Monday, January 23, 2017

Mystery Hunt 2017: The One Built For Us

Last weekend I participated with in Mystery Hunt, an annual giant word puzzle competition hosted by MIT on behalf of the team Left as an Exercise for the Reader.   Despite being much smaller (officially 35, unofficially who knows) than almost every major competitive team, we, for the first time ever, got to the end.  There are some spoilers ahead, but not all that many, given the large portion of Hunt I barely saw. There is also probably song lingo that will be utterly meaningless to you, if you are not familiar with this type of thing.

To give some background, Mystery Hunt started in 1981 as a small event, one person creating a page of puzzles for MIT students solve.  It grew slowly for a while; one member of our team last won Hunt in 1991 with a team of 6. As it grew though, it became an arms race where puzzle writers tried to make larger and more difficult hunts to counteract ever growing teams and advancing technology.  

I first hunted in 2010 and have done it every year since then except one.  Mystery Hunt was my first puzzle event of any kind, and it was generally impossible.  I’m not sure if I contributed anything of value that year, and mostly felt dumb.  I’ve gotten better since then, but not enough that I don’t find Mystery Hunts to be mainly overwhelmingly difficult, with occasional satisfying moments of making a minor impact.  The worst one was 2013 where the designers underestimated the strength of their play testers, and decided to make their puzzles even harder with puzzles like this one.  That Hunt lasted 75 hours and required lots of free answers and hints to end it at all.  2014 and 2015 had better balance, but 2016 was written by another large team, and was rough on smaller squads.  I have found I have generally enjoyed smaller events like BAPHL and DASH, where I’m hitting my head into a desk less often.
Setec, this year’s host, is small for a winning team clocking in at around ~50.   They have a few wins in the past, but for a while were determined not to win.  If they got close, they started drinking more.  Eventually they forgot how painful writing a Hunt is, sobered up, and won last year.  They brought with them a new philosophy in writing this Hunt -- not-impossible puzzles are even more fun than impossible puzzles.  They wanted to make it challenging for the top teams too, but wanted to make sure everyone had fun, and that more teams got to progress to major milestones.

After partaking in my annual tradition of getting on the MIT Campus, I saw a cute opening sketch of a role playing game themed Hunt.  The first puzzle I worked on was Pic of the Litter, because I could identify pictures of objects and put it in a fill-in grid.  It was pretty straightforward and a nice start to the the Hunt.  Then I started on Hexed Adventure, a text adventure game.  Let’s see: desert to my left, mountain to my right, hills in the distance…  hmm that sounds suspiciously familiar.  I then spent most of the next 8 hours with a partner roaming around a Catan board.  The puzzle wasn’t hard per se, but it had a lot of steps and a mountain of text as you progressed to the actual solvable puzzle part.  We got stuck a couple times, most frustratingly missing where to get WD40 from, but for the most part we progressed steadily.  The solving part was also a little more challenging than it could have been, since we forgot wood could be lumber, and ended up with something that looked like it should give us an answer but wasn’t.  Still it was an incredibly fun puzzle, and I’d recommend giving it a try if that sort of thing appeals to you.  There was another puzzle with a mock BGG page, so for a while I was working on two board game related puzzles at once.  It was great.  

We were also visited by an economist, transfixed as Bob Barker, and played some Price is Right until we identified the exact price of a cup of our color pencils.   I nibbled at a few more puzzles afters, but didn’t make much progress.  My final useful contribution was standing behind someone who got stuck.  He proceeded to tell us what he done thus far, tell us what he just did that didn’t work, realize that in fact it did work, corrected a minor error and solved the puzzle.   Nice to know hovering works.  I left for the evening while we were trying to figure out how to use semaphore on elements in a circular periodic table.*

*Spoiler-- you don’t

Saturday morning, I came in knowing Death and Mayhem had already found the coin, winning the Hunt, which was a good sign for us getting far.   I bounced around a bit, trying to get back into the swing of things, not all that successfully.   Someone mentioned there was a Fantasy Baseball puzzle, Special OPS, so I joined in on that.  It involved figuring out the real components of fake stats.  We tried it both manually and through a program someone built, but we never had any success.  I spent a lot of time trying to make the numbers work, but they wouldn’t.  I felt bad for striking out on a puzzle that should clearly be in my wheelhouse, but it turns almost no one successfully solved it.  There were less impossible puzzles than normal, but there still was some, and even reading the solution, left me utterly confused.

Feeling dejected, I picked up a children’s book that we had received entitled The Puzzle at the End of the Book, modeled after the Sesame Street Book, The Monster at the End of the Book.  It was a small book, so I figured that it shouldn’t take too long to solve.  Turned out there were a whole lot of puzzles hidden in there.  It was a really cool puzzle in that everyone who looked at it found different things.   If you haven’t done any puzzle competitions, but the idea sounds interesting to you, this is a really good puzzle to start with.

We celebrated our team’s birthday at noon.  It was was our 10th entry into the Hunt.  We had tiaras and cake and invited some teams over including our neighbors, IIF, and Setec.  It went well, we even had an appropriate amount of food unlike last year, when we had to call HQ with an emergency cookie problem.  The only misstep was we accidentally sent Palindrome a blank invitation, which they assumed was a puzzle and proceeded to try to solve.  Palindrome finishes the Hunt early though, so we were just giving them something else to do.

I also made my biggest meta contribution that day, determining that the Crafty Criminal meta involved actually playing Mastermind.  Before I had to time figure out how that could work though, someone else discovered the black and white letters, and an offsite supersolver submitted a correct answer.*  Still, I feel good about that one.

*We don’t have a ton of off-site people but we have one who you know will solve quickly when he enters the google doc.
Late Sunday morning, I got an urgent e-mail, stating we were three meta puzzles away from the end.   When I arrived, we were working on several needed puzzles for the Meta like Tricky Wicket and Hamiltonian Path.  I took up a beast of a logic puzzle, Mirror Mirror,  that revolved around figuring out where mirrors reflected light.  Solving was slow, since many options were unlikely but not impossible, making it difficult to draw conclusions.  Our progress hastened when we realized the numbers spelled words, but halted when we made another incorrect assumption later on.  Then we solved the associated Meta, so we moved on to more pertinent puzzles.  Still it was interesting, and I’d like to solve it some other time.

We solved another Meta, so we were down to one final Meta in the last few hours of the Hunt.  We got  lot of visitors from Setec, including one who hung around poker faced, as we were the only team left on the brink.  I think they were both heavily rooting for us and enjoying watching us squirm.   The last Meta, the Broken Bridge, became a large group solve as everyone worked on it or the four remaining unsolved feeding into it.  We were told we had enough answers to solve it though, so we bashed our heads together on many a pointless path.

Finally, someone realized that each word had a 4 letter word associated with it and those words could be chained together by changing one letter between each.   Most of the words fit together, but there were a couple like Hugo that were way different, and we were missing some answers.  I noticed that tape fit between two words we had, and someone else saw duct could also work as well as role.  Aha Roll of Duct tape!  Nope, not quite.  We pondered other options, when Tree Rings, one of the four remaining was solved with the answer Winter Olympics.  Winter Olympics = Luge - Huge - Hugo…  We entered in HugeDuctTapeRoll, and for the first time, entered the endgame.  It was about 4:00 pm Sunday, 51 hours of the Hunt completed, 2 remaining.

We proceeded with almost our entire team to confront Mystereo Cantos the big bad, cheered on by IIF.  We gathered around him in a circle, answering random questions with thematic answers, while he made fun of us.   We then had to do one last thing to defeat him, which we had a little trouble working it out, mostly because our participants in this quest were completely different than an earlier one, which contained a necessary step in the solution.  We eventually figured it out though, gave him a “grouphug” and turned over a giant hex which had instructions for our final run around attached.

Anyone who saw us, a large group of people, wandering around the MIT campus, with someone reading off D and D sounding instructions from a giant hex, would have thought we were insane.  Of course if they were at MIT, they are probably used to these things.  We followed the instructions  which compared MIT landmarks to a fantasy setting.  Eventually we came across the row of lights, where our leader found the coin at 5:53, 7 minutes before the Hunt ended. We may not have been, but we certainly had the most efficient at pacing ourselves for an enjoyable weekend.