Mystery Hunt 2013

There's a lot of chatter about this year's MIT Mystery Hunt, which was the longest one on record at 73 hours.  They usually begin on Friday morning and wrap up on Sunday afternoon after the coin has been found.  This year, the coin wasn't found until Monday afternoon, and even then only after aggressive hinting, wholesale distribution of puzzle answers, and reduction of endgame requirements.  This has led to much hand-wringing about the Hunt being broken and despondent posts from Hunt organizers who are taking the outcome and backlash very hard.

Man, I've been there.

If you've never run a puzzle event like this, you may not realize the toll it takes on the organizers.  People who put Hunts together pour their hearts and souls into it.  They give up not only all of their free time for the year leading up to the event, but they sacrifice other things to create more free time they can throw onto the Hunt pyre as well.  It takes an enormous toll on friendships, families, work-- every aspect of your life.

I absolutely understand the depth of sacrifice involved in stepping up to make a Hunt happen.  And that understanding makes it all the more painful when fundamental decisions at the core of the event seem predestined to undermine all that work by making players unhappy.  And it all comes back to one thing: make it fun.

As an event creator, you're not there to show how clever you are.  You're not there to outwit the players.  You're not, I'd argue, even there to make things fair.  You're there to make things fun for the hundreds of people who are trusting you with their leisure time (and in some cases vacation time, plane fares, hotel fees, etc).  When you sign up to run an event like this, you're agreeing to provide fun.

A problem is that fun is not absolute.  There's no Unified Fun Theory that satisfies everyone.  A physical obstacle course might be tremendous fun for some people and loathsome to others.  Some people like manipulating numbers, while others like manipulating words.  Some people like to stand hip-deep in running water holding a pole for hours at a time, others like aligning pixels to be just right.  Nobody's going to get everything right for everybody.  So you need to try to maximize the fun for the most people.  Which means you make some assumptions, put some stakes in the ground, and try to hew to those guiding principles as you create your event.

And if those principles turn out to be wrong... your event's in trouble.

This happened with Puzzle Hunt 123.  I still believe there's a viable model in having teams self-sort into competitive or recreational divisions.  The impact of that division was too stark in 123-- an all-or-nothing switch that made people feel like pulling it was crying uncle and dropping out of the competition.  We did that because if you can get unlimited hints-- even answers to puzzles-- how do you score things fairly?  How can you maintain a competition when anyone can decide to just get all the answers at will?  We prioritized fairness.  Our players prioritized competition.  The feeling of contending for something, of jockeying for position on the leaderboards, was an essential element of the fun for them.  Without it, they had no motivation.  Had we thrown fairness to the winds and just moved teams from one leaderboard to another, instead of dropping them from the leaderboard entirely, I think teams would have been perfectly fine with that.  So what if there would have been no way to tell the difference between a team that solved a puzzle on their own and one that got hinted all the way through it?  The teams' own pride at wanting to solve without hints would probably have provided all the regulation we needed.  Maybe we could have reduced the value of a puzzle a token amount for taking a hint.  There were any number of things we could have done.  But our assumption-- that people wanted to see puzzles and be unblocked more than they wanted to compete-- was wrong.  People wanted both.  And since we weren't aligned with our players, our event had trouble.

The Hunt this year was too hard.  That manifested in many ways.  There were too many puzzles to solve in the time allotted.  The puzzles got unlocked too quickly, by a failsafe timer rather than team achievement, so that teams were flooded with puzzles (making it hard to focus on what you had).  Many puzzles went one, two, or a few dozen steps too far, overstaying their welcome.

Here's an example of the latter case.  The idea of a fractal wordsearch is interesting.  Solving it through one or two levels is fun.  Solving it through three or four is pushing it.  Going WAY beyond that, beyond even the ability of most computers to solve the problem in the allotted time without superior programming skill, it going way beyond the fun.  There were probably a few people at the Hunt for whom writing an efficient fractal wordsearch solver was an enormously fun challenge.  But I think it's reasonable to assume that such a person won't be found on many teams, and therefore many teams won't find that puzzle fun.  Moreover, teams will probably expend quite a bit of work on the puzzle before getting to the point where they realize the need for a programming solution, and abandoning all of that work is a bitter pill.  Including such a puzzle in the hunt, therefore, sets your players up for unhappiness and un-fun.

Another puzzle consisted, in its entirety, of 263 MP3 files.  It's hard for me to imagine that listening to 263 MP3 files-- even at only a few seconds each-- sounds like fun to anyone.

The difficulty problem is foreseeable and solvable.  It usually requires a trusted editor (or editorial board) empowered to make decisions about whether or not something makes it to the final event, and a team that agrees to trust their judgement (or at least abide by their decisions).  That's a very difficult thing in a volunteer event.  People get disappointed or disgruntled.  Egos get bruised.  But if it's done in the best interest of the event, it's the right thing to do.  At the end of the day, if you're running the event to serve your own ego, you're there for the wrong reason.  It's not about you.  It's about your players.

I think the principle that wasn't aligned with players here was that the Hunt had to be longer and not end as early as it did last year (late Saturday night).  But I'm not sure that's true.  If the Hunt continues even after the coin is found, giving more teams the chance to see all the goodness, then you actually want an event that ends earlier for the top teams so that other teams-- slower, smaller teams-- still have a chance to finish.  Worrying overmuch about length means you're focused on a small percentage of your players.  There's no cap on team size.  If a team finishes too early for their taste, that's not the organizers' fault-- it's the team's problem.  If winning is most important to you, join a big team that races to the end.  If puzzling into Sunday is more important to you, join a smaller team.  The event can hardly be expected to gracefully scale up along with the sizes of the teams.  That just isn't tenable long-term.  Players have to take some responsibility themselves and self-organize into teams that align with their interests.  Don't join a team of 150 people and then complain that the event ended too soon.