Fort & Forge

Mystery Hunt 2022

Jan 22, 2022

This post is about the 2022 MIT Mystery Hunt. You can see the hunt website with most of the puzzles and solutions at bookspace.world.

After over ten years of participating, my team finally won the MIT Mystery Hunt for the first time last week! In this post I’ll talk about my hunt experience.

A brief history of my involvement with puzzlehunts

…but first, indulge me for a moment as I detail my involvement with puzzlehunts over the course of the past decade.

I was first introduced to puzzlehunts in March of 2010, when a friend forwarded me an email inviting me to join his team for the Bay Area Puzzlehunt which took place on Stanford’s campus, and which I now realize was run by none other than Sam Bankman-Fried. Our team didn’t do particularly well, but we had a ton of fun and I was hooked. In May of that year I participated in the CiSRA puzzle competition with a few friends; I’ll never forget assembling the 3D sculpture described by the puzzle Simple Instructions in my middle school’s library.

In the summer of 2011 I attended the Canada/USA Mathcamp. They hold a day-long puzzlehunt in week 4 that basically the entire camp participates in (this is also where SBF got interested in puzzlehunts, I suspect). Mathcamp had a semi-associated Mystery Hunt team called the Manic Sages, so I participated in the 2012 MIT Mystery Hunt remotely with them. I didn’t contribute much—I still had school on Friday, and was pretty out of my depth anyways, but I did provide the key insight that led to our solve of the Okla-Holmes-A! meta.

Incidentally, we actually won that year, so I “helped” write the 2013 hunt. I’m credited on two puzzles; the first one is halfway-decent, though nothing I’m super proud of. The second is one of many puzzles in the 2013 hunt that is actually nigh-unsolvable and in retrospect probably should never have made it into the hunt.

In 2014, the Sages disbanded, and I hunted with One Fish Two Fish Random Fish Blue Fish (their current name is Hunches in Bunches). I do remember participating, but I also remember checking out of the hunt fairly early on because I felt like I wasn’t adding anything meaningful. They also ended up winning that year, but this time I didn’t stick around to help write the 2015 hunt.

In Spring of 2014 I was accepted into MIT and visited it in April during Campus Preview Weekend (CPW). One of the dorms (Simmons Hall) held a small puzzlehunt, which I took part in along with some friends and friends of friends. This was the birth of teammate, the puzzlehunt team. Most of us chose to attend MIT (and some of those who didn’t ended up at a certain school not far down the river…), and we continued participating in puzzlehunts in the fall of our freshman year (Palantir, SUMS, and BAPHL 11).

We were all excited for the upcoming 2015 Mystery Hunt, and we decided to join ✈✈✈ Galactic Trendsetters ✈✈✈ since we had a number of friends on the team. This was my first in-person Mystery Hunt, and it may forever be my favorite hunt of all time. (Some moments that stand out: the do-it-again step of the Coral Reef Meta, our initial breakthrough for the Colorful Tower Meta, and the key insight of the Atlantis Meta Meta about the four seahorsemen of the apocalypse).

Teammate continued to do puzzlehunts and continued to steadily improve. (When we first participated in the Palantir puzzlehunt, I don’t think we even finished, but by our senior year we were consistently in first or second). In 2016 and 2017 we again hunted with the Trendsetters for Mystery Hunt. The 2017 hunt famously finished early, so some of us on the team used the extra time on Sunday to start writing our own hunt, which eventually turned into the first Galactic Puzzle Hunt.

In 2018—in part because of how fast we finished in 2017—teammate decided to split off from ✈✈✈ Galactic Trendsetters ✈✈✈ and form our own team. We were joined by a number of friends from CMU and Berkeley who had also been hunting with the Trendsetters in prior years. We continued hunting separately in 2019 and 2020, and all the while, I continued helping write subsequent Galactic Puzzle Hunts.

When ✈✈✈ Galactic Trendsetters ✈✈✈ won Mystery Hunt in 2020, I and a few others decided to help them write the 2021 Mystery Hunt. I talked a lot about my experience constructing hunt in this blogpost. I’m credited on 22 puzzles, of which I’m probably most proud of Nutraumatic, ⊥IW.nano, and MacGregor House.

In the meantime, the rest of teammate wanted to try their hand at writing a hunt of their own and produced the 2020 Teammate Hunt (followed by the universally-acclaimed 2021 Teammate Hunt). Teammate eventually participated in the 2021 Mystery Hunt (without me) and did quite well.

I rejoined teammate this year, and WE WON! All of this is to say: even though I technically was on a team that won in 2012 and 2014, and I’ve technically written for hunt twice already, in my mind this is the one that counts.

Hunt Weekend

In 2021, teammate (like all other teams) hunted almost entirely remotely. Everyone was in their own isolated spaces, and all communication took place on Google Sheets and Discord. This year with vaccines available we decided to organize some hubs where people could gather and hunt together, even though hunt was still remote. We organized 3 such hubs: one in Boston, one in the Bay Area, and one in Seattle.

I flew out to join the Boston hub. We booked a 3-story Airbnb in Allston and had about 20 people in attendance over the course of the weekend. Honestly, it was the best hunting setup I’ve ever experienced: easy access to beds, showers, a kitchen, a coffee-maker; lots of nice desks and tables perfect for hunting; being surrounded by friends, some of whom I haven’t seen in person in 2 years. It was so good that we’re seriously considering doing doing a similar thing in future years even if hunt stops being remote. The downside is that it’s difficult to get such a nice place within walking distance of MIT’s campus. This wasn’t a problem this year, but it’s something we’ll have to think about in the future.

It was, of course, a risky time to host such a large gathering. During the weekend of hunt Boston was probably just past the peak of the Omicron surge. We went in clear-eyed about the risks and did take lots of measures to mitigate them as best as possible. We required proof of vaccination for everyone and negative COVID tests taken no earlier than the day before hunt. We also administered rapid tests to about 1/3 of the hub on Saturday and Sunday to detect spread; none of the tests were positive, so it seems like we managed to escape unscathed.

I arrived in Boston on Thursday afternoon and took a Lyft to the Airbnb, turning in early in order to get a final good night’s sleep. The start of hunt felt like a series of never-ending countdown timers, but eventually kickoff did begin in earnest and we were introduced to this year’s theme: books!

This was, in my opinion, a shrewd choice. Running teams have been moving away from specific pop culture franchises towards more general concepts in order to ensure widespread appealability and to encapsulate rounds of varying content. This year’s theme allowed for rounds about horror, romance, and food, that each felt distinct but thematically linked. I also greatly appreciated the connection to MIT via the integration of the five ministers, each of whom represented one of the MIT campus libraries.

Structurally, the hunt took the form of ever-increasing rounds of puzzles, the culmination of each of which made for a good finishing point for teams of increasing sizes/skill levels. First there was the Star Rats prologue, released a couple of weeks prior to Martin Luther King Jr. weekend, a first for Mystery Hunt. Then there was The Investigation, a 10 puzzle round with a charmingly-simple meta, followed by The Ministry, a 25 puzzle round with 5 submetas, a meta-matching gimmick, a metameta, and a fruitaround. All of this was a precursor to the main event, Pen Station: 10 substantial rounds each themed around a different literary genre.

I can’t speak for smaller teams, but I appreciated the effort Palindrome took to build out these smaller milestones while still generating enough content (~196 puzzles!) to satisfy teams like mine.

The Ministry (spoilers!)

Teammate made quick work of The Investigation and the meta-matching portion of The Ministry, so we quickly found ourselves working on The Ministry Meta Meta. Early on when we were solving this, someone asked if there was a mural outside Hayden library, since the flavortext seemed to be strongly clueing such a thing. I replied that there wasn’t, to my knowledge. I later re-read the flavortext and realized that it was clueing a mural—this one outside of building 14.

Engineers looking at an elephant

This gave us an ordering on the 5 ministers which allowed us to solve the puzzle. In my defense, I think Palindrome was a little bit confused about the distinction between Hayden Library and building 14. Hayden library is located within building 14, and this mural is located just outside building 14, but it’s really not accurate to say that this mural is outside Hayden Library. (Incidentally, the Lewis Music library is also located in building 14, so as someone pointed out, if all of building 14 disappeared into the plot hole, Minister Lewis should probably be freaking out as well!).

Regardless, I thought that this was a great puzzle. Every time I passed this mural, I always had this thought in the back of my mind that it was useful puzzle fodder. I’m glad to see that someone from Palindrome thought similarly! I also loved how the answers to the 5 minister metas were both semantically meaningful, and could be re-applied to the 25 puzzle answers.

The answer to this meta was A VORACIOUS BOOKWYRM RUN AMOK. I was very proud of myself for correctly surmising that this was referring to The Very Hungry Caterpillar (one of my favorite books growing up), but apparently just about every team guessed this. Palindrome said at wrap-up that something like 60% of teams submitted A VERY HUNGRY CATERPILLAR to the subsequent fruitaround puzzle.

Caterpillar

My team really found it weird when we finished this puzzle and had 0 puzzles open. We were blocked from moving forwards by an interaction, but Palindrome seemed to be really overwhelmed at this particular point, and we had to contact them twice in order to schedule it. Looking at the stats, it seems like lots of teams were probably finishing around this time, so this was pretty understandable.

Pen Station

We eventually did unlock and solve the Fruitaround, and found our way to Pen Station.

Pen Station

Now’s a good time to talk about the art this year, which I adored. There’s a certain type of nerd who loves public transit and libraries, and I suspect they were gaga about the homepage for this hunt, shown above. Initially there were only two “lines” open on the Information Kiosk; the rest were “under construction.” This did a good job of unifying the subsequent rounds, while previewing the awesome scope of it all. The art for each of the individual rounds was also wonderful; every time we unlocked a new one, it was a joy to check it out.

From this moment on, we never really got seriously stuck at any point (unlike in 2020). Some of this is probably that teammate has gotten better at meta-solving, but I think it has more to do with the particular metapuzzles that Palindrome wrote. There were only a few instances where we knew what we had to do to solve the meta, but simply needed more puzzle answers in order to do so. Furthermore in all of those puzzles (Masquerade Ball, How to Find a Component, and Introspection), the reason we needed more answers was purely for the mechanical reason that we unlocked new information every time we solved a puzzle rather than needing the answers in and of themselves, which generally meant that Palindrome had written the puzzle in such a way where they could guarantee it was solvable with some reasonable proportion of puzzles solved.

There also weren’t any metas that asked us to do something truly insane like simulate an electronic circuit out of the puzzle answers or crack a chaocipher with only part of the key, which meant that once we had solved enough puzzles and made the initial breakthrough about how the meta worked, we’d usually have the answer within 40 minutes or so.

As such, the rest of the hunt was kind of a blur. I grabbed about 5 hours of sleep on Friday night, and pulled an all-nighter on Saturday night. Mystery Hunt is unique among puzzlehunts in not giving you any feedback on your relative ranking while you’re doing it, so we were constantly speculating on how we were doing all weekend. There were a number of clues:

All this suggested we were near the top of the pack, but maybe not in 1st (this ended up being basically correct), so it wasn’t obvious to us when we finished Tollbooth and had our final interaction with Palindrome that we had won.

This, combined with our collective sleep deprivation and the limits of a virtual experience, did lend a bit of anticlimax to the finale. Although I wish my reaction was closer to one of pure joy, it was similar to Alex’s:

“oh, we won, wooooo, I’m going to sleep.”

Looking forwards

In the summer of 2021, when Palindrome announced that MH22 would be remote, I was disappointed by the decision and thought it was wrong-headed. Obviously, I was wrong. With omicron surging across the country there was no way that hunt could have taken place in person. I continue to remain optimistic about the course of the pandemic, and my most fervent desire with respect to MH23 is that we can hold it on campus. Realistically though the seasonal nature of COVID-19 does not interact well with the schedule of hunt.

Regardless, I can promise that teammate will work hard to deliver the best hunt we can. This is the first chance many of have had to write for Mystery Hunt and we’re all excited by the opportunity. Thanks again to Palindrome for a fantastic hunt, and see you all next year!