MIT IAP Mystery Hunt Cambridge, MA, MLK Weekend, yearly

On the Friday before Martin Luther King weekend, at noon, hordes of eager puzzlers magically appear in the lobby of MIT's Building 7. So begins the MIT IAP Mystery Hunt, run during MIT's Independent Activities Period. At noon, the victor of the previous year's hunt introduces the hunt and theme to the tens of teams listening in the lobby and the hunt begins. The goal of all Mystery Hunts is to find a coin hidden on campus. (Here's a little hunt history, and a MIT Tech Review article about the hunt. Wikipedia has an article entitled MIT Mystery Hunt as well.)

Typically the puzzles are released in waves with many puzzles on each wave. If you finish the puzzles from a wave before the next wave is released, you get the next wave early. Eventually, the answers to all the puzzles feed into one giant (or several cascading) metapuzzle/s and a final puzzle that leads to the coin. It's not necessary to solve every puzzle; occasionally you can deduce the solution to a puzzle from the rest of the puzzles in a group by backsolving from the metapuzzle. The puzzles are usually fairly difficult and involve esoteric knowledge; web access is necessary for many of them, but most can be done without leaving one's seat. However, there's usually a few that require treks around campus and one or two that require leaving campus. With so many puzzles outstanding (there were over 80 puzzles in the 2002 hunt), teams are usually fairly large, at least 10 people and often many more, with varying degrees of dedication (and sleep). The hunt usually takes on the order of two days (that's night-and-days) to finish.

I'd always wanted to participate, but it's across the country and I needed a team. Fortunately, Kiran Kedlaya of Mystic Fish was a regular and offered to hook me up for the 2001 hunt (The Hunt of Horror). I was introduced to the team from the Third East dormitory (aka "Tetazoo"). Their team was called the "Vermicious Knids" and over the course of the weekend I met 30-someodd people on the team who wandered in and out of the math dept. lounge during the 40-someodd hours of the hunt.

2001, The Hunt of Horror

Our team had tons of energy, lots of people, and very little organization. I got to see a little of Cambridge as I asked to go on the puzzles that led around campus: to a CVS supermarket for Blood of the Creatures and around campus for the very excellent The Roads Through Hell and Stumpophobia. The puzzle with the most participation among team members was We Have Enough Twists, Thank You. We placed second overall, a few hours behind Setec Astronomy, who found the coin at 2:16 am on Sunday.

After the hunt I sent a friend the following listing my favorite puzzles:

Some of my favorites: Gojira, The Thing, To A Monster in round 1. Disturbing Universe, Eeek! was clever, Hex: very good, Twisting Passages. We Have Enough Twists was neat but took FOREVER. "This Puzzle Needs No Title" was a great metapuzzle. Agoraphobia. Xenophobia was a BITCH. Dialoguophobia was very cool. Numerology was clever.

(2001 hunt homepage)

2002, Monopoly

Setec Astronomy ran a great mystery hunt in 2002. (Great webpage too.) Largest hunt ever - 83 puzzles. There were 8 waves of puzzles, each culminating in a hotel puzzle, and a final puzzle when all 8 hotels were solved.

The Third East team was called "Excuse Me, Is Thouis There" and contained many of the same folks from the 2001 team. We secured a larger conference room this year and went to work. Each puzzle's answer implied a Monopoly square (first one we got, "carpet", led to "Oriental Avenue" for instance).

When you got a monopoly of properties, you were given 5 more puzzles: 4 "houses" and a "hotel" metapuzzle. The answers to the house puzzles + the answers to the property puzzles for any color were used to solve the hotel puzzle. The name of each hotel metapuzzle fit into the same pattern as all the subpuzzles for that hotel. The endgame started when your team got all 8 hotels.

When that happened, you got a tax form metapuzzle. That used all the answers to the non-colored squares (water works, chance, etc.) and told us the starting location for the runaround.

Now, each puzzle was associated with two dice. And it turned out that you could map the 40 board puzzles and their dice rolls to a game of Monopoly in which every square is visited once. (My notes from our reconstruction.) It involved figuring out what cards were drawn (i.e. "Advance to Nearest Utility") at Chance and Community Chest squares. This was used in the final metapuzzle.

But first, before we were allowed to get the final metapuzzle, we were told we were in jail. It turned out that the Community Chest square in which we had to pick up that card during our Monopoly game to make the rolls work out right had a secret message in it to "COME TO HQ AND GET A CARD". So we had to figure that out and then go get the card and move on.

Final puzzle was a 40-spoked wheel with an instruction per spoke. If you picked the right starting location (decodable using a number per spoke from your Monopoly game traversal) and followed the Monopoly game rolls to move around, you got the right order for the instructions, which led to the coin. We had sent two teams to start the walk given what we guessed were two likely correct starting points, and were decoding the starting location, when the Acme team found the coin. We were 25 minutes back (and HAD guessed the right starting location).

The coin was found at 5:29 am on Sunday by ACME. (GC describes the 2002 endgame.) I set two personal records: two all-nighters in a row for the first time ever, and a sleep record afterwards, 15 hours sleep the next day.

My favorite puzzles from the 2002 hunt (again, from a list in an email I sent to friends after the hunt):

  • 2/3-1: Some extremely weird juxtaposition of 5 board games at the same time, swapping pieces between games. Clue, Scrabble, chess, etc.
  • 2/2-3: Programs written in bizarre computer languages (REALLY weird languages) .
  • 3/1-1: "Anguished English" as in GNASHING ALGAE OKRA FAKE == "National Geographic".
  • 3/3-1: Puzzle with the Get Out of Jail Free card (+ its own sol'n)
  • 6/3-5: Any reference to "Red Meat" is just fine with me.
  • 6/6-6: Thrice-recusive Scrabble puzzle.
  • Orange/A: cross-word puzzle with lots of solutions, but only one corresponding to the appropriate song they had in mind
  • Orange/B: An amazing amount of work to put this together (AND to find all the rubbings), and an incredibly beautiful solution. I think this was GC's favorite clue.
  • There are several good crossword puzzles. I didn't work on any of them.
  • Purple/Hotel: This was the last hotel puzzle to be solved (by any team). We had it for over 24 hours. Nasty.
  • Red/B: My vote for coolest puzzle. I didn't work on this. Clever solution to true/false type puzzle.
  • Blue/C: I actually brought my "Lost Treasures of Infocom" CD.
  • Magenta/B: I had no idea there were so many kinds of rock music. These were playing for HOURS in the middle of the night as it was being solved.
  • Magenta/C: Based on Al Hirschfeld drawings. We never got this but it was sure cool.

(2002 hunt homepage)

2003, ACMECorp / Matrix

The 2003 hunt started at noon on Friday, January 17th, 2003. I played with Third East again, as team Kappa Sig, which found the coin at 7:26 am on Monday morning, January 20th, in the longest Mystery Hunt ever.

The hunt started with ACMECorp CEO N. Ron Aldephia beginning a speech and getting shot down by the fastest-firing Nerf dart shooter I've ever seen. Our task was to decipher who wanted to kill him based on documentation he left in 7 places on the 7 previous days.

However, puzzle wave 1 not only had puzzles and a metapuzzle but also two letters hidden per puzzle, which when strung together spelled "TAKE THE RED PILL". (metapuzzle answer) This led to my favorite moment of the hunt, where 5 of us ran from building to building inside MIT, first logging into a computer and receiving instructions to go to a lobby, then corralling the Airborne Express delivery person there, who sent us to another lobby where the phone rang, then finally to a dark conference room where we met Morpheus and took the red pill.

ACME's counterorganization, led by Morpheus, was called VILE and they let us know that we had been solving puzzles in the Matrix. We would continue to solve the puzzles in the Matrix, but also had to solve training puzzles and reality puzzles. Training puzzles altered the Matrix puzzles' answers ... the metapuzzles, with altered answers, led to a runaround involving the reality puzzles that ended at Gilbert Strang's door, which led to a Twister game, which led to the final runaround and the coin hidden in a pair of pants in building 16 ... a little complicated, but definitely understandable, and a very cool plot point. (Endgame, as described by ACME.)

In 2001 and 2002, I had bought a Monday ticket home and the hunt ended early on Sunday. So this year I bought the last ticket out on Sunday. Bad idea. I left on time, got home, and the hunt was still going on. In the end, our team did find the coin, so we'll be planning next year's hunt.

Favorite puzzles:

(2003 hunt homepage, MIT Tech story about hunt)

2004, Time Bandits

Because our team won in 2003, we ran the 2004 hunt. Tons of fun to run, despite a hunt that, well, was poorly received. Nonetheless, I'm glad I got to be a part of it.

Our opening skit looked like a Pirates theme, and each team received a map of a pirate island. As time went on, however, they discovered a wormhole on the island that led to a wholly different map. The theme was in fact a Time Bandits theme, and the teams each visited 6 other spacetimes (Aztec, Vatican, Timbuktu, Las Vegas, Yukon, Neo-Tokyo) with associated puzzles. At the end was a Titanic spacetime (in conjunction with the movie's plot).

2 teams were neck and neck at the end, with Random Hall beginning the final (Titanic) phase of the hunt about an hour before 2-time champs Setec Astronomy. Random was about 1/2 hour ahead when the final runaround began, but Setec pulled it out by about 15 minutes in the end. The hunt was about 68 hours long, one hour longer than 2003's and thus the longest on record.

We had a huge screen with the hunt status on it during the endgame (well, throughout the whole hunt, actually) and were cheering. Unencumbered by experience, it was like being at an off-track betting parlor and watching results come up on a big screen.

I did a fair amount of phone support, a little coding for the results, acted in most of the spacetime intro skits (Pirates, Aztec, Vegas, Vatican, Yukon, and Timbuktu), did some of the runaround with Setec, figured out some policy things (we needed to speed up the hunt in a big way as it turned out to be too long - even so, longest hunt ever, lasting an hour longer than the year before).

The team put in a huge amount of work to make this all run smoothly. Unfortunately, most who played didn't like it (Greg Brume, Francis Heaney, Jeff Barrett, Aaron Fuegi, Dan Katz, Anand Sarwate, etc. -- all of whom make really good points and are among the more civil of the folks who commented on the hunt). The hunt was too long, and the puzzles weren't well written, and these two things happen to a lot of new puzzle writers which we were. Compounding the problem was that we sped up the hunt to make it finishable, and we think some of the teams didn't get to see some of the coolest elements. I think our puzzle-writing skills will improve from our experience, but feedback indicated that we still have a while to go. I think the feedback was good, but brutal. One thing I learned in running an event was that it's really important when giving feedback to puzzle writers, particularly first-time ones, is to try to make it constructive and try to also concentrate on the good things they did. I think the team was pretty floored by the brutality of the comments we had, particularly given that the time commitment from the main folks planning (not me) was obscene. Hopefully worth it for all the participants; it was sure a lot of fun for me, if a little sad that we didn't do a better job.

I wrote five puzzles in my first puzzle-writing experience: English Class, Janus, Girl, Infirmary, and the near-universally despised Leftovers.

(2004 hunt homepage, Tech article, Headquarters pictures)

2005, Normalville

Setec Astronomy ran the 2005 Mystery Hunt with a fantastic effort all around. Normalville was beset by a meteor shower. Solving puzzles opened up more puzzles, and the creeping progress through Normalville's map was a lot of fun. The metapuzzles allowed us to acquire powers of various sorts that we would then use in the endgame.

We played remotely from the West Coast with a small (by MH standards) team of about a dozen Bay Area Gamers, most of whom had never done a Hunt-type event before. Martin Reinfried hosted, Brent Holman and I organized our team ("Left Out"), and Mike Springer captained the tiny Boston-based "Army of Very Few" that was essential to our success. We weren't anywhere close to winning (results, looks like we got about 12th) but I'm fairly sure we did awfully well on the puzzles per person metric, as most teams were substantially larger than ours. PhysPlant (from Random Hall) won and will run the 2006 Hunt.

2005 hunt homepage

2006, SPIES

2006 hunt homepage

John Owens | Game Home Page | Last updated .