2020 MIT Mystery Hunt

The “metameta” (which didn’t actually involve any regular or meta answers, so I might prefer to call it more of just a “capstone puzzle” that you unlocked components of as you finished each round) consisted of a bunch of custom-pressed pennies, each with three icons and arrows on them, on cards with descriptions. A lot of what people wanted to do involved physically rearranging those pennies to make the arrows point at each other and such, which was limited by the number of people who could stand around a table with all the pennies, so I didn’t get involved in that part of the puzzle. Nevertheless, there was a lot of data to work through. I think it took us at least an hour, probably two, to notice that each card description had exactly three sentences.

We tried arranging the coins in lots of ways. At some point, to increase the number of people who could arrange coins, I hacked together an HTML file with rectangular “coins” that you could drag around thanks to the magic of jQuery. I think we figured out the right coin arrangement (making arrows point at each other and the phrase “PENNY PARK” oriented upwards on each coin) after a few false starts, and gradually gained increasing confidence in it. What gave us the most confidence was when we made a bunch of solvers who had just woken up try arranging the pennies independently of the rest of us, and they converged to the same arrangement. On the other hand, there were also a bunch of other things people tried doing on the side, including matching the descriptions or rounds to Disneyworld locations or real MIT locations, as well as removing two pennies from the correct configuration (so, those two pennies would be Left Out…) to make a thematic cent sign.

Meanwhile, I spent most of the time I remember in of an offshoot group that focused on the three icons on each coin and three sentences on each card, and simply tried very hard to match sentences to icons. However, we didn’t think of the idea of adding words. Instead, we just tried really hard to force sentences to icons in any way possible, even if the icon was on a different coin, which we got pretty far with — the most “creative” example I ended up with was that “We have three goals: fun, fun, and FUN!” could be linked with the ANGEL icon, because the way the same word “fun” was repeated three times represented the Holy Trinity. (Having gotten 5½ hours of sleep the previous “night” and then been awake for maybe 20 hours, I had a lot of fun playing sleep-deprived word association.) I am also certain that we said multiple correct two-word phrases in this process, but it never clicked.

At something like six hours into the process, I was brainstorming in a corner of the room with Anderson and a few other hunters, since Anderson had just woken up. My memory is pretty fuzzy here, but I think Anderson pointed at Safari Adventure 2’s “full-fledged entertainment center” and said that it could be a MAN CAVE. I might have privately thought that phrase before, but only fleetingly. This time, though, hearing another person say the phrase aloud made it feel like a real possibility in a way it wasn’t before. I think I looked at the first row of our spreadsheet — CASTLE, STAR, FISH — and recalled how I had previously argued a NORTH STAR could “help [patrons] navigate the sprawling grounds” (this wasn’t even the correct word to add to STAR, but it made enough sense at the time). Then I looked at the sentence before it, a sentence about a “temporary structure” that we hadn’t convincingly connected to any icon, and it actually clicked. The world came into focus, the black letters snapped at the end of the Revisualization sequence:

SAND CASTLE.

Everything after that was a blur. The room shifted. Scrambling, everybody converged onto a Google spreadsheet called some variant of “the pog zone”. We started labeling the connections between coins with two-word phrases. The off-by-one-letter realization was easy to get, and eventually a phrase started to emerge.

At this point, we had dozens of excited solvers spamming Wheel-of-Fortuned guesses, including guesses that simply couldn’t be correct by virtue of being not exactly 24 letters long, and were quite afraid of getting locked out of answer submission. Unfortunately our team was big enough and had enough remote solvers that nobody really knew who was spamming guesses, which is why we sent multiple “@everyone”s into the Discord and also threw bright read STOP GUESSING signs all over the spreadsheet. Still, we didn’t actually get locked out and I think it took just 20 to 30 minutes from when we figured out the last for us to guess the final phrase.

The rest is history. We split up into groups to do 10 mini-runarounds, signed up for an appointment with Left Out, and went on the final runaround, which was already covered pretty well during wrap-up, so I won’t dwell on it. And that was the 2020 Mystery Hunt for us.

(metameta spoilers end here)