging and satisfying for veterans while still approachable to newbies?

Innovation in puzzle design is one answer. But creating entirely new paradigms is hard, and a lot to ask of people who put together an event in their spare time. Perhaps the answer is to change what we think of as a puzzle. More physical manipulation puzzles, interactive locations, computerized challenges. Puzzles where you're given the answer in exchange for accomplishing a task. Maybe we need to shift out thinking away from the traditional paper-and-pencil puzzle mold.

Maybe there is no answer. Perhaps it's a natural life cycle and eventually veterans become so well-versed in the language of event puzzles that their interest in the entire genre dies, extinguished in a kind of puzzler's Carousel ("Renew! Renew!").

I'll be reflecting on these and other mysteries, particularly those of the ancient Maya, for the next two weeks on the sunny Yucatan peninsula.

Puzzle Hunt A

This weekend was the tenth Microsoft Puzzle Hunt (Puzzle Hunt A. A is 10 in hexadecimal. Oh, those crazy computer geeks!), the theme this time around being Atlantis (although the hunt was essentially unthemed, as aside from a beginning-, mid-, and closing-game skit, there was no Atlantean flavor anywhere in the body of the hunt). The puzzles were generally good, including some very nice metas and a nifty structure which, when assembled properly, caused a message to appear when quinoa was poured into it. The organizers did a fine job and produced a smooth, error-free event. Our team finished second, six minutes behind the winning team, with the third-place team a scant few minutes behind us.

Fifteen hours after the hunt began.

In fact, as I type this the hunt is still going on. I've had time to get a good night's sleep and eat a leisurely breakfast, and I'll be heading back to MS for the closing ceremonies in a few hours.

But I had a hard time getting to sleep last night, because I was a little depressed. Not because we finished second-- it was a close, good finish. In fact, throughout the hunt the top four teams kept flip-flopping positions, which is always more exciting than when a team pulls out to an unassailable lead. I was depressed because it was over so quickly.

The MS Puzzle Hunt used to be designed so that the winning team would finish in the late afternoon on Sunday. In truth, the first few hunts took things to the wire. Then things started to change. We reached the final meta of hunt five 8 hours before the event ended (and then banged our head against its inscrutable ambiguities for the next 7 hours and forty-five minutes). But nobody else was even close to us. When we ran hunt 6, the winning team was also far ahead of their closest competition, restoring the timestream early Sunday morning, but a few other teams also finished before the hunt was over. Hunt 7 had issues, and the winning team had to be pushed to the finish even as all the other teams were coming in for the wrap-up.

With PH8, hunt design took a turn for the easier. We won PH8 at 4:30 AM on Sunday morning (with other teams coming in throughout the day). The disappointment of finishing the hunt so soon was mitigated by the euphoria of, well, finishing the hunt so soon-- and by the superb quality of the hunt. We won PH9 around 11 AM, and that seemed about right to me. Other teams also finished before the 6 PM deadline, and we still got a full 24 hours of puzzle-solving in.

Fifteen hours, however, is just too short. I'm sure a lot of other teams will get to finish this hunt, which will be great for them. But it feels to me like we've reached the end of an era, and it saddens me.

I recognize the problem. The disparity between the top teams and the next tier continues to grow, which presents a genuine dilemma for hunt organizers. If they create a hunt that will keep top teams occupied for 24 hours or more, the rest of the pack will not get to experience the full event. It will continue to overwhelm the lower teams. Organizers don't want to spend time creating and testing puzzles that a small number of people will get to see. And the needs of the many outweigh the needs of the few ("...or the one"). It's hard to argue with the notion that the right place to target is closer to the center (or right of center) of the bell curve. And so future hunts will likely continue to be shorter events for the top teams.

Top teams can shrug and say, "That's the way it is," adjusting their expectations. If I'd known it was to be a fifteen-hour event ahead of time, I'd have come into it with a different mind-set and reacted more favorably to its conclusion. Instead, I spent the past 15 months anticipating this event, and while it was qualitatively good, in the end if wound up being quantitatively less than I'd hoped for. To be fair, the organizers didn't expect it to be so short. Their playtests, with experienced solvers, bigger teams, and no loss of time for running around campus, still came up against the 6 PM deadline-- so even after streamlining further, we shattered their expectations. The timing of something like this is very hard to get right. The old-school philosophy was that if you make it hard, you can always make it easier on-the-fly by providing hints, but if you make it too easy there's nothing you do about it. That philosophy seems to have changed.

Perhaps the problem is that the top teams are just too good. Maybe we should just break ourselves up and form new teams with less experienced solvers. If the goal is to spend more time puzzling, that would probably do the trick. With fewer top players on the same team, each of them would also get to see and participate in more of the event's puzzles. But of course, people like to play with their friends, and they also like to play with people of comparable skill. If you're the superstar of your team, your experience of the hunt will change. You might spend most of your time helping other people instead of solving things yourself. You might get frustrated by slower progress. You might offend teammates by swooping in to help. Most of the people I'd want to play happen to be good puzzle solvers, and I suspect the same is true for many other top players. Breaking up the lead teams would just shuffle the top players around without leveling the playing field.

Am I in it to win it, or am I there to have fun? As long as the hunt remains a competition, I want to be on a competitive team (that's "competitive" in the sense of being of the same caliber as other top teams, not in the sense of being win-at-all-costs). The adrenaline rush of trying to outsolve other teams is one thing separating the hunt from a weekend with P and A Magazine. But I always want to be satisfied. I want to be materially involved in solving as many puzzles as possible. I want to collaborate with teammates and share the thrill of a great insight. I want to work through the night and rally the team at 4 AM to come together over a tough puzzle and push through the invisible wall. As the top teams become more experienced and adept, I want the event itself to grow with us and challenge us further.

But I'm not sure how that can happen. There are people like me who solve a Monday NY Times crossword in under 5 minutes (hell, there are people in the world who solve them in under two), and there are people who finish them in an hour. How do you accommodate both groups in the same event? How can you possibly slow a 5-minute solver down without being unjustly inscrutable, or speed a one-hour solver up without feeding him answers and taking away his fun? You could give the 5-minute solver Thursday-level clues, but now you're almost making two different puzzles, which of course takes much longer. Beginners could be given full instructions for every puzzle (which often come without any at all), or more hints.

The truth is, little is likely to change. This is only an issue for a fairly small percentage of players. The needs of the many. The real goal is to find simple, easily-made changes that create a better experience for the top teams while keeping the event accessible to the rest of the pack.

Iron Puzzler

I've been extremely busy lately-- too busy even to blog. But not, apparently, too busy to devote an entire weekend to a new Seattle puzzle event, Iron Puzzler. Inspired by Iron Chef and a similar event in the Bay Area, in Iron Puzzler all the puzzles are created by the competing teams themselves. Four secret ingredients were announced at 9 AM on Saturday: CLOCK, L, MERCURY, and SPOON. Each team had 15 hours to create one paper puzzle and one non-paper puzzle, each using at least one of the theme ingredients. Sadly, neither Alton Brown nor Will Shortz was on-hand for color commentary ("I see a bunch of ones and zeroes on the challenger's side, I believe he's going to turn that into Morse, a fairly traditional but versatile preparation.").

Fourteen teams participated, meaning each team had to turn in 30 copies of the paper puzzle (2 per team, including the organizers) and 15 copies of the non-paper one. Then, at midnight (as it turned out, closer to 1 AM), the puzzles were distributed and the 15-hour solving period began. Teams scored points for each puzzle solved (with a small bonus for being one of the first three teams to solve each puzzle), and each puzzle earned its creators points according to how many teams solved it (with the sweet spot at 10 teams). At the end of the event teams also awarded each puzzle points on quality/fun, presentation, and use of the ingredients.

Creating a puzzle by committee under time pressure proved to be a challenge. Our group largely focused on the physical puzzle first, because we knew we'd have no trouble putting together a paper puzzle. Meanwhile one member of the team went off on her own and produced a terrific paper puzzle independently. In the end, only two teams solved our physical puzzle (which was a solid puzzle that we overstreamlined, removing a couple of hinting elements that we should have left in place) while everyone solved the paper puzzle and voted it their favorite. A rather disheartening result, actually, since it undermines what for me is the most interesting aspect of the event-- creating puzzles collaboratively. These results suggest that we'd be better off splitting into individuals or pairs and developing puzzles independently, then coming together to test them and pick the two strongest. That's a less interesting challenge than producing puzzles as a group. On the other hand, we could also interpret the results as an indication that next time we should value elegance less than solvability and internal hinting.

As expected, there were a lot of puzzles that involved periodic tables and clock faces, but surprisingly little semaphore. I think most teams thought as we did: "Oh my God, CLOCK-- everyone's going to do semaphore, so let's do something else." Interestingly, that kind of reasoning resulted in two spoonerism puzzles presented as bags of plastic spoons with parts of spoonerisms written on each spoon. We also got no less than four crosswords (one involving spoonerisms!), but sadly none of them cryptic.

The overall quality of the puzzles was surprisingly high. One of the things I found most interesting about the event was the different interpretations of the ingredients. Quite a few puzzles incorporated the periodic table or other chemical elements, for example, but none focused on the planets or car manufacturers. Analog clocks were prevalent, but nobody made a digital clock puzzle. And regrettably, nobody did a superhero battlecry puzzle ("Spoooooon!"). As a constructor, it also seemed we could leverage the fact that everyone would view every puzzle through the CLOCK/L/MERCURY/SPOON filter, enabling them to make leaps that might otherwise seem unfair.

The event was put together in about a month. It was an undeniable success, but the rough edges show. The scoring system in particular needs some massaging. The number of teams solving a given puzzle does not seem like a good basis for awarding points. The premise that a puzzle with 10 of 13 teams solving it is more desirable than one solved by all teams is specious. I'd argue that a puzzle that everyone solves is more desirable than one only 75% of teams solve. The catch is that you want puzzles to be a challenge-- you don't want everyone to solve it immediately. Nobody wants to see teams submit a 6 letter anagram and call it done. But I don't think we need an entire scoring vector to capture that. If a puzzle is too easy or too hard, teams can reflect that in their ratings. If a puzzle is just right in difficulty but not fun to solve, teams can dock it points.

This event was created to fill a gap formed by the delay of the next MS Puzzle Hunt, but it was intriguing enough in its own right to warrant repeating. Since it requires far less effort to organize than traditional events, the chance of that happening seems high. In fact, if someone creates a back-end that allows teams to register their puzzles and answers and then allows teams to submit answers to solved puzzles electronically, organizers could play along, too. Allez cuisine!

Hogwarts and the Draconian Prophecy

Last weekend I escaped the Muggles to participate in Hogwarts and the Draconian Prophecy. This Game was run by Snout, the same team that ran the Justice Unlimited Game a couple of years ago. With such a richly themed subject and an experienced GC at the helm, we had high expectations for this event and were not disappointed. Snout hit many balls out of the park.

Their biggest success was with theming. The world of Harry Potter is rich with detail and flavor, and Snout leveraged that to create an intensely thematic Game experience. We were Sorted into Houses, rode the Hogwarts Express, attended classes, used wands (by far the coolest thing about the Game, and described in detail below), snuck into Hogsmeade, collaborated with other teams in our House, mixed potions, withdrew funds from Gringott's, and evaded Dementors. Surprisingly, however, there was no quidditch. Some kind of faux sporting event involving all sixteen teams would have been a blast, and this seems like a real missed opportunity.

The entire staff remained in character throughout the Game-- headmaster Curtis never once slipped from his faux British accent. In their last Game, I felt like Snout dropped the putative premise of the Game-- that we'd been recruited as substitute superheroes in the wake of the disappearance of the more established ones-- almost immediately and instead just presented us with puzzles that incorporated references to established Marvel and DC superheroes. With Hogwarts, they let the theme inform the Game design and incorporated it thoughout the event. And they did so without once using any of the established Hogwarts characters but Hagrid (if Curtis' headmaster was Dumbledore, I don't recall ever hearing the name used).

The second area in which Hogwarts excelled was in story. The Game wasn't just themed as Hogwarts, but followed the basic outline of a Harry Potter novel. Imagine our surprise to discover that the Defense Against the Dark Arts instructor had gone missing even before the Game began, or that the substitute turned out to be evil! The flow of the Game followed the story very well, and while not every puzzle made strict "in-story" sense, a remarkable number of them did. The plot actually advanced and got resolved in a dramatic and entertaining way.

By dividing teams into Houses and awarding House points, Hogwarts was the first Game I've played in which encouraged collaboration among teams (Our Mooncurser's Handbook Game encouraged interaction among teams, but not cooperation to this extent). We've competed with The Burninators in past Games, but at Hogwarts they were in our House and we wound up collaborating with them on many puzzles. Our playing styles are very different, so it's unlikely this would ever have happened under other circumstances. We enjoyed getting to know them better. I particularly liked the one puzzle where all four teams in each House needed to work together to get the solution, and wish there had been more such collaborative moments in the event.

The clues in Hogwarts were perhaps the least successful element of the Game. Most were perfectly fine, but none were especially innovative, surprising, or delightful (although the wand device was amazing). We often referred to a textbook given to us early in the Game, and in fact this crutch was overused. Having only two copies of the book also limited the number of teammates who could actively participate at times. On the other hand, very few clues came on paper-- Snout did a great job of delivering interesting materials to manipulate.

Snout didn't get everything right (more on that below), but the magnitude of their successes far outweighs the things that could have been done better. This was a stellar Game and a terrific experience. All involved should feel exceptionally proud.

Now for the play by play.

Sorting Hat: The Game began at platform nine and three quarters of the Emeryville train station, where we checked in and walked a short way to a movie theater. The sorting hat ceremony was supposed to happen inside the theater, but a snafu with the management forced the staff to go to plan B and we had the ceremony outside instead. A computer-animated sorting hat sang a unique sorting song for each of the sixteen teams, placing us in the four Hogwarts Houses (my Briny Deep teammates and I chanted "Anything but Hufflepuff! Anything but Hufflepuff!" as we appeared before the sorting hat, and to our relief were placed in Slytherin). There was much cheering as each team was sorted and given red, yellow, blue, or green bandanas as House identification to wear during the Game (all of the bandanas also had a simple semaphore puzzle giving us our House password). We were then given our breakfast, which is the first place Snout went awry. The Game began at 7:30 AM, which means we were up by 6:30 and had a long day ahead of us. We'd been promised breakfast, however, so we didn't grab anything before we left. The breakfast they provided-- a bottle of juice, a shrinkwrapped danish, and a string cheese-- was woefully inadequate. When we provided teams with a bagged breakfast in Mooncurser's, we gave them fresh bagels and cream cheese, milk, cereal, juice, and fruit, in ample quantity. So when Snout promised a bagged breakfast, we expected we'd be well-fed. I couldn't believe my eyes when I saw how meager their bagged breakfast was, and my heart sank a little.

Hogwarts Express: Escorted by prefects dressed in robes of their House colors, we walked back to the train station to wait for the Hogwarts Express. Speculation abounded as to where we were going. Some people thought Tahoe was likely, while our team felt UC Davis-- a school campus that could act as Hogwarts, especially since classes were not yet in session-- was a top candidate. The mystery of where we were headed was unfortunately spoiled when an Amtrak employee got on the loudspeaker to ask, "Will the leader of the Hogwarts group to Sacramento please come to the ticket counter?" The waiting players let out a collective groan at the unintended spoiler. This is the second place I thought Snout miscalculated. The Game began at 7:30, and there was quite a bit of waiting around before the sorting ceremony began. The train didn't leave until 9:15, and there was more waiting before that. Almost two hours into the Game, we'd done a lot of waiting and hadn't yet received a single puzzle. Some delay at the start is inevitable as teams arrive and check in, and obviously Snout needed to make sure they had enough time to complete the sorting ceremony before the train departed. But at the start of a Game you're primed and ready to go. Some kind of puzzle to work on before boarding the train would have been welcome. Perhaps the bandanas were that puzzle, but they were cracked so quickly it's hard to think of them that way.

Once the train got underway an owl traveled the aisle to deliver that day's Daily Prophet which brought us up to speed on the story so far. Professor Cross, the Defense Against the Dark Arts teacher, was missing and wanted for questioning by the Ministry of Magic. Wizards and witches were losing their magic. Mugglium, an element hazardous to wizarding folk and thought to be a myth, had been discovered. And oh, look-- a crossword! Solving it told us our class schedule would come from a flavorful source. A little while later, we received a huge package of Bertie Botts' Every Flavor Beans. Teams had been divided into third, fourth, and fifth-year students, and each year received a different package. We needed to sort the flavors (which included such treats as Sardine, Upchuck, and Garlic) and count them to create an order, then read the first letters of the flavors in that order. In the case of the fifth year students, however, the flavors were already listed on the package in the correct order, so no puzzle solving was actually necessary. So far as we know none of the fifth-year teams noticed, however, until well into their sorting process, so this didn't wind up spoiling the puzzle (except that we didn't "solve" the puzzle so much as we "noticed" the answer was already there for us).

Upon arrival in Sacramento we followed a picture trail to lead us to the location of our first class.

Care of Magical Creatures: Hagrid (who looked and sounded the part quite well) welcomed us to class and gave us our textbooks-- custom-made books with information about creatures, spells, potions, wand use, encoding schemes, wizardly lore, and the like. Over the course of the Game we'd refer back to the book frequently and repeatedly. After quizzing us on the properties of various creatures, which we'd answer by searching through the book's bestiary, we moved on to the class assignment. Hagrid played a recording of weird sounds, and we needed to identify which creatures made them based on the descriptions given in the bestiary. The bestiary listed 26 creatures, each beginning with a different letter of the alphabet, so identifying the creatures gave us the location of our next class. Matching the sounds to creatures was an inexact process. Which one might sound like a ratchet on steroids? Which one like an underwater gurgle? When finished, we received a CD with a complete study guide of all 26 animals and their sounds, which we expected would come in useful later. Andrew spent the remaining class time listening to the guide and learning what everything sounded like.

Defense Against the Dark Arts: Our substitute, Professor Guzzany, taught us the fundamental of wandcraft. Each team was issued a wand, which was an incredibly cool device building on the basic concepts of the DRUID and Bat Blinker from Justice. The wand had five red LEDs at the tip, and was turned on by tapping it against something. Built into the wand was an accelerometer which enabled the wand to detect and recognize gestures. Our textbook had a lexicon of phonemes and the corresponding "mwanemes"-- looping or flicking motions with the wand. Stringing together the proper sequence of mwanemes "cast" a spell and generated a response from the wand which could be seen by waving the wand back and forth to create a persistence of vision optical illusion of scrolling text. Answers during the Game often took the form of spells we needed to cast with the wand. Upon casting the correct spell, the wand would then supply us with instructions on where to go or what to do. This device alone brought the theming rating to 11. Some teams had difficulty mastering the wand and found it very frustrating, but Andrew mastered our wand quickly and we loved it. Sure, it took longer than typing an answer into a Palm and sometimes required multiple attempts to get correct, but the coolness factor and thematic rightness trumped any such quibbles. The varying degrees of wand mastery among teams created additional opportunities for collaboration, as teams with lesser ability at wanding would assist others with puzzles in exchange for the information yielded by casting the result successfully.

The wands also had built-in clocks with preprogrammed skip schedules for clues. Solve a clue beyond a certain time, and the wand would automatically and transparently skip you over the next clue to maintain the schedule. Right out of the gate the schedule got a little tight-- I think the train was delayed, which cut into the class timetable-- and our Defense class was curtailed to keep the schedule. Unfortunately, lunch was served very late during this class, which meant we were rushed and didn't have much time to eat. We were famished in the wake of the light breakfast, so this was less than ideal but not really under Snout's control.

Potions: Our last class was Potions. Our professor was not Snape, unfortunately (we Slytherins would have had a House advantage), but he was entertaining and in-character throughout. A kit of eight colored liquids, 8 plastic pipettes, and four receptacles was set before us along with a logic puzzle. Solving the puzzle told us what liquids to mix together. The resulting four potions-- one of which foamed and expanded dramatically-- each had a different color matching some of the colors around the border of the puzzle. Ignoring the colors not represented by the potions, the potion colors appeared either alone or immediately next to one other-- dots and dashes of Morse code telling us what spell to cast for our next instructions. Once again, class was a little too rushed here and we'd have enjoyed having more time.

House Portrait: Each House was faced with their own House portrait and a box with a combination lock. The frame of the portrait was decorated with colored stones, and the figure in the portrait wore a necklace featuring chains of colored stones. Each chain represented a letter in the password we got from the bandana. The same stone patterns could be found on the frame, allowing the frame to be decoded into a message: CAST ARCSTONES. Casting ARCSTONES with the wand told us to cast the name of our House, which then gave us the combination to the lock. My big problem with this puzzle was that some of the stones on the frame weren't used-- they were just noise. Since the authors had complete freedom to compose the puzzle as they wished, there was no reason why all the stones couldn't have been used. This would have made it much cleaner. Instead, we kept thinking we were doing something wrong because we had unused stones. The puzzle itself was therefore unsatisfying, but the process-- in which all four teams of each House worked together-- was terrific. I was in the center of the Slytherin clump so I was directly involved and really enjoyed it, but I heard from teammates on the fringe of the clump that they were too far away from the action to participate, which is unfortunate. Perhaps if the portraits had been poster-sized instead of 8.5 x 11, more people could have felt involved. But the underlying idea of getting all four teams of a House to work together was outstanding, different, and really fun.

Dorms: The locked box contained the keys to our dorm rooms, which were parked in a nearby lot. Aha, vans! Upon reaching our van we discovered three things: 1) a cooler full of beverages and munchies, plus a first-aid kit, 2) a package labeled "I solemnly swear that I am up to no good" containing a bunch of maps, including a Marauder's Map of Hogwarts and Hogsmeade, and 3) a CD with a message from the Ministry of Magic telling us to sneak into Hogsmeade for an important meeting, but to avoid getting caught by any Hogwarts staff who would take away House points for being off the grounds without authorization.

A word about the vans: ours was a Kia, with barely enough room for 6 people. When we play in full Games, we always rent a larger, 12- or 15-person van. We remove a bench to create a workspace in the van, and the extra room allows people to sleep in the back. These vans afforded none of that, so things were quite cramped. Since the van forced us into three rows of two, it made it difficult for everyone to work on some of the clues. This was exacerbated by Snout's choice to provide only two copies of each paper-based clue. Given that we were in three rows, three copies would have been far preferable. To be fair, there were very few paper-based clues in the Game-- which makes the question of why they didn't supply us with more copies of those few all the more puzzling.

Hogsmeade: Another exceptionally cool part of the Game. Sneaking from our vans to the clandestine meeting was a ton of fun. We made life harder on ourselves than it needed to be by not realizing that our Hogsmeade map was a Marauder's Map with footprints representing the locations and routes of the Hogwarts staff. Instead, we just flew by the seat of our pants, which may have been even more fun. Thematically, the choice of location was spot-on. Our vans were parked at a Sacramento mall. Just outside the mall was a tunnel that led to Old Sacramento, a touristy area of old-time frontier-looking shops. This jived perfectly with the books, in which a secret tunnel leads from Hogwarts to the retail town of Hogsmeade. I think the sheriffs or park rangers or whoever they were in their brown uniforms and flat-rimmed trooper hats didn't quite know what to make of all the oddly-dressed people playing a big game of hide and seek in their town. We hid from them almost as much as we hid from the Hogwarts staff! When we finally reached our rendezvous, we were told that Professor Cross was researching the Draconus Device-- a magic-sucking artifact legend said was disassembled and hidden by the four Hogwarts founders-- when she went missing. We needed to find her.

Floo Network: We got on Cross' trail by tracking her movements through the floo network. We received a diagram of the network, along with cryptic traces of her passage that turned out to be ultra-compressed word associations. This was a nice puzzle, well-presented on scorched parchment (but a third copy would have made it easier for all of us to be involved).

Broomsweeper: At our next location-- the Nimbus Fish Hatchery-- we picked up a broom and our next puzzle, a game of minesweeper with brooms instead of mines and many values already filled in. Once we found all the brooms they formed braille letters telling us to correct the mistakes. Many of the given values were blatant errors that appeared in clumps. Connecting the members of these clumps in numerical order created letters giving us our answer. This was a straightforward puzzle we dispatched quickly.

Magic Mirror: A mirror with strange symbols around its perimeter, along with a message telling us that the spell should be drummed, not waved. The symbols were from a code given in the textbook, but each symbol was a mirror image and the alphabet was listed backwards. Another easy puzzle, but we were impressed that not only could the wand detect waving gestures, but it also had rhythm and could detect different drumming patterns-- in this case, the first few bars of Puff the Magic Dragon.

Traveling Tunes: An audio CD with 18 tracks, paired up on the jewel case with a roman numeral between track numbers. Also included was a map of the London Underground system, with each stop replaced by the name of a musical artist. We identified the songs and artists fairly quickly, but had trouble figuring out what to do with them. It was a complete brain fart. My call to GC went something like this:

Me: "We've identified all the artists and songs. We've tried treating the routes between pairs as semaphore, we've counted the number of stations between the pairs and used that as an index, we've noticed that all the pairs have an odd number of stations between them, but we're not getting anything sensible."
GC: "Ok, so you've counted the number of stations between the pairs--"
Me: "--and we need to use the roman numeral as an index into the name of the stop at the middle of the route connecting the pairs, don't we. Nevermind!"

We arrived at that site 30 minutes ahead of the next team, and left about 15 minutes after they did. Not our best moment.

Gringott's: In the back of an organic cafe we found, of all things, an old mine tunnel! We hunched over and crept across the planked floor by the light of bare overhead bulbs to find a Gringott's teller at the far end. He allowed us to make a withdrawal from our account, handing us a sack of gold Gringott's galleons (foil-wrapped chocolate). Each galleon had a unique date, all divisible by 3, and the wrapper bore the quote "What hath God wrought?" The quote was the first message transmitted by telegraph, cluing Morse code. Express each date in base three, treat zeroes as spaces, ones as dots, and twos as dashes, and presto-- each coin is a letter in Morse. This was also the dinner stop, but instead of providing us with dinner they gave us $15 each to go buy dinner ourselves nearby. We chose the Mexican restaurant next door, which had a terrific apple jalapeno carnitas special and friendly waitstaff.

Rygbi Klew's Potions: Five bottles of juice-- er, I mean potions-- with various patterns of colored stripes on them. Arranged in the right order, the stripes could be read as 5-bit binary by reading only red stripes as ones across the top of all 5 bottles, then only the yellow stripes across the second row, then green across the next, then blue, then indigo, then back to red. We didn't much care for this puzzle, because the decoding scheme seemed entirely arbitrary. This puzzle seemed like it was intended to leverage what we'd learned in potions class, but it didn't. In potions class we'd been trained to ignore the colors that weren't represented in our potions. All of the juices were red, so ignoring the non-red colors would have made more sense.

Honeyduke's Bars: Three big chocolate bars stacked atop each other, many with red and/or blue sugar candies affixed to some of the squares of the bar. This was obviously Braille, with each chocolate bar representing the top, middle, or bottom row of Braille, but we couldn't get it to work without GC's help. This was for two reasons. First, we kept on trying to solve using an encoding system that seemed more natural and elegant than GC's: treat each square as a letter; red dots in the left column, blue in the right (the Honeydukes logo on the wrapper showed HONEY in red and DUKES in blue, hence red=left, blue=right). The real encoding was that the red dots were one data stream reading forwards, the blue was another reading backwards, and each letter required two chocolate squares from each bar instead of just one. Our second problem was that we oriented the bars exactly as they were given in the wrapper, using the Cadbury logos on the chocolate as a double-check to make sure they were all aligned, but our middle bar was incorrectly constructed in reverse. So we kept getting garbage no matter what we tried. Very, very frustrating.

The Draconian Prophecy: On the banks of a river we finally tracked down the missing Professor Cross, who had been driven insane by exposure to Mugglium. The person playing her part really threw herself into the role and was absolutely fantastic, never once dropping out of character. She told us the Draconus Device was divided into 16 pieces, so all 16 study groups would need to work together. She gave us the spell we'd need to cast in unison (DYNAMITE) once all pieces were obtained, and handed us a complete copy of the Draconian Prophecy. The prophecy contained veiled references to a series of spells, the first letters of which gave us our answer. The puzzle itself was easy, but the combination of the strong performance and the elegantly printed and origami'd prophecy made this a very fun, thematic location.

Animal Sounds: Elsewhere along the river we found a matrix of small spheres strung up in the air. A series of animal sounds could be heard, and with each sound a different set of spheres lit up. Andrew's study of the CD from our Care of Magical Creatures class paid off here, enabling him to quickly identify the animals. When the animals were grouped by threat level and the lights for each threat level were displayed together, they formed the letters of our answer spell. We figured out pretty quickly that the threat values were important here, but we wanted to treat the spheres as inputs into the squares to which they were adjacent in the matrix, using the threat values of the animals as input values and summing them. That, of course, never worked (yet again, in the process of solving a puzzle we invented a completely different one), and it took a nudge for us to try something simpler.

Gryffindor: Professor Cross informed us that each of the four founders left challenges behind to protect the fragments of the Draconus Device, and we'd need to overcome each of them. FIrst up was Gryffindor, where we received a talisman that blinked in three colors. We very quickly discovered that squeezing the talisman turned it off, and squeezing it again not only turned it back on but changed the sequence of lights. We cracked this one virtually immediately, reading each triple-blink as a base 3 value with red as 0, green as 1, and blue as 2. Each squeeze sequence was one word. By the time two of us had transcribed one sequence, another pair of us had decoded the previous one.

Ravenclaw: Time for tea! Curtis served us six cups of tea from a coffee truck. Each cup had "tea leaves" marked on the bottom. We needed to identify the shapes of the tea leaves, then find the associated meaning of those shapes in the divination section of the textbook. The section on reading tea leaves mentioned that a symbol's rotation affected its strength, so arranging the symbols in order of their rotation and reading the first letters of their meanings provided the answer. I thought this was a nice puzzle, but I was a little tired of the whole "look it up in the textbook" mechanic, which by this point had been overused.

Slytherin: Two Slytherin guardians blocked access to all but those demonstrating Slytherin qualities. So I went with a classic: "Look, over there!" The guardians obediently looked in that direction, giving us a chance to sneak behind their turned backs and swipe our next challenge (and a woefully inadequate breakfast of Trader Joe's breakfast bars, further cementing our belief that Team Snout are in fact birds). The challenge was essentially a tangram puzzle, once again referring back to the textbook for the shapes. Two parts of this puzzle's design really bothered me. First, each tangram had an eight-letter nonsense word associated with it. When I see over a dozen words all the same length, that jumps out as being important. In this case, I expected to read down a column of their letters. Nope-- their names were completely unused, except for an acrostic from their initial letters. So why make all the nonsense words equal length? Second, each tangram wound up getting associated with a number between 10 and 260. With all those juicy names of equal length, made of unusual letters, those numbers were just begging to be indices or a way to order all of the tangrams. Nope. We had to divine, unhinted, that we should drop the trailing 0 so that we had numbers in the 1-26 range, which we could then treat as letters. Perhaps someone will tell me that the numbers were in fact an ordering scheme, and that reading down the proper column of the names produced a clue to the 1-26 approach-- in which case, my objections are withdrawn. Barring that... blech!

Ceramic Tiles: We expected a Hufflepuff challenge, but this was completely unrelated-- perhaps it was a slow-down puzzle for teams ahead of the curve at this stage. We received a 6x6 sheet of ceramic tiles with pigpen cipher symbols on the back. Read in the proper orientation, the symbols told us to reconstruct an image with the front of the tiles. We promptly separated the tiles, flipped them over, and assembled the image (a Hogwarts "H"). In fact, it took us more time to squabble about how to flip the tiles back over to read the rearranged pigpen for our next destination than it did to assemble the thing.

Hufflepuff: The final founder's challenge gave us a bag of ball bearings and colored magnetic rods. Each rod bore a symbol at each end and a positive or negative value at its center. A slip of paper gave us four rows of symbols divided into sets of 3-5. Solving was a simple matter of assembling a polyhedron by connecting the symbols in each set, then adding the values on all rods of the same color. These produced values in the 1-26 range, giving us the spell to cast. This was a fun little puzzle to work on, but unfortunately it wasn't a great group puzzle-- only a couple of us could really work on it at a time. The rest schmoozed with other Slytherins at the site and collaborated on figuring out the right deciphering steps.

Dementor Tablet: We had to sneak by a pair of roaming dementors to recover the next clue, a tablet with 16 lines of Roman numerals and cuneiform numbers. We noticed that there were only 4 different Roman values-- 54, 60, 70, and 75, but we had no idea why. The directions that brought us here said we might need first aid, and our first aid kit included 4 squares of chocolate (protection against dementors), but because we didn't get tagged by the dementors we didn't think it mattered. A quick call to GC suggested otherwise, so we took another look at the chocolate and discovered that each square had a different level of cocoa content-- 54%, 60%, 70%, and 75%. Aha! The cuneiform values indexed into the names of the four chocolate squares, and presto! We were off.

The funny story from this location involves the Burninators, who were at the site with us and were equally stumped. When we told them GC said our first aid kits were needed, they gave us blank looks. The conversation then went something like this: "You know, the first aid kits in the cooler." "Cooler?" "The cooler full of snacks in the back of the van." "We never looked in there!" Turns out each of them had thought the cooler belonged to someone else, so nobody ever opened it. They promptly ran to the cooler, grabbed the first aid kit, and attacked the stockpile of snacks they hadn't known had existed, shoveling down in 5 minutes all the food that had been intended to tide them over for the past 20 hours.

Ziggurat: The final challenge combined a bunch of terms culled from many of the puzzles we'd solved during the Game, and required us to search through the textbook and other materials to resolve each term to a single letter. This was a fun, quick team effort that brought everything together nicely. Upon getting our final answer we received a piece of the Draconus Device.

Finale: All teams gathered for a banquet at a riverboat hotel, where we turned over our fragment of the artifact to the headmaster who in turn placed it at the head table. Steaming trays of scrambled eggs, potatoes, sausages, muffins, and fruit arrived on the buffet and a horde of ravenous players finally got to fill our bellies. Once all teams arrived, Professor Guzzany began the ceremony to permanently destroy the Draconus Device. But when he finally cast an incantation, the assembled throng was shocked-- shocked, I tell you!-- to hear him say "Reparo!" The device came together into one piece and the assembled wizards and witches collapsed unconscious as Guzzany began to gloat. The sixteen wand-bearers among us stepped forward and began casting DYNAMITE, but many fizzled. A confident Guzzany taunted and goaded us as we made a second attempt which also failed. Finally, in our third and more orchestrated attempt, the spell succeeded. The Draconus Device was destroyed, and the power-hungry Defense Against the Dark Arts professor was taken into custody.

Shinteki: Decathlon II

The pirates of Briny Deep made port in the east Bay area of California last weekend for the latest 12-hour puzzle event from the good people of Shinteki: Decathlon II. And the operative word for the weekend was HOT. The west coast is in the throes of a heat wave (even here in Seattle, where it's a balmy 90 degrees by day and a scant 82 at midnight). The heat dragged all teams down, making it difficult to find a comfortable place to sit down and solve. While everyone would have preferred a few clouds, the overbearing heat didn't deter players from enjoying the well-organized event.

Briny Deep fell victim to a classic (for us) blunder, letting a single bad data point deter us from pursuing the correct path. The 60-90 minutes we lost on that puzzle wound up being costly, as we ran out of time and failed to reach the final puzzle or solve the penultimate one. That sent us from a possible first place finish to third.

But enough about us. You want to hear about the puzzles.

Sprint: A four-part relay on the athletic field/track of Merritt College. One team member had to work with players from three other teams to solve an oversize jigsaw puzzle, then run a lap. Yes, even in the absurd pre-noon heat. The next team member had to solve a Rubik's cube (we'd been prewarned about this and provided with solving algorithms), then run a lap. The third team member (me) had to eat twelve saltines without any water, then run a lap. The final team member had to research six facts in a provided World Almanac, but no lap was required. Aside from the cube, which some teams had great trouble with (Andrew solved ours in about 4 minutes), none of the tasks was very hard and it was fun to heckle from the stands and cheer on our teammates. I could have done without running the laps, however-- really, the last thing I want to do before spending the rest of the day in a van with three other people is get all of us sweaty. Good times. To their credit, the Shinteki crew spent less time than normal on an introductory spiel (and the heat silenced any hecklers in the crowd, who probably sensed that their rapier wit played better to an audience that wasn't medium rare) and got us going as soon as possible.

Classic: An acrostic. I do loves me the word puzzles. The twist to this one was that the grid consisted of the names of gold medalist decathletes (for which the alamanc from Sprint came in handy). Making that connection was a nice little aha moment, and since three copies of the puzzle were provided, it was easy for everyone on the team to contribute.

Orientation: The delivery for this clue was a little odd. We received it at the Berkeley Rose Garden, but the text of the clue included a conceptual street map of nearby Emeryville. With all of the intriguing placards of rose species throughout the garden, it seemed very bizarre to us that we'd just pick up a clue at this location and then immediately leave without doing anything else. But that's exactly what we were expected to do, and 5 minutes later a free hint told us as much. Why not send us directly to Emeryville and give us the clue at one of the places on the map? A cool thing about the puzzle itself is that it had us doing something we'd never done before-- gather data by driving around the city. We cruised the designated area in search of utility boxes bearing stick figures engaged in various mysterious activities. When we found an intersection that had such a box, we extracted a letter from the corresponding square of the map grid. A simple puzzle but one that all of us in the (blissfully air-conditioned) van were able to contribute to without leaving the (blissfully air-conditioned) van. Woohoo!

Manipulation: This was a really nice puzzle. We were given a set of hard plastic strips and ball-shaped connectors, and we had to assemble them into an icosahedron. Each strip was numbered and had something written on each end-- a city, food, mayor, year, or landmark. When properly assembled, all the items related to the same Olympic city would be attached to the same connector. Once that was assembled, we had to identify each face of the icosahedron by the sum of the values on its three edges. An accompanying clue had lists of these values, and we needed to roll the die from value to value according to that list to draw letters in the final answer. A very good team puzzle and a fun way to construct an object. It also contained a good implied puzzle-- once the die was assembled, how do you associate each sum with a face to facilitate rolling? After chuckling at the Burninators' frenzied attachment of numbered slips of paper to the struts, we wound up doing the same thing.

AIM: More Rubik's Cube fun. Now that we all had an unscrambled cube, we needed to label it in a certain way and then manipulate it to duplicate 18 patterns. As it turned out, each pattern could be created by finding the appropriate face (which, on a Rubik's cube, is always the color of the center square) and then rotating the rightmost column 1-3 turns. We initially tried finding a message on the final cube, and one face read "EXIT". But the hint device told us that was inconsequential. Next we reset the cube and tried to extract a message from the new letter brought into view to create each pattern, but that gave us garbage. Finally we tried the right approach, which was to look at the old letter being rotated off of the designated face, instead of the one being rotated onto it.

That brought us to a new location where we were given a reflective, filmy square and told to use it to look inside our problem. Also at the site was a giant metal cube, each face a solid color matching a Rubik's Cube. Three other teams were also at the site, and none had yet figured out what to do. Within seconds of getting our film, Andrew slapped the thing down on the surface of the giant cube and PRESTO-- letters appeared on the film, which reacted to magnets inside the cube. By moving the film around the cube, we revealed different letters beneath the surface at the same positions as the letters on our smaller cube. Unfortunately, we made another classic blunder-- leaving the van without a writing utensil or clipboard. By the time we rectified that problem, the other teams had gravitated toward the cube and there was no way for us to get our information without revealing the secret to the other teams. So we just shrugged, told the other three teams they owed us, and slapped down the film to start taking notes. Once we had that info, all we needed to do was replace the letters in our initial answer from the small cube with the letters from the same locations on the bigger cube to get our final answer.

Knowledge: A CD contained "news broadcasts" describing fictitious Olympic scandals in which medal winners in various events were found guilty of cheating. Their countries of origin weren't stated directly, but were clued obliquely. We had to transcribe the key data-- events, medals, and clues-- and then solve for the countries. Then we were stumped. We knew the medals were indices-- 1 for gold, 2 for silver, 3 for bronze. But indexing into the countries or the event names produced garbage. We banged our heads on this for over an hour, finally stopping to get a bite to eat and try to figure out what we were missing. We listened to the messages again and I noticed that "platform diving" had been transcribed as just "diving". When the event indexing looked like garbage (IILEDGEAL), Andrew had abandoned that approach partway through the data. This changed the second indexed letter from I to P, giving us IPLEDGEAL, which suddenly looked more promising. We finished the indexing and got I PLEDGE ALLEGIANCE, kicking ourselves for not being thorough in the first place since we would have spotted ALLEGIANCE at the bottom of our data had we just completed the process. I PLEDGE ALLEGIANCE clued us in to look at the flags of the clued nations. The fact that RINGS was the only repeated event in the puzzle and was clearly set apart from the others by spacing suggested that Olympic Rings were important. Putting two and two together, we realized that each flag contained some of the 5 colors of the Olympic rings, and that allowed us to convert the flag into 5-bit binary values, and thus letters.

Enigma: A yellow piece of paper curled inside a sealed tube, meant to suggest a urine sample, held our next clue-- seemingly a cryptogram. The flavor text about drug tests had us trying to decrypt the text as names of steroids or other drugs, but that went nowhere. Then Dave and I simultaneously noticed that in the two lines of flavor text, "justify" happened to be directly over "right". The light bulb went off, and we typed in each word of the puzzle text and justified everything to the right. Sure enough, one column of the text gave us a message, telling us how to decrypt the text. Doing so revealed a message in another column, telling us to Caesar shift all the letters in yet another column. That gave us instructions to shift yet another column, and so on, until we finally got an answer string.

Wild Card: A stash of arcade tokens led us to mini-golf course where we had to earn 400 tickets in redemption games to earn our next clue, a pack of baseball cards with Shinteki players on them (each team had been asked to send in photos of their members in baseball poses), and the scorecard for a fictional baseball game. We immediately set about extracting all the key data from the cards: the players' fake names, their positions, the numbers they were assigned, and the location of the baseball icon on their card (left, center, or right). There were 26 players, and their first names all began with a different letter of the alphabet. Eureka, an order. Now for their last names. If their baseball was on the left side, take the first letter of their last name. If it was on the right, take the last letter. Players with a ball in the center had the same first and last letters, so it didn't matter which one was used. That gave us a message telling us to construct a sudoku grid. The two numbers assigned to each player represented grid coordinates, and in baseball each position is associated with a number from 1 to 9. Plugging each player's position number into the grid at their given coordinates gave us our givens, and then we had to solve the grid. We'd used all our baseball card data, so now it was time to look at that scorecard, which showed the score for each inning. We noticed that only one team scored per inning. If the top team scored, say, 4 runs in inning 1, we needed to extract the 4th number down from column 1. If the bottom team scored 6 in inning 2, we needed the sixth number from the bottom in column 2. This gave us a 9-digit number. The final step was realizing that the digital font used for the scoreboard values was a hint-- we needed to turn our 9-digit number upside-down like a calculator and read the digits as letters for our final answer. We liked this puzzle right up until the scorecard. Unless you knew that you had to read the number upside-down, there was no way to be sure that you were doing the right thing with your number extraction. I'm sure many teams just extracted the nth number from each column from the top down-- going bottom-up for the bottom team was an arbitrary and unnecessary wrinkle. And our first guess was that each number would turn into a letter from A-I, so nothing seemed like it was working. An additional built-in hint for the calculator trick was called for here, I think, instead of relying on teams to make that leap from just the font.

Endurance: We had less than an hour remaining when we got to this puzzle, so we took more hints than we would have normally (and still finished about 5-10 minutes too late to get credit), so it's hard for me to judge the puzzle's fairness. I loved the presentation, which was an album-like sheet of poster board with a 10x10 grid of photographs. If you correctly identified the 10 photos across the top, the first letters spelled CATEGORIES. The idea was that the photos could be divided into ten different categories, with ten photos per category. Each picture had a ten-letter name. Create a 10x10 grid for each category, with each item listed in the order found in the grid, and the main diagonal of each grid gave you another 10-letter word. Recurse, putting all of these new words into another 10x10 grid, and you got the final answer. We had a number of problems with this puzzle. We never got all the categories correct, we had no idea who some of the people in the photos were, and some pictures could have gone into multiple categories. Differentiating became a matter of "the only way for us to get 10 items in category A is if photo X is an A instead of a B," which isn't very satisfying and feels messy. Category-based puzzles are very hard to construct without this kind of ambiguity. This ambiguity made the photos we couldn't identify even more frustrating, because we couldn't say for sure what category they went into. This seemed like it would have been a better puzzle for a puzzle hunt environment. The group-participation phase of identifying the photos was much less than half the solving time. Once that phase ended, the Excel phase began and it was hard for more than 2 people to be engaged. That's fine in a puzzle hunt with multiple puzzles happening in parallel, but in a Game I want the Excel phase to be small so that everyone on the team is having fun.

Teamwork: Time ran out before we could get to the final puzzle, although we took a copy home with us and look forward to solving it later-- possibly as a warm-up the night before Hogwarts in September.

Despite the withering heat, I thought this was a great event. The puzzles were a varied mix and generally excellent, and the entire course was packed into a manageably small area for minimal drive times. I was somewhat disgruntled after Shinteki: Untamed a year and a half ago, but Shinteki: Decathlon and Shinteki: Decathlon II were both solid, entertaining events. The easy availability of hints at your own pace make the Shinteki events very newbie-friendly, so if you've been intrigued by my puzzle event exploits I encourage you to find a trio of like-minded friends and take a weekend vacation to the Bay Area the next time the Shinteki wagon pulls into town (although with Shinteki principals Linda and Brent's impending baby, that might be a while). I continue to look forward to future Shinteki offerings.

Paparazzi

Paparazzi was a strange game for me. Once my allergy kicked in, my focus shrank to encompass my nose and very little else. I'm not even sure I've remembered the clues in the proper order. My perceptions of the Game come through a mucus-tinged filter.

Aside from my discomfort, there were a few noteworthy things about this Game. My team, Briny Deep, finally scored a first-place finish (and what a boost for my ego that it happened in a Game where my level of contribution was lowest). Let me tell you, it is far more exciting to be in the lead than to be anywhere else in the pack. We're used to arriving at a clue site and finding the Scoobies, Burninators, Blood and Bones, or Advil already there or there-and-gone. Nothing gets the blood pumping as much as arriving first at a clue site, except perhaps leaving first. We were strong right out of the gate, and were nipping at Blood and Bones' heels for much of the pre-dinner Game. We didn't rush dinner, so we left after some other teams, but we caught up to and passed them at the late-night poker clue and stayed in first place from then on. It was a real thrill to arrive at a clue site, solve the clue, and leave before any other team arrived.

Most of the puzzles in this Game were simple. By that I don't mean that they were easy per se, but rather that they didn't require obtuse leaps of insight, masochistic levels of grunt work, or multiple stages to navigate. While some were a bit too basic or familiar, most felt elegant in their simplicity. Most importantly, they were fun to solve. Solving a puzzle quickly-- without being at a site for over an hour, say-- is both satisfying and energizing. You feel like you're moving forward, making progress, and not falling behind.

For the first time in our experience, a Game faced the problem of the top teams getting way, way ahead of GC's expected schedule. Games typically end around 4-6 PM on Sunday. We arrived at the finish around 10:40 AM. The Burninators were only ten minutes behind us, and by 12:30 six other teams had arrived. But I don't think the next wave of arrivals started until about 3. There was a very large and distinct gap between the front and the rest of the pack. There have always been teams who have outperformed, but when they stop becoming the anomaly and the field splits into two clearly separated groups, that's a huge problem for organizers to deal with. Teams start arriving at sites that aren't open yet, or a GC equipped to handle five sites simultaneously finds itself with teams at seven or eight. For most GCs, the big problem is what to do about teams that are falling too far behind the expected schedule. Now they'll also have to plan for teams who get too far ahead.

XX-Rated handled this as best they could. They skipped trailing teams over some sites, which is the standard way to close the gap. A couple of clues were designed to be adjustable to slow teams down or speed them up. GC took full advantage of this, holding teams at one location for over four hours with hyper-accelerated dance pad puzzles. In one case things just slipped through the cracks, and the top team idled for an hour (and another team-- us-- waited twenty minutes) until a site could be staffed. Everyone on XX-Rated felt terrible about the scheduling snafus, but as part of a team that was always on the front lines of the problem it never really bothered us. There was some waiting, but we understood why and we'd much rather be breaking things from the front of the pack than be skipped from the back.

I know what it's like to have things in your Game go differently from how you planned them. I beat myself up over things that didn't go quite right in The Mooncurser's Handbook, but the truth is that those things didn't matter to most players. The little bumps are forgotten amid the overwhelming number of other things that go right. XX-Rated did a terrific job. Did everything go perfectly? No. But players had fun, and that's what's important. This was a solid Game, and everyone on our team enjoyed themselves a lot (modulo allergy problems-- future GCs, please no more Napa!). We're grateful to XX-Rated for all their hard work.

Here's a rundown of the clues.

Fashion Show: A series of LED displays that seemed to be blank. In fact they were infrared, and could only be seen on the LCD screens of digital cameras. Our team couldn't help but laugh because we had planned a puzzle for Mooncurser's that took advantage of this same property, but the location got cut. Unfortunately, with all teams present at the same time, there was little in the way of "aha" here-- any team that didn't know about this property of a digital camera was able to pick it up easily from the activities of everyone else.

Headlines: Sets of newspaper headlines. Each headline described a movie, and each movie title contained either a number or a mathmatical operation, allowing each set to resolve to a number in the 1-26 range. This was a fun puzzle well-suited to group solving.

Passport: After filming a brief bit of dialogue (obviously to be used later in a puzzle), we received a passport stamped with circular and rectangular visas from many foreign airports. These formed a Morse message hinting that the duration of each stay was important. When sorted in chronological order, the length of each stay could be converted to letters (1-26), instructing us to remove the Opec countries. Doing so and reading the 3-letter airport codes from those left over yielded the final answer. This was also a very good and thematic puzzle, with each step internally clued and leading to a satisfying finish.

Shoots and Ladders: A checkerboard with letters and velcro dots, onto which variously-sized velcroed "shoots" and ladders had to be properly attached. Playing the game according to provided die rolls spelled out our next destination. This was nicely conceived, although the assembly proved more difficult than we expected until we got the hint that the position of the velcro dots within their squares indicated the direction that the attached shoot or ladder would go.

Grape-stomping and encrypted word search: In an I Love Lucy moment, one of us had to stomp "grapes" (water balloons) while a teammate collected the water until three jugs were filled. Popping those balloons was much harder than it looked-- those suckers kept sliding out from underfoot. Once done, we received a bottle of wine with an encrypted word search on the label and a list of search terms on the inside. Unfortunately an old version of the wordsearch with a spelling error snuck in which slowed us down, and frankly I've seen encrypted wordsearches many times before and this particular version put no new twists on the genre.

Wine menu: A nice multilayered puzzle in the form of a wine menu. Adjectives in the wine descriptions could be paired with adjectives in the names of the wines (e.g. SPICY and KICKIN'). The relationships of these sets can then be mapped onto a watermark of grapes in the background of the menu, which acts as a key to the subtly highlighted grapes around the perimeter. The set graph was an innovative approach that I hadn't seen before, and everything you needed to solve the puzzle was there to be discovered. Very nice.

Discs: The biggest clunker in the Game consisted of a set of seven segmented, transparent discs which had to be colored in and then rotated atop each other such that no transparency remained. Unfortunately there was no system to follow here, just a brute force search among many, many combinations. So many, in fact, that our team finally resorted to writing a computer program to find the right one. Any puzzle that can be solved faster by writing a computer program than by hand-solving has serious problems in this kind of event. For one thing, only one person can manipulate the discs at a time, leaving the rest of the team to watch, wander off, or try to simulate the discs in another medium (or write a computer program...). Each disc had one radial line that was thicker than the rest, and the puzzle would have been tractable if none of the thick lines overlapped each other in the correct configuration, but that wasn't the case. The result was a puzzle that far outstayed its welcome. Many teams got mercifully skipped over this one.

Jigsaw: A jigsaw puzzle and accompanying spaceless cryptogram that was fundamentally broken-- even without spaces, the crypto fell in seconds when plugged into SCBSolver. We assembled the jigsaw before even trying the crypto, but Blood and Bones attacked the code immediately and was gone in two minutes. The image on the puzzle was intended to provide the key to the cipher. Teams definitely got ahead of GC's schedule here.

Marina: At a marina with views of multiple shipyards and engraved monuments all over the place, we received digitally altered photos of these engravings and had to perform the same modifications to another piece of text to produce an answer (so if one photo had a word with the R replaced with ON, we had to do the same to all Rs in our text before applying the next alteration). A nice divide-and-conquer to find the right engravings, then a satisfying solve as we applied the changes.

Typesetting: Four-word sets in which one letter of each word could be changed to form another word, with the resulting changed words having something in common (ex: BED, GREED, MELLOW, BLUR). A four-column grid appeared at the sides of the page, and shading the boxes corresponding to the words whose changed letters are part of PAPERAZZI (a step clued by a message formed from some words that were printed every so slightly bolded) created letters in the grid to form our answer. A good group puzzle to solve over dinner.

Fashion Set: A customized game of Set with five tableaus. A Set tableau is 4x3, making it possible to encode Braille letters by looking at which cards are or aren't used in any sets in the tableau. I felt like I'd seen this puzzle before-- I believe there was a Set puzzle in a recent Microsoft Intern Puzzle Day-- but it was nevertheless fun to do and nicely conceived.

Baseball: At the statue of Willie Mays in front of the SF Giants' stadium we received a baseball "autographed" by many famous players, and a story which put various world cities in a sequence. Close examination of the ball revealed that each signature had a star by it, and that if the baseball was treated as a globe those stars corresponded to the cities in the story. Listing the players in the order of their associated cities and taking the first letter of the first player, the second letter of the second, and so forth produced the answer. Using the baseball as a globe was clever and this was a satisfying solve.

Letters to the Editor: We were given a copy of the current issue of Shape magazine, and a bunch of letters to the editor that didn't make it into print. We had to count the number of words in each letter, then turn to that page in the magazine and find a photo matching the description of the letter's author ("Jumping for Joy", "Reaching for the Stars", etc). Treating that person's arms as semaphore flags yielded the answer. Counting words was a pain, but could at least be parallelized across the team and was something even I, in my allergy daze, could help with.

Girlz Dance Mix: An empty CD case with thirteen sets of four words on the back. Each set had some common characteristic, and one word in each set didn't belong. Each of those words, when the right letter was added, could be anagrammed into a woman's name. Those extra letters could then be arranged to form the answer. It was here that the nicer clothes we'd been told to bring came into play, as our team was ushered into a limousine and driven to a downtown dance club where we had to find and dance with the right person to get the CD that went along with the initial case. Each track on that CD contained a woman's name, giving you a hint about what to do with the odd-man-out words and an order in which to read the added letters. Our team, however, solved the puzzle without the CD-- but we went along for limo ride anyway. For one thing, GC needed us to so we wouldn't get even farther ahead of schedule. For another, we'd all brought the nicer clothes as felt like we'd might as well use 'em. And who's going to turn down a nice limo ride? The payoff for all that effort was lacking, though. Once at the club, we spent very little time inside and didn't have to do anything interesting. We felt like there should have been something more involved going on at the club to warrant all the extra effort taken to get there. A dance marathon, karaoke, an embarrassing group dance, a back-room rendezvous... you get the idea. The puzzle itself was great, but the trappings didn't live up to the promise.

pr0n: A video CD of various animals mating. Each clip's length was in multiples of 5 seconds. Divide the length of each clip by 5 and index into the word for that animal's offspring to get the answer. The clues came a little too quickly here-- I think we'd barely got all the data entered into Excel before the device told us that offspring were important. It would have been nice to have had more time with the puzzle on our own before the hint pushed us into the puzzle's major aha.

photo rebus: At Walgreen's we got a packet of photos culled from teams' applications, with elements of each photo highlighted and notations in the margins. This was a photo rebus, and was pretty easy but fun to solve.

ILM film strip: A film canister retrieved from the campus of Industrial Light and Magic contained a long strip of paper with photos on them, with each photo representing a letter from the phonetic radio alphabet (ALPHA, BRAVO, CHARLIE, etc). This fell in minutes and seemed too basic for a Game.

poker: One team member got some chips and played limit hold 'em against other teams and house players while the rest of the team solved some Paint By Numbers puzzles and the start of a logic puzzle. The Paint By Numbers supplied some info that was missing from the logic puzzle, but not everything. The rest-- parts 2, 3, and 4 of the logic puzzle-- could be purchased for 100 chips, and every so often each team was given a chip infusion-- 10 chips the first time, 20 the next, etc. This was a great stop for our team, since our player had a great run at the table and we managed to guess the info that part 4 would provide, solving the puzzle without ever buying that part. This is where we took the lead, arriving in third place but leaving in first. The activity design had some problems, though, mainly in that only one player could play poker while the rest of the team, once they'd done all they could do on the logic puzzle, sat idle. Since teams were at this site for a couple of hours, that wasn't ideal. Generally you want to keep players busy, not have them sit around waiting. When the waiting is because you're excelling at the Game and are ahead of GC's schedule, you don't really mind. But when the waiting is designed into the very nature of the puzzle, it's much more frustrating. My frustration, of course, was with my nasal passages.

Starcrossed crossword: A simple, star-shaped crossword puzzle in which all occurrances of "STAR" needed to be filled into a single square. The squares adjacent to each star always contained the same letter, and these letters could be anagrammed to spell the answer. Another puzzle that was almost too simple for a Game, but coming right after the marathon poker site it was just the right length to get our energy up again.

Donuts: At a donut shop we got a box of fresh donuts and a newspaper story listing a bunch of actors and their diet plans, each of which contained a number. Each actor had appeared in a movie whose title could be matched to one of the donuts (TWISTER for a twisted donut; BLUE CRUSH for a blueberry donut; etc). The numbers indexed into the movie titles, and the arrangement of the donuts gave us the order in which to read them. And then we ate the donuts. The puzzle required IMDB, but the donut shop had free wireless (supposedly-- it wasn't working when we were there, but we hitched a ride on The Burninator's mobile network. Thanks, guys!).

Stanford: The slowdown from hell. Upon arrival we found a room full of computers set up with dance pads, and were told we needed to complete five different dance games (traditional DDR, a Simon-like memory game, Whack-a-Mole, trivia, and a maze). Only one person from each team could be dancing at a time, but our team could have a second dancer thanks to our nifty application. The problem? Each of the games was hellaciously fast. Andrew's almost good enough at DDR to pass as an Asian teenager, and even he had trouble nailing the dance game. Not only did you have to succeed, but you had to record the series of arrows so you could transcribe them onto a map of the Stanford campus, where each path led you to a bag of tiles. We arrived at 4 AM. We were the first team to leave, at 8:20 AM. That's a LONG time. GC needed to keep teams there for logistical reasons, which is why the games were set to such a high speed. But it was very frustrating, and I imagine it was even more so for teams who could have only one dancer. As more teams arrived and all the stations filled, teams had to switch off to give each other a turn. It would have been agonizing, except that no teams were being allowed to leave anyway-- so even if we'd been able to solve everything, it wouldn't have done us any immediate good. Knowing that made repeated failure tolerable. At 7 AM they reduced the speed a bit, and they planned to do so again at 9. After all the practice at insane speeds, we finished the games we were missing as soon as the speed dropped. The tiles were a straightforward, 4-sided domino assembly puzzle that went smoothly. Once assembled and inverted, the other side showed a photomosaic of Donkey from Shrek composed of real and cartoon animals. Each title contained six tiny photos in 2x3 grid, and the animated/real status of each photo created Braille letters spelling the final clue.

Word association strips: A set of vertical and horizontal strips with word chains, with some words missing in each chain. When properly filled in, some words appeared on multiple strips. Overlap those strips to form letters. This puzzle bugged me because I would have designed it such that all the overlaps happened on words that needed to be filled in, and in this puzzle the overlaps were inconsistent. Some were with two provided words, some with two filled-in words, and some with one of each. That inconsistency was like nails on the chalkboard of my sense of puzzle elegance, and my teammates had to drag me along kicking and screaming. "I see... FOUR... lights!"

Book store: At Borders we got a marked up table of contents where five chapters had been highlighted and annotated to form titles of other books. We had to find those books in the store, where we discovered bookmarks in each one at a certain page. Those five 2-digit numbers could then be strung together to form an ISBN number (my big contribution to the team, with my allergies feeling a little better) of another book, which contained another bookmark with the answer. The concept for this puzzle seemed problematic. There was nobody else in the bookstore when we did this puzzle so we had no problems, but with multiple teams there at once it seemed like it would be impossible to solve this without giving it away to other teams nearby.

Erin T's gossip: A gossip column in which the Scrabble hook EINRST was anagrammed multiple times, along with one other letter, into words incorporated into the text. This one fell in minutes and was just a blah puzzle. We've seen the same basic concept of spotting anagrams within a larger text before (Shinteki: Untamed, for example), and like the encrypted word search this puzzle simply added nothing new. It was also a pure aha. Our team saw it very quickly, but if a team didn't spot the anagrams there was no internal cluing to guide them. I'd have cut this one (but then, having too many puzzles wasn't a problem this time).

Casting call video: All of the snippets videotaped the day before got spliced together to create an argument between two characters. Each snippet contained a quote from a movie, and we had to identify the actor or actress who was doing the talking and read the first letters of their last names. We nailed the last 7 solidly but had a bunch of misidentifications up front. We were very solid on _OLLF_______ACTRESS though, and Kenny just focused on that and pulled MOLL FLANDERS ACTRESS out of his butt, and sure enough ROBIN WRIGHT was correct. It's always fun seeing your teammates and friends on film and identifying the quotes was fun, but this was the third IMDB puzzle in the Game and that's really too many, even for a Game with a paparazzi theme. Note to future GCs: In a Game environment, where wireless access can be spotty and not necessarily available to all team members, puzzles where the gating function is database access are perhaps not the best idea.

Da Vinci Choke

It's a good thing I didn't want to pay taxes on the grand prize, because my final time of 64 minutes is definitely not good enough. The final challenge consists of hard versions of all the earlier puzzle types except for the observation challenge, without any trivia or research. This means you actually need to be able to solve a chess puzzle, for example. For fun I noted my times after each puzzle:

Symbols (9x9 Soduku): 19 minutes
Restoration: 27 minutes
Curator: 47 minutes
Chess: 59 minutes
Jigsaw: 64 minutes

The soduku was challenging just because they weren't numbers, which made it harder to scan for the missing elements and hold that information in my head. When I got down to most of the grid solved except for the lower-right corner, I converted to numbers so I could "see" things better and that let me finish.

Restoration was easier than I expected. It took me 4 or 5 attempts before getting to a point where my gut said I had a solvable configuration, and then I just stopped and thought it through before making any more moves.

Curator KILLED me. I kept getting everything arranged and needing one more nail. Finally I got systematic about it and figured out the only 2 or 3 places the 3x3 painting could go given the other nails and paintings, and it quickly fell apart after that.

Chess was the one I was most worried about. I know the rules of the game but I'm not a chess player and don't do chess puzzles, so I had to slowly and methodically reason through the options ("In order to get a mate, I need to attack these squares. How can I do that in just two moves?"). If I'd gotten desperate I could have just cycled through the 64 possibilities as quickly as possible. That might have actually been faster. But I was nevertheless pleased with my solve time here.

The jigsaw was a nifty variation where the picture kept changing, making it harder to match up the pieces. The traditional approach of assembling the outer edge first still worked.

I'm guessing the winning time will be posted by a team approach, with a group of people collaborating on the Soduku, Curator, and Chess puzzles (I don't think more brains will help with Restoration or Jigsaw). It'll be interesting to hear what that winning time is.

The Apprentice: Zorg

This weekend I was in the Bay area to participate in a 15-hour Game called The Apprentice: Zorg. This Game was notable for two things. First was the very strong theming throughout the event, which essentially followed the plot of The Fifth Element with occasional artistic licenses. Four "ancient languages" (Braille, binary, Morse, and semaphore) and the periodic table were recurring themes, we traveled from Earth to Fhloston Paradise, we recovered and activated the four stones, we got a clue from Ruby Rhod, and ultimately changed a doomsday weapon into an inert ball of carbon. Most Games I've played in have fallen flat on the theming, which tends to be tacked on via supplementary material and not incorporated into the puzzles themselves, or else the subject matter is used in puzzles but without any coherent story behind them. The Apprentice: Zorg managed to follow a recognizable plot AND connect the puzzles to that plot and theme, which certainly enhanced my experience.

The second notable thing about the Game is that, while many people helped run the actual event, all of the puzzles were created by only one person. That singularity of vision helps explain the theming success. Ian, my headwrap's off to you.

One of my personal takeaways from this Game (and this isn't a criticism of Zorg, merely a realization I came to while playing) was that, when we're being scored for completing puzzles, getting skipped really stinks. Even without a scoring system, skips bum us out because it means we're experiencing one fewer puzzle. That puzzle might be offered at the end of the event, but in practice we never get together to do them later. In this case the skips were particularly frustrating at the time-- for one we were near the front of the pack, in another a timed puzzle came when our team was separated, and in a third we were encouraged to take a dinner break while solving, but then got skipped over the next puzzle as a result. So the skips seemed especially capricious. I understand that from a design perspective, some of what we skipped might have been "bonus" puzzles expressly for the purpose of slowing down the fastest team, never intended for all teams' consumption. And having run a Game, I understand why skips are necessary. I think managing teams' expectations can make a big difference. When the device showed us that there are 21 puzzles at the start of the event, we expected to see 21 puzzles. If the device hadn't shown us that information and the puzzles hadn't been numbered, we might never have known about the skips until afterward-- a situation that's preferable to knowing in the moment that we're being skipped. I'd also be reassured if I knew GC had planned skip points that only occurred at their weakest puzzles.
.
Speaking of which, The puzzles:

ZF5: An instruction manual for the ZF5 weapon, in which photos and text could be expressed as two-letter answers and entered into a grid. A perfectly fine get-your-mind-started opening puzzle.

Presidents: A sudoku puzzle in which the data for each row was found by identifying wives, pets, daughters, foods, running mates, election years, and quotes from the last nine presidents. We took much longer than we should have needed for this-- we made the classic Puzzle Hunt mistake of focusing on all the component parts first, rather than plugging our data into the sudoku grid as we acquired it which would have allowed some parallel solving.

Paths: Gather crossword clues posted throughout a park, solve for words having all five vowels, plug them into a circular path, and extract a message by following branching trees from the center outward according to all possible arrangements of five vowels (in alphabetical order). An error in a hint convinced us we had the wrong approach, causing us-- and most other teams-- to time out on this puzzle.

Magic: Perhaps expecting a better spread from the previous two puzzles, this array of Magic cards was located in a small interior space that couldn't accommodate more than 3-4 teams at once. When all teams started arriving within a few minutes of each other, GC shut down the location. This lack of scalability felt like a rookie mistake. The puzzle could easily have been distributed as a paper printout. We never solved this sodoku-like puzzle. Instead, as Dave described it to a teammate who stayed behind in our van, he intuited what the right answer was in a flash of insight based on the fact that the first letters of all the cards were either A, B, C, D, or R. ABRACADABRA, we were on our way.

Berkeley runaround: A series of puzzles utilizing the "four ancient languages" began with a fake BART ticket with four colored stops listed and a series of numbers which we quickly decoded to SEMAPHORES. Finding each stop on the BART map and interpreting the corresponding colored line at that point as semaphore gave us the answer. Next we were directed to a gelateria to order six specific flavors, which we did (yummy!). When we got nothing special along with the gelato, we realized the positions of these flavors in the refrigerated case formed Braille. Next was a traditional-style logic puzzle creating 5-digit binary, and finally a game of charades in which each word being charaded was a typographic symbol that could be interpreted as Morse code. Each of these puzzles yielded a set of six wooden cubes with letters on four faces, a number on the fifth, and a photo on the sixth. The photos in each set were thematically linked to earth, air, fire, or water. But we didn't know what to do with them. To figure that out, we had to play the Game of Four-- a game of 20 questions to figure out a six-letter word, but where our own questions couldn't have more than 4 letters per word (which, of course, we had to figure out for ourselves). The correct answer was TETRIS. The blocks could be grouped, one from each element, according to their pictures ("Things that move through their element", "Things that destroy their element", "Things that expel their element", etc) and arranged into Tetris shapes by matching letters on adjoining faces. These pieces could then be arranged, number-side up, into a 5x5 grid. Translating each number into the corresponding symbol from the periodic table gave us our answer. The runaround was fun-- the puzzles were the right level of difficulty for such a structure, and the cubes were a satisfying meta-puzzle. I don't think the meta was sufficiently clued internally-- there's no way we would have discovered the cross-element categories without being told to look for them. This was the first puzzle where we felt the hints started coming a little too quickly, before we had adequate time to utilize earlier hints. Getting that timing down is very hard. Too slow and you risk teams getting frustrated, too fast and you deprive teams of the satisfaction of solving for themselves.

Next we went to a children's museum on the Berkeley campus, where we were skipped over a puzzle involving the giant DNA strand outside the museum and went directly to a meeting with Zorg in "the boardroom" inside. There we found an array of numbered pieces from various board games. We made short work of this one, recognizing that each of the games was played on a grid and turning the size of each grid into a letter (only afterward did I realize that this step had been elegantly clued by placing the puzzle in "the boardroom").

Next was a scramble inside the museum, with our team working on two puzzles at once-- a trivia search and another use of the four ancient languages. Solving the first gave us a challenge to compose a rap incorporating a subset of stuff from a list of possibilities-- a reference to an Apprentice contestant, a shout-out to another Zorg team, a line using all letters of the alphabet, a reference to an exhibit from the museum, etc. Andrew dropped what he was doing to work on this, producing and performing a fine, fine rap and giving us enough info to backsolve the meta and skip the fourth puzzle involving a map of the area.

The aforementioned timed puzzle came next, before the whole team was together or back at our van. The device fed us word pairs every minute, and after four of them we realized they were thematically linked to the 12 Days of Christmas. This, along with hints, sent us on a wild goose chase trying to count the number of total gifts in the song and trying to index into each pair according to the number with which that pair is associated in the song. Only after the puzzle's virtual bomb "exploded" did we try the far simpler approach of extracting the common letter from each pair to get the answer, but since the bomb exploded we were skipped over the bonus scavenger hunt puzzle which could have been solved in transit to the next location. I liked the idea of a timed puzzle a lot, but we it started at a moment when we weren't prepared for it. I would have preferred the device to give us a very strong warning not to enter the start code until we were ready.

I thought the next puzzle was clever and fun-- a flow chart disguised as 12 pages of a Choose Your Own Adventure book. The page numbers and goto references on each page, however, used a different numbering system that was clued by the content of the page. So an encounter with mole people told you that the numbers on that page were written in moles; a page with lots of traffic signs told you to translate the numbers as sines. Once translated you had a graph of letter nodes and paths between them, and you had to navigate the graph to generate a message. I think the puzzle would have been more elegant if each path was used only once, but otherwise I liked this one quite a bit.

A box of "Gemini Croquettes" contained a bunch of animal shapes with data in them. This,was an eastern/western zodiac puzzle, and was very nicely clued internally by messages on the box. We solved this pretty leisurely over dinner, and so our time on this one wound up being much longer than it would have had we all hunkered down on it at once, but it hung together nicely. Unfortunately the price of our dinner turned out to be skipping the next clue, the Multipass, which was a Roman numeral multiplication table with symbols instead of Roman numerals.

A CD from Ruby Rhod was one of the most disappointing puzzles in the Game. The material was terrific-- a radio broadcast from a very good Ruby Rhod impersonator-- but the puzzle it contained was overly simple and resolved to SPIN THE CD. The CD label had an obvious rainbow pattern of boxes and lines on it, and spinning the disc didn't produce anything interesting other than the rainbow effect you'd expect. We stared at it for a minute or so, not seeing anything, until someone said, "Let's just try RAINBOW," which was the correct (and anticlimactic) answer. It was just so obvious that none of us thought it was the answer-- we were expecting a bigger eureka moment.

Next we went to Fhloston Paradice for a pouch full of 20-sided dice with some values on each die colored in. Mapping these values onto a grid yielded a picture: SNEAKY 124533631. We thought SNEAKY was part of the answer, and it would have been nice for the device to tell us that it wasn't. We eventually solved it by plugging the numbers into TEA to yield SNAKE EYES, and only later realized that the numbers were indices into SNEAKY.

We went to the Diva for a concert, distributed on CD, but before we could listen to it mercenaries arrived and it was off to Q-Zar for a game of laser tag while trying to read the letters written on players' arms in UV ink. This was thematic and a lot of fun.

I also enjoyed the subsequent puzzle in which team names were combined with each other and represented pictorially (Briny Taxi, Sharks on a Plane, etc). The coolest aspect was that teams weren't allowed to solve the puzzle alone-- we had to wait until another team arrived, and then solve it together. Nice!

The STONES puzzle-- in which we had to activate the four stones by applying the appropriate element to them-- was inevitable, and was a valiant attempt, but one that fell completely flat. We were given four numbered canisters, each containing something different-- a balloon, a capsule, water, and a lot of a white powder that turned out to be Borax). The markings on three of the canisters looked identical, and while it was clear that the one with the balloon represented air, it wasn't clear which was fire, water, or earth. A hint finally told us that we would get a color from each canister after applying the right element (we'd already extracted an orange Skittle from the innermost of the three nested balloons) and that the powder burned yellow when ignited. The capsule was clearly one of those foam things that blooms in water, but adding dirt to the water in the last canister didn't seem to do anything. We were supposed to add the red powder we'd been given at the very start of the Game, but there really wasn't any way for us to know that. A for effort, but the whole puzzle just fell completely flat.

Next came a set of five puzzles, one for each sense. I took the TASTE puzzle and quickly discovered that all five of the cupcakes were, in fact, chocolate-- just like they looked. Ripping the top off each cake revealed Skittles inside, which allowed each cupcake to become a letter when the Skittles were arranged in ROYGBIV order and treated as bits. The other four puzzles also involved 5-bit binary via sign language, Japanese characters, scratch-and-sniff stickers, and audio wave forms. The result told us to combine the white powder from the stones with the glue we got at the start of the game. This produced a viscous foamy substance. We entered FLOAM into the device and were told we were close, but they wanted a brand name with two five-letter words-- SILLY PUTTY (which I think is nothing like the substance we created, but why quibble when we got the right answer?).

Finally, a soccer ball with some hexes numbered was supposed to be a game of Minesweeper, but Dave once again pulled the answer out of his butt. I'm not even sure where the inspiration came from this time, but 7 minutes after getting the ball he just plugged in CARBON. Sometimes it's better to be psychic than good.

We enjoyed ourselves quite a bit, and appreciated the Mooncurser's Handbook shout-outs Ian scattered throughout, from the "____ module" format of the hints to the presence of "tsnuamic hydrogel" as one of the rap terms. We liked the 15-hour length more than Shinteki's typical 12. We're getting to know more of the Bay area players now, but I'm horrible with names and wouldn't mind if future Games made everyone wear name tags so I can start associating names with faces. I look forward to seeing everyone again in six weeks for Paparazzi.

SNAP 2: Fools Rush In

April 1 marked the start of puzzle season and the beginning of what will hopefully become a new recurring puzzle event in Seattle, SNAP (Seattle and Neighboring Areas in Puzzling). This debut outing, SNAP 2: Fools Rush In, took place in Seattle's Fremont neighborhood, the self-proclaimed center of the universe, from 1 to 6 PM. Twenty-five teams of four players each traveled throughout the area on foot, solving twelve puzzles before arriving at the finish. Three hint envelopes were provided with each puzzle; each opened envelope cost a team points. No electronic devices or reference materials were allowed other than the code sheet provided by the organizers.

Despite the event's title, our team decided not to rush. Our goal was simply to finish all the puzzles within the time allotted without taking any hints. We succeeded and even finished in third place behind two other hintless (but hardly clueless) teams, having a great time in the process.

The puzzles were all of appropriate difficulty for the event and most were conducive to team solving. The notable exception was the cleverly titled "Fool's Russian"-- a cryptogram utilizing Cyrillic-looking characters obtained at Fremont's bronze Lenin statue-- which (like all cryptograms) was difficult for more than two to work at once. I particularly liked a puzzle at the sundial atop the hill at Gasworks Park consisting of a dozen letters posted around the sundial and a dozen photos of different views from the sundial. When the most obvious approach-- face the same as each photo and read the letter you point toward-- failed, discovering the correct solution was a very satisfying process of utilizing all the information contained within the puzzle. A puzzle involving braided strands bound by rings was also elegant, with a satisfying "aha" upon realizing the datastream was encoded with each strand, rather than each ring. I also liked the Center of the Universe puzzle involving chains of compound words and phrases, which is one of my favorite puzzle paradigms. BRAIN _____ PROOF stumped us, but happily the puzzle was designed so that we could work around such a roadblock.

A "count the number of words before the word FOOL in these song fragments" puzzle suffered from its outdoor location, where it was hard to distinguish syllables from words. The puzzle would have been more successful in a more controlled environment, or with headphones. I heard a number of teams complain about a data collection puzzle utilizing a local mosaic, but our team had no complaints. In fact, since the data to be gathered was distributed on twelve strips of paper, I liked that the work could be split among the team and all of us could contribute. Only 3-4 teams were there when we arrived, however, and since the mosaic was on the ground I can understand how it might have been a nightmare if it got more crowded.

I have no idea how effective the hint system was, since we used none, but preparing three clue envelopes for each team for each puzzle must have been a bit of a pain. Since each site was staffed, the effort could have been avoided and the system streamlined by having the staffer hold one copy of each hint and a checklist for each team. A team requesting help would be shown a hint and checked off on the staffer's list.

Snap 2 was a great inaugural effort-- Shark Bait put together a solid, fun event. I hope we SNAP again soon. Next up this weekend... The Apprentice: Zorg.

Seattle Puzzling Alert

If you live in the Seattle area, have been intrigued by the accounts of my exploits at local and Bay area puzzle events, but have been put off by the travel costs or the time committments, I've got some great news for you. A new event format is being introduced this April 1 that's tailored just for you. Teams of four, walking only, five hours long in an area under five miles wide, and $25 per team. It's being run by some swell folks and should be a great time. Only 25 teams will be accepted, however, and those slots will probably fill very quickly-- so if you're interested, you should jump on it as soon as registration opens on February 15. Visit the SNAP site for more details.

Puzzle Hunt 9

This weekend I participated in Puzzle Hunt 9: Doomsday on the MS campus. Captain Micropolis, in the midst of announcing his retirement, was seemingly dispatched by a new threat named The Puzzler and it was up to us to get to the bottom of things and foil The Puzzler's plot.

Rather than using the traditional "wave" structure for distributing puzzles, this hunt used a self-paced structure in which solving a puzzle unlocked access to one or more new puzzles in six different parallel storylines. For us and the other top teams, this structure worked very well. A steady flow of new puzzles to look at kept interest high. I've heard that less successful teams had quite a different experience. To get new puzzles you had to solve the ones you had. Teams unable to do that had to bang their heads against the same puzzles for hours, and didn't get the kind of help they wanted from the organizers to unblock them. That's a shame, and I'm surprised the organizers didn't implement a time-delay back-up system, where if a puzzle hadn't been unlocked by a certain time it would get unlocked automatically regardless of whether or not the team had "earned" it.

Aside from the unlocking system, the structure of this hunt bore a frightening similarity to Puzzle Hunt 8:

The puzzles were, by and large, solid efforts. I didn't think any were particularly innovative or surprising, but with perhaps one exception (a Googlefest involving many, many dates) none were real stinkers, either. I did especially enjoy a Scrabble puzzle in which each play was a bingo. Finding the bingos was usually trivial (thanks to anagram software), but searching for the right spots to place them was fun and provided the most satisfying moments of the hunt for me.

I arrived at the hunt after staying up all night at an event in Dallas. My flight left Dallas at 7 AM and arrived in Seattle at 9:30, just barely in time for me to get to the 10 AM kickoff. So with no sleep except for a few sporadic winks on the plane-- and anticipating no sleep on Saturday night either-- I expected to fall unconscious during the Hunt or else be so hopped up on caffeine that Robin Williams would look somnambulant. Somehow, however, I remained coherent and functional throughout. I suspect this is at least partially thanks to the unlocking Hunt structure, which provided a steady stream of new puzzles to look at and avoided the ennui that can set in when all that's left are the inscrutable, intractable beasts. We spent much of the hunt flip-flopping with Scrubbers for first place before defeating the Puzzler around 11 AM and claiming victory, marking the first time a team has won two hunts in a row. Of course, usually the team that wins a hunt runs the next one, so that's perhaps a dubious achievement at best.

Most importantly, our team had a lot of fun during this hunt. We made continual progress, experienced very few moments of homicidal rage at the organizers, and kept in good spirits throughout. Perhaps easy to do when you're in the lead, but significant nonetheless. Many thanks to Everyday Heroes for putting the event together. I know how much work it is, and how thankless it can seem-- especially when everything doesn't go perfectly. Good job, friends.

The Mooncurser's Handbook

The Game is a 30 hour mobile puzzle challenge in which team of 6 players each travel in a van, solving a clue at each destination that leads them to the next. Information about the Game's history can be found here, and official details about our event can be found at The Mooncurser's Handbook web site. Early in our planning process, we agreed that we wanted our Game to be more social than most. Often in a Game you see very few other teams, and even when you do you're all embroiled in whatever puzzle is facing you and may not take time to chat. We wanted teams in our Game to have a chance to get to know and interact with each other. Towards that end, we tried a few experiments. We knew going in that reactions would range from love to hate, and decided we were OK with that. Hopefully more people were in the former camp than the latter.

The big experiment was dividing the game into four legs, with a pit stop between each one, and serving teams a meal. When teams arrived at the pit stop, they were fed a complete meal and given time to relax, chat with other teams, and participate in an activity that would translate into a slight advantage in the following leg. This structure meant that all teams would start on roughly equal footing four times during the Game, and no team would get too far out in front of the pack. Everyone would be seeing each other throughout the event.

The second big experiment was running a trading game concurrent to the traditional puzzle-based Game. The trading game was inspired by Sid Sackson's Haggle, but tweaked to better fit a Game format. Instead of getting a bunch of stuff at the beginning and then first trading for info and then trading their commodities, teams got a few goods at the start of each leg and more upon arriving at each clue site. Upon leaving a site they got a manifest of what would be available at the next one, with the idea that figuring out what they wanted at the next stop would give teams something to do during the drive between sites. Information was also doled out a little at a time through the Handbook itself (see below), but all teams got the same information at the same time-- so trading information wasn't really a factor. Teams [mostly] didn't have to worry that everyone else knew something they didn't. At each pit stop teams had to hand in their goods to us for scoring, and anything they kept was worthless in the next round.

The goods themselves were custom-made trading cards with the name of the item, an illustration (all drawn by GC member Dana Young), and pertinent stats about the item (in round one, this meant whether the item was ANIMAL, VEGETABLE, MINERAL, or MANUFACTURED; in round 2, items were given some combination of additional qualities FOOD, DRUG, SEX, UNETHICAL, DANGEROUS, and RARE; in round 3, each item had a QUALITY rating of 0 to 10). Additional information about each item could be found by looking it up in the Mooncurser's Handbook.

The trading game existed to give teams something to do during drives, to encourage teams to interact, and to manifest the smuggling/trading angle of the Game's theme. Rather than have an overarching story, we decided to make the trading game, and the universe revealed by the Handbook in the course of pursuing it, our story. Some teams found it to be overwhelming, with too much data to synthesize effectively. Others really got into it and were actively trading right up until the last possible moment.

The Mooncurser's Handbook itself was a Zire app loaded with data, none of which was initially available to players. Upon finding each clue teams got an arrival code which, when entered into the Handbook, started the automated hint system for that clue. At predetermined intervals, the Handbook Intelligent Neural Tracer, or HINT system, sifted through players' unconscious thoughts to detect any brilliant ideas just below the surface and alerted players to insights they were about to have. Entering a correct answer into the Handbook unlocked a page of content about that answer in the Mooncurser universe, providing GPS coordinates of the next clue site and possibly unlocking additional Handbook entries. Some Handbook entries contained information about bonuses that could be achieved through certain combinations of trade goods. The full content of the Handbook will eventually be posted online.

Another experiment involved bringing teams to a hotel on Friday afternoon and hosting a banquet and pre-game event there on Friday night, with teams staying overnight to start the Game itself early Saturday morning. Something similar had been done once before for the Jackpot Game, but we had something a little different in mind. The combination of the hotel rooms and pit stop meals meant that other than keeping their vans gassed up, teams would incur no incidental expenses from their moment of arrival to their moment of departure. This "Club Med" all-inclusive aspect of the experience was one of our big design goals, carrying through all the way to such details as creating a Galaxy Today newspaper for teams to enjoy with their Saturday morning breakfast.

Another design goal was to make the route more than just a bunch of clue drops. Whenever possible, we designed puzzles to fit our sites. Big Rock Garden, Binary Garden, Conway (skipped by most), Rock Paper Scissors, UW Bothell, Tolt River, Galaxy 12 theater, the corn maze, Seattle Center, Guild 45th theater, Sunset Bowl, downtown Ballard, Sturgis (skipped by everyone), Constellation Park, and the Museum of Glass sites all had puzzles designed specifically for those locations, usually incorporating features found there.

While I created a couple of puzzles (the signpost, which most teams skipped, the corn maze tiles, and blackjack) and collaborated on a few others (sculpture garden, Galaxy 12, early charades concept, bowling, some crossword-style clues at Elements (the split-up-each-team site), and Blinky), my primary responsibilities were for the trading game, the Handbook content, the pit stop activities, the Friday night post-banquet event, the newspaper, and the soundtrack audio CD.

The soundtrack CD was inspired by a similar disc we received during Shelby Logan's Run. I realized pretty early on that each track was connected with a site. That Game was entirely over the top, with helicopters, scuba diving, power boating, machine guns, and ATVs-- so when one of the last songs was Van Halen's Jump, I was convinced a tandem skydive was in our future and spent much of the game psyching myself up to do it and not let the team down (the song actually referred to a bungee jump). I'd really enjoyed using the CD to try to anticipate what was coming at each site, so I convinced the rest of GC to provide a similar CD for our Game. We told teams outright that it wasn't a puzzle, however, so I fear most teams may not have given it any attention.

More thoughts beyond the [More...] link.

What Went Well
The banquet food was surprisingly good. I expected mediocre hotel food, but everything was really quite tasty (mmm... escalloped potatoes...). Most people seemed to enjoy the friday night event, and all six challenges got accomplished almost within the expected time frame. Saturday breakfast delivery went without a hitch. I know of no hotel-related snafus. The Game started on time. People seemed to enjoy the games at the first pit stop. The corn maze, at night under the light of a full moon, had a spectacular vibe. We only needed to use footage of ourselves for 4 of the 30 charades in the charades puzzle-- the rest were culled from footage shot the night before during one of the friday night activities (and some of that footage was quite terrific). The movie theater location itself worked very well (thanks, Kim and Nate!) and really upped the ambiance of that puzzle. Everyone loved bowling. Blackjack held teams' interest for longer than intended. All teams got to the finish. All teams got fed at the pit stops. Staffing each site meant we got to interact with teams a lot, and were always on hand to swiftly handle problems as they arose. Using text messaging to spread the word about what was happening at each site worked very well (at least, until the SMS server went down for a bit before sunrise). The friday night stuff, pit stops, and trading game accomplished their goals of providing the means and motivation for team interaction-- I've heard from many players that they met and talked with lots of different people.

There are probably other things that went well, but these are the ones I observed directly.

What Went Wrong
The hotel kitchen had the wrong time for dinner on their schedule, and so the banquet started about an hour late. We misjudged the difficulty level of some friday night activities, resulting in some getting less play than anticipated thereby increasing the lines at others. People wound up waiting in longer lines, and we had insufficient staff to rectify the problem swiftly. Teams had trouble mapping the completed strips puzzle to the inlaid grid at the clue site (four beta teams had no problem here). Teams ignored the map of the artwork where the Codex puzzle was distributed, and therefore didn't understand that the "3" in the decoded message referred to the part of the artwork labeled "3" on the map. Once there, some teams still had trouble locating the proper book. There was some staff confusion about which teams were entitled to laminated decoder cards at the river. Helpers accidentally gave out only one answer card at the end of the river instead of two. Teams got to the corn maze about an hour late, so dinner was cold. Symbols in the maze were more difficult to locate than expected. We cut the puzzle immediately following the corn maze to try to get us back on schedule. The Tile Trouble puzzle at Seattle Center was too similar in feel to the corn maze search that immediately preceded it. Teams had more trouble assembling the Seattle Center cube than our beta teams had. Teams got skipped over the last 2, and in some cases 3, puzzles in the third leg (which we had planned for as the worst-case scenario, so it could be argued that this actually went well in that sense-- we knew in advance that we might skip those 3 locations, and so were prepared to do so-- but of course we'd have preferred teams got to see more puzzles). We didn't receive all the griddles we'd ordered, and so pancake production at breakfast ran at 50% expected capacity, thus prolonging the blackjack activity and delaying the start of the fourth leg. The custom electronic devices for the Blinky puzzle never got made, and so a software CD version was used instead. The CDs were burned on a Mac, and apparently PC laptops had trouble reading them. We were chased away from the Blinky site by the owner, who didn't like lots of vans in his parking lot. Some teams found the Blinky CD early, when the tape holding it to the underside of their seat failed. Even after applying a water mist, some teams couldn't read the message written on Rain-X on the inside of their van windows (contrary to some complaints, it was there, just hard to read unless you looked from the proper angle). Some teams got skipped over the final "meta" puzzle, which used information from the Handbook entries they'd been reading all day long. Some Handbooks crashed multiple times-- not sure how widespread the problem was, or if it was isolated to just a few devices. Fortunately recovery was fast and simple without any data loss. Staffing each site in person was physically demanding. Final standings have not yet been determined, due to timing imbalances caused by the Blinky problems.

I don't know how much of this stuff was noticable to players. Some might have been noticed, but made little or no impact. Others, like the Blinky problems, might have been major fiascos. I certainly hope all the things that went well outweigh whatever problems players may have encountered, but I know from past experience that even a single problem can sour someone's perception of the entire event.

From my own perspective, all the little (and some not-so-little) problems I saw made it difficult for me to celebrate at the end. Objectively I know that hiccups always happen in an event of this complexity, and I should be able to look past them to see everything that went right. But since I didn't play in the Game, I can't react to the players' experience because I don't know what it really was. I can only react to my experience as part of GC, trying to mitigate problems and keep things running smoothly. I play a lot of board games, and I always say that it doesn't matter to me whether I win or lose. What's important to me is that I played the best game I could play. I kick myself for the mistakes I make. In this case, at this moment, I feel that although we put on a good Game, we at GC could have done better. Some of the gripes players might have-- whether or not they liked the trading game, for example-- I accept as the cost of doing business. We were never going to please everyone in that regard. But we'd all run puzzle events before. As organizers we have, collectively, five MS puzzle hunts, 1 Game, and multiple MS Intern Puzzle Days under our belts. Many of our problems were foreseeable-- in some cases, even foreseen-- and should not have been allowed to happen. In the future I might discuss them at greater length, in the hopes that future organizers of similar events might learn from them and not repeat them.

Perhaps the most unexpected thing is that I think I enjoyed the planning of the event far more than the actual running of it. Other big events-- game shows, puzzle hunts, etc-- I've enjoyed running as much or more than planning. This time it was the other way around. Maybe that's because planning was a social group activity, and running was more solitary and stressful. I only saw bits and pieces of the actual event, and most of what I did see were the pit stops where I was stressed about keeping things going smoothly and getting our schedule back on track. I think I most enjoyed staffing the theater, where I didn't have to orchestrate as much and so my level of stress was lower.

I'm looking forward to honest feedback, both good and bad, from players about their experiences.

The Griffiths Collection

This weekend I visited San Francisco to take part in The Griffiths Collection. This was a full-length, 32-hour event that sent us driving and puzzling throughout the SF area. There was virtually no theme. Ostensibly we were searching for the stolen McGuffin diamond and journal, and a few pages of the journal were recovered at each clue site. The contents of the journal, however, had no bearing on the Game itself, and consequently most of our team never read those pages.

This was a schizophrenic Game. The first 4 or 5 clues were reasonable. The opening clue was a great concept-- translating text into Braille to form a picture-- marred a bit by its execution. There were some unintentional red herrings and nothing to prod us into snapping out of a key assumption-- that we were trying to translate from Braille. Clue #2 was a nicely conceived and executed polyominoes puzzle. Clue #3 was the first of many puzzles that had too many layers for their own good. In this case, the flavor text strongly implied an approach that wasted a lot of time for us. Clue #4, a CD of TV show themes, was terrific. We completely missed the ciphertext engraved on the shiny side of the disc, and so wasted a lot of time trying to extract meaning from nothing, but when we finally found the ciphertext it was a classic V-8 moment.

Clue #5 is where the wheels started falling off the cart. Clue 5 through clue 10 took us 15 hours-- from around 4 in the afternoon to 7 in the morning. We (and most, if not all, other teams as I understand it) were skipped over one of them involving a stereogram and a mirror, so that's an average of 3 hours per clue, which I think most Game players would agree is far too long. And lest you think our team was performing below average, even the best team took 14 hours on these clues (depending on how you interpret the results we finished 4th or 5th out of 18 teams). GC (that's "game control", Game-speak for the folks running the event) clearly underestimated the difficulty of their clues. But each of these monsters had multiple layers to unravel, including unnecessary Caesar shifts and a superfluous Vignere cipher, often with no internal clues to the necessary steps.

The fact that these puzzles made it into the event suggests that this was the first time the organizers were acting as GC. Underestimating difficulty is a classic rookie mistake, but these puzzles just screamed for simplification. With luck, the lessons learned from this event will help them make their next effort run more smoothly.

After sunrise, it was a like a completely different Game. Suddenly the puzzles were less convoluted, more approachable, and more fun. A couple of them were terrific group puzzles, and solving them was a wonderful collaborative experience. If the whole Game had been like the last 10 hours, it would have been a fantastic weekend. But to compound the problem, GC didn't manage our time very well. Instead of calling teams in within the pre-announced window, they let teams stay in the field longer. Understandable, since the Game was running long and they wanted as many people as possible to get to try the stuff they'd spent so long putting together. But we had a plane to catch, and as a result we skipped a clue near the end and still didn't have time to attend the wrap-up and chat with GC and other players-- often one of the highlights of the experience. So we finished with an anticlimax.

Our hope is that we'll learn from this experience and that all the mistakes we'll make in our own event next month will be completely different. We believe we know what our biggest problems are likely to be, and hope that our mitigation strategies will be effective. We're trying things that, to our knowledge, have never been done in a Game before. In and of themselves, these decisions may be controversial and some players may not like them. But that's the risk you take when you try something different.

Shinteki: Decathlon Recap

Now that the final run of the event is over, I can talk in more detail about Shinteki: Decathlon. As I said a few days ago, it was generally a great event with none of the hitches we experienced in Shinteki: Untamed. Our team finished a very strong second, and if it hadn't been for a complete shutdown of higher brain function on clue #6, we might even have tied for first. The clues, while often not well-themed, were nevertheless solidly constructed and fun. None of them had us scratching our heads for long. When they wouldn't give up their secrets right away, there were nevertheless enough bread crumbs to follow to find the correct path. This resulted in a great head-scratching:fun ratio for us.

A big change in this event was that it was not a race. Aside from having 12 hours to solve the 10 clues, time was irrelevant. And I found that a big disappointing. There's nothing like the adrenaline rush of arriving at a clue site after another team, but leaving before them. That euphoric moment of passing other teams and moving up in the rankings was definitely missed. But then, we were at the front of the pack. Teams farther behind might have had a different viewpoint.

One reason for the change was the new hint system, in which teams were able to purchase hints through their Palm. Each clue had a maximum value of 100 points. Each clue a team asked for reduced that value. This system worked remarkably well, and special kudos to the Shinteki team for thinking the system through. Early hints often became free after a certain amount of time, and the Palm recognized partial answers so that, once entered into the device, clues which were no longer useful to the team also became free. This greatly reduced the likelihood of purchasing a hint that told you something you already knew. Whether or not to take a hint became a factor of a team's level of frustration and overall time remaining in the event. A nice balance for an event intended to be more recreational than competitive. In a full 30 hour Game, I'd miss the rush of catching up to / passing other teams even more.

Here are some of the clue highlights:

After entering a code of FERNANDO at Ikea, we received a deck of cards with subtle markings on none, one, or both of the indices. After carefully recording the data in two Excel columns, it looked like Braille-- but that quickly failed to pan out. Ikea (Swedish) and Fernando suggested ABBA, which led me to Morse code (a-b-b-a as dot-dash-dash-dot). Treating cards with no markings as spaces, one marking as a dot, and two markings as a dash, the deck spelled out a message: PULL OUT FACE CARDS. Doing so and re-reading the remaining deck gave us PULL OUT RED CARDS. Reading the remainder again gave us ORDER ME. Putting the black cards in traditional deck order (clubs A-K, spades A-K) gave us the answer, MASTERY. I thought this was a terrific puzzle, with great recursion and elegance. And if we hadn't pulled the classic teen horror movie blunder of splitting up in the Ikea, and if the person who found the deck of cards hadn't left his Talkabout in the van, we'd have been out of there fifteen minutes earlier.

The next clue was a nice word search puzzle, but to get it we had to use a pedal boat to reach a raft floating in a lake. Even rotating our foursome so that each of us pedaled only halfway in each direction, it was exhausting. The customized Wheaties box we received, with our team photo on the cover, was a nice touch-- but I was a bit disappointed that it didn't factor into the puzzle as anything other than a delivery vehicle for the word search contained inside.

A set of four speakers at our next stop played a synchronized recording of someone shouting "Ready!" "Set!" and "Go!" at regular intervals, each speaker with a different series. To envision this, imagine a device that reads a Scantron sheet of rows of 3 circles each, and on each row 0-3 of the circles are filled in. Now imagine the device examines a new circle every half second, and calls out the appropriate word (READY, SET, or GO) if that circle is filled in. Now imagine four of these devices running simultaneously, each with a different Scantron sheet as input. That was the puzzle (if you replace the Scantron devices with iPods and speakers taped to trees out of reach). The member of our team with the best rhythm transcribed all the outputs. Another teammate looked at the data and noted that everything came in groups of five rows. I then looked at it and saw that they were 3x5 grids, and said at that size they really wanted to be letters. Almost immediately we began putting all the READY marks from all four outputs together, then the SETS, then the GOs. Whaddyaknow-- letters! We were out of there in record time, over an hour ahead of the average.

The next clue took the form of a Mad Libs book delivered at the top of a steep hill climb. Before we were an eighth of the way back down, one of our team members realized that the correct answer for each word in the Mad Libs was an anagram of the words filling the matching blanks in the other Mad Libs. We had our list by the time we reached our van and finished the solve in minutes.

Then... disaster. The next clue wasn't a puzzle, it was a challenge. 10 straws. 50 cm of tape. A model chasm. Build a bridge to span the chasm and support a 1kg brick for ten seconds. A van full of software engineers completely overengineered the hardware problem. Knowing that triangles are one of the strongest shapes, we set about painstaking creating a train-trestle bridge of triangles. We should have realized there were simpler solutions, especially when the leading team-- only five minutes ahead of us-- dashed off successfully while we were still drawing up plans. But we blundered blindly ahead. Ninety minutes later another team finally showed up and we still weren't done. And then our tape ran out, before we could shore up a key join. Our gorgeous bridge collapsed. We bought another 2 straws and 10 cm of tape for ten points. Then another. And another. Meanwhile, other teams had come and gone with vastly simpler designs, some just barely eking out 10 seconds before giving way. Incredibly, one team even succeeded with the most brain-dead design possible: make two long straws and tape them together. Ten point five seconds. We finally achieved success two and a half hours after arriving, well past the "we're not having fun anymore" stage. This wasn't the kind of activity I enjoy, and it's soooooo cliched. I was enormously disappointed when we found out what the challenge was. But our failure was clearly our own doing. Other teams succeeded in a tenth of the time we took, and we should have had the sense to realize our approach was silly. But dang, our bridge looked nice.

A later clue had us assemble a jigsaw puzzle (always a good team activity to which everyone can contribute) of a bunch of overlapping circles of various colors and sizes. Counting the number of circles in each color, arranging by the rainbow, and coverting to letters gave us WIDTH J. The device told us this was good progress. So we switched to the harder thing-- counting circles by width, and arranging from smallest to largest. This gave us OVERBLUE. Sure enough, there were precisely eight blue circles, one of each size. Counting all the circles on top of those blue circles gave us the final answer. Another very nice puzzle, especially the embedding of the WIDTH hint in the color data.

The final puzzle I want to comment on is the audio identification puzzle. Just as in Untamed, we got a CD full of music. This time, they gave it to us in an area where it was easy to find a wireless internet connection. One of the songs was impossible to ID, however, so we bought it for 5 points. This was the penultimate puzzle, and after the bridge debacle we were concerned about having enough time to go all the clues. Had not flubbed the bridge, we'd never have bought this hint and would have solved the puzzle without it, but we all decided it was more important to reach the final clue with enough time to solve it. But that's besides the point, which is that audio identification puzzles like this are not fun!. Or more precisely, they're great fun when you know the songs, but are utterly not fun when you don't. If you don't know the songs all you can do is try to search for the lyrics. And that's simply not entertaining. So if you're going to do an audio identification puzzle, the content being identified should be stuff 95% of the American populace-- or at least, the target demographic of Game players-- can be expected to recognize. Musical tastes vary so widely that identifying popular songs is not a slam dunk. We've had a pop music ID puzzle in the past two Shinteki events. Please, retire the form and find something else-- perhaps a clue where listening carefully to the music is more important than identifying the songs, artists, or albums.

Shinteki: Decathlon

This past weekend I went to Palo Alto and participated in Decathlon, the latest 12-hour Game event run by the good folks at Shinteki. Our team finished a very strong second, and had a lot of fun doing so. The game will be run again this weekend, so I can't discuss any details yet lest I spoil things for other players. I'll post a full recap next week. I can say that the clues were generally very well designed, and that most of the complaints I had about Shinteki: Untamed were fixed this time out. Teams playing in Decathlon this weekend should have a great time.

Puzzle Hunt 8

This weekend was Puzzle Hunt 8 at Microsoft-- a weekend of puzzle-solving debauchery shared by 57 teams of 12 people each. The event started at 10 AM Saturday and ended at 5:30 PM Sunday.

A couple of years ago my team ran Puzzle Hunt 6 which was, by most accounts, the best hunt ever. This one was better.

My hopes were high going into the hunt, since Games Magazine contributors Mike Selinker and Mark Gottlieb were both on the organizing team. I'd played in a 3-hour hunt they created for a National Puzzlers League convention a few years ago, and it was top-notch. This hunt ran flawlessly. We required no hints for any puzzles. None. And I don't mean just hints released to the general population-- we didn't get any kind of assistance from the organizers at any time during the hunt. No verification of partial answers, no reinforcement that we were on the right track, nothing. It wasn't necessary. The puzzles were sensible, well-designed and executed, and required no superhuman leaps of intuition to solve.

The first round was particularly elegant, consisting entirely of a Las Vegas travel brochure that, at first glance, didn't appear to contain any puzzles at all. In fact it contained seven, plus the most brilliant meta-puzzle I've ever seen. Each puzzle resolved to an instruction. Following all seven instructions resulted in cutting out parts of the brochure, folding them into rectangular tubes, cutting holes in them, and shining a light on them to cause them to cast a shadow in the shape of a 5-digit phone extension. A thing of beauty.

My team won hunts 2 and 5, ran hunts 3 and 6, and lost hunts 4 and 7. Given that pattern, we "expected" to win this hunt. We got the first solve a few minutes into the hunt, securing a lead we never lost. We were on fire, solving almost half the puzzles before any other team and ultimately finishing the hunt at 5 AM, 19 hours in. The next team finished three hours later. On one level it was a bit disappointing to finish so early (we WAY overbought on the food front, expecting another twelve hours of puzzling), but they say it's always better to leave 'em wanting more. The hunt never wore out its welcome, and we finished exhilirated and satisfied.

I pity the team that has to follow this act. Because it's us.

Shinteki: Untamed

Last weekend I went to San Francisco to participate in Shinteki: Untamed, a 12-hour Game event. This was explicitly for teams for 4 players, and I joined three of my fellow players from Justice Unlimited so that Team Briny Deep could set sail once more.

I'd heard nothing but raves about the organizers' two previous efforts, the full-length Jackpot and the beginner-level Shinteki: Aquarius, so my expections for this event were high. Unfortunately they were not met. I enjoyed the experience, but there were a number of very easy ways its flaws could have been eliminated and the experience improved dramatically.

We-- and thirty other teams of four-- gave up a big part of a weekend to participate. Fundamental to these kinds of events is trusting the organizers to be fair and keep player entertainment as their top priority. We put ourselves in their hands. The expectation was that Big Fun would ensue.

The opening was great-- a team scramble within the San Francisco Zoo to match clues to animals and locate the letters affixed to their exhibits. We got to see sleeping elephants and hyperactive birds and unflappable chimps and many other forms of wildlife, if we were willing to pause in our frenzy and appreciate them.

Things started going downhill from there. The biggest problem with the event was that it required too much Googling. The next clue had only a moderate amount-- we ultimately needed to identify eight professional athletes from their team and uniform number. With no Wi-Fi access at the golf course where this was needed, we had to rely on friends and family we had standing by at home. We'd been pre-warned to have phone support lined up, and it was simple to define the search terms over the phone and get the desired results without putting those helpers to too much trouble. One could argue, however, that it was completely unnecessary.

At the start of the event each team was issued a Palm device. Upon receiving each clue, teams entered an ID string into the Palm to start the clock. After a predetermined amount of time (which could be different for each puzzle), an alarm would sound and a new hint would become available on the device. When teams solved a puzzle, they entered the solution into the Palm and were told where to go next. The device worked flawlessly-- the first device in three Games I've played in to do so. And it would have been trivial to add to the device the ability to enter search terms and get pre-canned responses. So entering "Chargers 12" would give the name of player 12 on the Chargers, if that was the right player to ask for. Any other Chargers query would get no response or, if they wanted to be thorough, the entire roster of the team could have been available. Presto-- no Google, and teams get the satisfaction of solving the entire puzzle themselves without having to call out to friends for help.

A later puzzle couldn't be de-Googled so easily. Unless you're at a music store, identifying sixteen audio tracks and matching them to album covers will entail quite a bit of internet search. Distributing that CD atop a mountain overlooking the Pacific coast was not a terribly bright idea. While the organizers verified that there was ample cell coverage at the location, they completely missed the point. The four of us in the van are the players. Our friends and family at home should be used only for emergencies. Being tied up on the phone and the computer for an hour is not only above and beyond the call of duty for them, but a real drag for us. It is Not Fun. Forcing friends to gather large quantities of data for us is nothing close to fun. A puzzle where that gathering of data represents 95% of the solving time and no real creative thought is required is Not Fun. Planting a red herring message within the data-- intended to steer teams away from useless information but worded poorly enough to itself seem potentially meaningful-- was not smart design. This puzzle should simply have been axed.

But the most frustrating design gaffe had nothing to do with Google. The puzzle was a hard-boiled detective story with some very odd turns of phrase. My first thought upon seeing it was that there were cryptic crossword-style clues embedded in the text. But the content of the story also alluded to Gilligan's Island, with a margin annotation calling attention to that reference as well as two other parts of the story. That strongly suggested some kind of TV theme to the puzzle, but it was a complete red herring-- one that had us sidetracked for a long time. The real puzzle was actually quite clever-- the string "Shinteki Untamed" was anagrammed in multiple different ways and embedded throughout the text, and counting words between these markers yielded values in the 1-26 range. And while I came close to the correct approach right out of the gate, there were so many red herrings in the puzzle that it took a few hints before we got there. And the hints themselves were vague enough to send us on entirely new wild goose chases. The first one, "Hint: unmake edits" seemed to reinforce my cryptic theory, since "unmake" is a cryptic signal for anagramming. Anagramming EDITS yields, among other things, DIETS-- and the text contained an explicit reference to Atkins. "Aha!," we thought, "we're looking for names of diets in the text." Argh!

This puzzle would have been brilliant if only it had contained some internal clues to steer solvers to the right path. Perhaps all it needed was a title of "Shinteki Untamed" or an anagram thereof, and the elimination of that dreadful Gilligan's Island blind alley. As it stood, it was highly improbable that any team would have the necessary "aha" to solve the puzzle unhinted. And a puzzle that a) requires hints to be solvable, and b) so ofuscates the correct path with false trails that it's impossible to zero in on the right approach has significant design problems.

The last puzzle our team saw was essentially a crossword involving a mix of common and obscure facts. All players were required to wear nametags in plain sight throughout the event, and each tag contained a "secret word" assigned to that team. During the game, players were constantly trying to spy and record other teams' secret words. These words were answers to many of the clues in this puzzle. Since teams were spread out from the get-go, nobody was able to collect them all-- so even the best spies still wound up with obscure facts that needed researching. Which meant more Googling over the phone. Let me reiterate this for posterity and for the benefit of future puzzlemakers:

Googling is not fun. Calling a friend and having them Google for you is even less so.

If your puzzle in a mobile event relies on internet search to be solvable, it needs to be redesigned. Find a better way. Limited Googling is more acceptable in a stationary puzzle event where players have high-speed internet access, but even there it should be used sparingly.

Some might consider the above to be nitpicky. On a charitable day I might agree, but I spoke to some teams for whom this was their first Game-like event who were very turned off by the same issues I mentioned above, and are disinclined to participate in future events as a result. My feelings are less extreme. Except. There was one design decision the organizers made which, independent of any other problems, is sufficiently and egregiously wrong-headed as to make me seriously reconsider whether I'd ever be willing to trust them with my time and money again.

At some point during the day, the top two teams were given a puzzle that was completely unsolvable-- not through an error in design, but by intent. It was specifically designed to offer many intriguing avenues of approach, but to have no actual solution. After some time bashing their heads against this wall, these teams got the hint that "You're overthinking it." After over an hour of this for the top team and forty minutes for the team close behind, the Palm informed them that their time on that puzzle was up and simply directed them to the next location. As a result of this boondoggle, one of those teams missed a clue they otherwise would have seen because they ran out of time. The organizers called this "The Demoralizer." Having claimed that Shinteki's challenges were mental, physical, and psychological, this was their psychological challenge-- to give the leading teams the taste of frustration that trailing teams often experience.

Ok, class-- all those who think this is Fun, raise your hands.

Torturing the players who paid you money for a challenging, yet ultimately fair puzzle experience is so fundamentally wrong-headed I don't even know where to begin. The key here is that players did in fact pay an entry fee to participate. What might-- and I strongly emphasize the word might-- be acceptable in an event run at no cost is unforgivable when you're charging me for the privilege. The players aren't there to entertain the organizers-- they're there to be entertained. Being intentionally demoralized does not qualify.

There were some very good puzzles in Untamed. I loved the nighttime sound-sensitive Braille puzzle that had multiple teams gathered atop a hilly clearing screaming into the night at the top of their lungs (apparently the neighbors loved it quite a bit less, as the police shut the place down not long after our team left). Another puzzle involving threading a series of characters along a Celtic braid pattern was very well clued internally and was therefore satisfying to solve. The locations we passed through were terrific-- we had some great views of the Golden Gate Bridge, we walked among giant redwoods in Muir Woods, and enjoyed passages along the coastline. Our team was dressed as pirates and our van decked out in full buccaneer regalia, and the hearty "Yarrrrr!"s we dished and received from passers-by (players and bystanders alike) were tremendous fun.

People whose opinions I respect loved Shinteki: Aquarius. But after Untamed, I'm far less likely to shlep down to the Bay area for another 12-hour event. It's not just the distrust I've developed for the organizers. Around 6 or 7 PM, our team finally started firing on all thrusters. When time ran out at 10, we were just getting revved up. We could have gone a few more hours easily, and were eager to do so. I think I prefer the more epic 24-hour experience, even with the spanking that awaits around 5 or 6 AM.

I'd love to hear comments from other Shinteki players about their view of the event.

Justice Unlimited

Last weekend I went to San Francisco for the latest Bay area Game, Justice Unlimited. The SF Game community is much larger and more active than Seattle's, and frankly seems more fun. In Seattle all the team names are colors which, while faithful to the source material, aren't much fun. SF teams get more creative with both naming and costuming, with teams like Orange Crush (clad in bright orange), Blinded By Science (wearing sunglasses and lab coats), and the Scooby Doobies (their van is, of course, a replica of the Mystery Machine). The Games themselves are also more publicized, without the cloak-and-dagger don't-talk-about-Fight-Club attitude prevalent in Seattle. All of these things are, in my opinion, Good Things which help make the experience more fun. We are team Briny Deep, the puzzle pirates, complete with pirate shirts, Jolly Roger head wraps, an anchor to toss overboard whenever we pulled up to a clue site, and a hearty "Yarrrrrr!" for all we meet.

At the start of the event, each team was given a groovy custom-made device that strapped to your forearm. It contained a two-line dot matrix display, a knob/button, and an infrared sensor. The device came preloaded with potentially useful information (Morse code, Braille, Semaphore, tide data, ASCII codes, etc), thereby ensuring that each team had any decoding data they might need. Each clue also contained a code which, when input into the device, yielded supplemental information (mostly just for flavor, unfortunately, but very occasionally containing a hint). Puzzle answers were intended to be input into the device as well to produce the next clue location, but sadly that aspect of the device didn't work and teams had to phone in our answers instead.

The event ran from 10 AM Saturday though about 4 PM Sunday and took us all around the Bay area starting at the SF Municipal Pier a stone's throw from Ghiradelli Square. Location-wise, the next stop was one of the coolest: the Bay Model, an enormous scale model of the Bay area used to simulate tidal and other natural effects. Teams had to identify parts of the map by matching line drawings of topography to the right part of the model-- a fun scavenger hunt activity that got all teams moving amongst each other.

Other highlights included the Bat Blinker, a custom-made electronic persistence of vision gizmo supplying an elegantly-designed AHA puzzle; searching the beach at night for a hidden clue, flashlight beams cutting through the mist to create an X-Files-like vibe; a wonderfully collaborative solve on an audio puzzle where high and low sound effects overlaying dialogue from the Daredevil movie mapped to-- what else?-- Braille; a clever puzzle in which multicolored plastic strips had to be threaded through like-colored metal connectors to create letters; and a very nice balancing puzzle involving coins and a ruler.

With very few exceptions (the beach, Bay Model, a mountainside, a park, and a rooftop), however, the locations were largely unremarkable and underutilized. We rarely had to really search for a clue, and consequently our out-of-van exposure at any given location was minimal. I'd have preferred the clues to either be better integrated with the environment or better concealed, so that we interacted with each location instead of just hopping out to grab the clue and then piling back into the van.

A couple of clues really spanked us hard. One, a rainbow jigsaw that had to be cut out and assembled, was just poorly designed. When the pieces were properly assembled, the resulting shape had the same number of rows as the letter grid we were to use to extract an answer, and the shape had irregular square gaps that just screamed to be utilized. Was the shape a physical mask to be overlain on the letter grid? Were the gaps some kind of mathematical encoding to determine which letter to extract from each row? No. They were just noise. Each piece of the puzzle was a different color, and each row of the grid was a color. If the color sequence was red, yellow, purple, orange, purple, green, blue, red, we were supposed to take the first letter from the red row, the second letter from the yellow row, the third from purple, and so forth. That's it. We wasted a lot of time analyzing the gaps which had absolutely no purpose in the puzzle. That's just poor puzzle design-- in this kind of event, where you're working without instructions, you don't throw false leads in front of people that can rathole them for an hour. In fact, you remove as much of that noise as possible to help funnel solvers to the correct approach. This puzzle aggravated us and left us disgruntled when we finally hit on the solution.

We also got spanked by a puzzle incorporating Heroclix figurines. The puzzle itself might have been fine a few hours earlier, but it hit us around 4 AM when our energy was at its lowest ebb. Four of our six players were asleep, and the other two were barely staying conscious. It was ugly-- especially coming off a semaphore puzzle that also spanked us because it appeared to be a completely different sort of puzzle, and consequently we overlooked an important bit of information.

There were some nice opportunities to draw on useless superhero knowledge gained through years of comic collecting in my adolescence-- a crossword puzzle where the clues were superheroes and the answers were their secret identities, and a quiz on superhero origins. I pray no Game ever has a Pokemon or gangsta rap theme.

Team Snout, Game Control for Justice Unlimited, generally did a terrific job. Aside from the DRUID problems (the whole infrared sensor thing never really worked right, either), the puzzles appeared to be error-free and Snout staff were unerringly upbeat and helpful when reached on the phone-- impressive, considering 25 teams were playing the Game (for comparison, only 9 teams played in Shelby Logan's Run in Vegas, the last Seattle-based game). Aside from reporting our answers, we generally called under two circumstances: when we were about to embark on a lengthy decoding process and wanted to make sure it wasn't going to be a blind alley; and when we were stumped and collectively agreed we weren't having fun anymore. Game Control was always happy to unblock us and keep us energized, which was exactly the right attitude.

I look forward to playing in the next Bay area game. If you'd like to join, Briny Deep may have a couple of spots available...

PH6: Time Corps

Needed some time to decompress after the MS Puzzle Hunt before I could go over it all again. For the curious, here's the set-up.

The hunt was announced via a letter to players, congratulating them on being accepted to the Time Corps-- a government research think tank established to exploit new Chronoscope technology that can peer backward through time. Time Corps is gathering the brightest historians, sociologists, and researchers to help analyze the mysteries of the past and create a better future.

Upon arrival the hunt, players are greeted by Chronoscope inventor and Time Corps founder Dr. Jacob Goode and given an orientation spiel about the Time Corps' history and mission, including a goofy call-and-response pledge (which, gratifyingly, the entire room of ~400 people played right along with). During that orientation, dissheveled and disgruntled researcher Dr. Richard Naste burst into the room demanding to know why he'd been locked out of his lab. The ensuing argument between Goode and Naste revealed that the Time Corps advisory board felt that Naste's research was dangerous and unethical, and that he'd been shut down and his credentials revoked. Naste, angry and egotistical, stormed out with a promise that this wasn't over. Teams were given their first set of puzzles-- "extracts" from the Chronoscope, each representing a particular period of time and named after historic events-- and dispersed to their separate conference rooms to begin working on them.

Five hours later, players were interrupted when their computers were taken over by a video message from Dr. Goode, informing them that Time Corps was in a state of security alert. Dangerous levels of chronoton emissions had been detected within the facility. All researchers were advised to don time-protective gear and proceed to the orientation area immediately for a security update. Players were warned this was not a drill and urged not to panic. (We later heard from a number of teams who were just stunned at how their machines had been taken over by our video, and who just sat there when it finished asking each other "How did they DO that?")

When teams arrived at the orientation area, they were met by Dr. Naste-- now immaculately attired in a fine suit, and attended by a number of other men in suits. The Time Corps logo behind him was subtly changed-- the S was missing. Naste proceeded to give a status report on TimeCorp's progress in various outrageous schemes to get wealthy-- extracting all diamonds from South African diamond mines centuries before they'd be discovered; inventing the lightbulb and phonograph, selling the patents to Edison, inventing improved versions, driving Edison out of business, and reacquiring the patents for a song; and so forth. Naste lead the group in a chant of the corporate motto, "Time is Money," before giving them their next set of puzzles, another set of historic events. (The script for this was very funny, and the audience obligingly shouted out two questions we wanted asked without our having to resort to the shills we'd planted: Where's Dr. Goode? and What happened to Time Corps?)

When players returned to their conference rooms, they discovered that the Puzzle Hunt web site, which had previously been the Time Corps home page (with info about the campus, advisory board, and so on) had changed. It was now the TimeCorp home page, with completely different personnel and history described therein. This included a password-protected "supply center" link which players couldn't access.

A few hours later another video interrupted players. Dr. Goode, apparently on the run and hiding somewhere in the TimeCorp facility, explains what happened. He didn't really invent the Chronoscope-- he found it frozen in ice in Siberia. In the process of analyzing it and learning how to use it, they theorized that it was part of a time machine that had exploded and become fragmented throughout time. Dr. Naste became obsessed with reconstucting the whole device. That's what the argument at the orientation was about. Naste had no concern about safety or ethics, and the board shut him down. But too late-- he'd obviously succeeded, activated his device, traveled through time and changed history so that Time Corps never existed. But his changes had more far-reaching consequences, and altered many events throughout history. Worse, these ripples in the timestream were becoming waves, which were becoming tsunamis, which would eventually shread the fabric of space-time with devastating results. Goode was sending this message to the new Time Corps employees because their labs, like his own, had the heaviest temporal shielding and he hoped they'd remember the old time stream as he did. Goode needed the players to figure out a way to reverse what Naste had done.

Goode also gave players a new set of puzzles. There were a total of 17 puzzles in the first two groups. This final group also had 17 puzzles-- one for each event in the first 17. But although their titles were the same, and in many cases they even looked the same as the previous puzzles, they were in fact completely different. As Goode said, history had changed-- all of these historical events had been altered.

Later, the players were emailed a status report which had been filled out for them by Dr. Goode. This was the meta puzzle. Solving it using answers from the other puzzles guided players to an Xbox kiosk on the MS campus, which we'd modified. To access the kiosk, players had to enter a series of 10 codes on the Xbox (which the meta puzzle revealed to them). Once they did, the Xbox gave them a game of Tetris to play. After clearing 10 lines, the right side of the screen became another video of Dr. Goode (if the players stopped playing at this or any point, or lost the game, they got kicked back out and had to reenter to code). Goode told players that Naste had managed to track down all the pieces of the time machine that had been scattered through time, and players had to do the same-- assemble their own machine with spares from the TimeCorp supply lab, go back in time to yesterday, and stop Naste from ever activating his own time machine. To find out what parts they needed, they'd have to do what Naste did and find them scattered through time. One time machine part was visible somewhere in each time period they'd studied, either in the original or alternate timeline. Goode gave them a password to the TimeCorp supply center web page which had, until now, been unavailable to them. There, players found an inventory list of 40 parts (only 17 of which were "real"), and a "Time Machine Schematic" puzzle. Once players searched through the puzzles and found the 17 time machine parts and entered them into the schematic, they got a 17-digit number which broke down into a building, room number, door lock combination, and chamber number. Dr. Goode from the original timeline met players at that location, having been sent a message from his future self. In that chamber was the time machine, and behind a glass wall was Dr. Naste frantically preparing to activate it. Players removed a key part from the machine, causing the machine to explode (sending the pieces scattering throughout time, thus closing the temporal causality loop at the heart of the hunt) and winning the game.

Phew.

Puzzle Hunt

I've been quiet for the past few days because I've been busy running a puzzle hunt at Microsoft for about 400 people. The event began Saturday morning at 10AM and concluded at 5PM on Sunday, for a total of 31 straight hours of puzzling (yes, most teams stayed through the night). There were 34 puzzles and 2 meta-puzzles, 3 filmed video segments, 3 live performances, nine organizers and 43 teams of up to 12 people each. Four teams "solved" the hunt, with about 5 more teams getting very, very close as time expired. The hunt was planned over a period of about eight months, although as usual with group activities much of the work actually got done near the end of that time period. I wrote six of the puzzles and both metas, and overall I'm pleased with the way things turned out. Players seemed to have a great time. I crashed at about 8PM last night and woke up at noon-- there's nothing quite so blissful as the long recuperative slumber after an all-nighter. Now I need to feed my body and catch up on various things that I've let slide while focusing on hunt prep.