MIT Mystery Hunt 2026 Recap

Another year another great time hunting during the frigid throes of January in Cambridge. Once again I hunted with NES and for second year a row we both managed to finish and also notch (unofficial) 4th place! I’m not sure how long we can maintain this consistency.

In-personness

As for what Cardinality did with the hunt this year, I’m very happy to see that they maintained the ethos of Mystery Hunt having a lot of in-person experience that D&M set last year. With so many fantastic online hunts throughout the year, I think it’s good for Mystery Hunt to lean in to what makes it special: to me that means lots of in-person things like interactions/events/physical puzzles. I think Mystery Hunt also means larger teams where puzzles can get more niche and the metas can get more hairy, but I don’t think that’s ever not been in hunt.

I really liked the idea of translating the traditional hunt “Events” into more smaller events throughout the weekend that more people can participate in. For most years, when a small dorm team is doing hunt, everyone can get involved. And I do think that these events/runarounds/interactions are what can hook a newbie into puzzlehunting for the long-term. It’s how I really got hooked on hunt! But I worry sometimes that if someone’s first experience is joining a larger team, they sometimes miss that part, because they don’t want to volunteer to be one of two people to do an event. With the Research Point framework, I think there was plenty of room for anyone to get involved. I also think the progression of the Scavenger Hunt nicely fit the Research Point framework–I always think it’s good to spread it out throughout the hunt and some previous hunts have found their own ways to do that–this worked just as well too!

That said, I also think the hunt-wide all-team events can also have their place–they have been some of my strongest memories of my first hunts. But I think the tradition of having four big canonical “events” doesn’t necessarily have to stay unless the running team has ideas they really like for them.

Spreading things out

This idea of taking the singular events of “Scavenger Hunt” and “Event” and smearing it across the weekend into smaller “microdoses” of RP fun I think also touches on another theme of this hunt (and many others) that I found myself thinking about. I think with the nature of Mystery Hunt being an experience with much larger teams–teams that often are comprised of a spectrum of puzzle experience and enthusiasm–it’s good to try to offer something for everyone for as long as possible. As an example, the Kingdom could have been a ‘traditional’ opening round (albeit a fairy long one), and then after finishing (or getting most of the way through) you would get to the crazy “dimension” rounds with their gimmicks.

Certainly this structure is more narratively appealing for many themes and many hunts have felt like that. However I think it’s almost always better for Mystery Hunt to try to do as much in parallel as can be managed–which this hunt did well. The Kingdom lasted for quite long into the weekend, and the Dimension rounds worked in parallel. This is many practical benefits, since some people would rather look at the MonQuest video game and work on easy puzzles, and some people want to see crazy rounds and fun structures. And also because if you are a small team that doesn’t actually progress very far into the hunt, you can get a taste of everything.

This is far from the only hunt to have this as a feature: the classic “fish” round like 2024’s Hole in the Ceiling of Hades run parallel to the main-track of harder puzzles. 2017’s Character puzzles ran parallel to the Quest puzzles. Last year’s choose-your-own-adventure mechanic allowed teams to recreate this for themselves–and I think NES’s choice to mete out the Stakeout puzzles for as long as possible instead of blowing through the round was great for team-wide morale. But I also remember, for example, Penny Park from 2020 whose “cool” rounds were all hidden at the back of hunt, and while I did love post-solving the Safari (one of my favorite hunt rounds of all time!), I was quite disappointed to find out about all this cool stuff at wrapup that we just barely didn’t get to.

All this to say, it’s certainly not a unique hunt-structure idea, but I want to call out its great execution in this year’s hunt. Along with the RP system, it felt like it contributed to a feeling that this hunt tried to spread as much out throughout the weekend as it could, which I think is a good goal for specifically Mystery Hunt. After all, who doesn’t love drinking from the firehose! So much was going on Friday afternoon in our classroom all at once, in a very good way.

Choose your own adventure

I liked the “puzzle blurb” unlocking that was used in the Kingdom, as well as all of 2025’s hunt. I appreciate that Cardinality felt they could copy this, but not feel the need to apply it to the whole hunt. I think it’s good because there are pros and cons (in my opinion)

Pro: A sense of agency and decision-making is always fun in any event
Pro: Smaller teams can just focus on the puzzles they want to do
Pro: For niche puzzles (crocheting or cryptics or whatnot) we can make sure people who want to do it are around when it gets unlocked
Con: For a team that’s going to do everything anyway, it can be hard to resist the temptation to eat your desserts first and then be stuck with the Sunday vegetables. A curated experience can be good
Con: Some feeling of progression and surprise and “oh! what’s coming next” is fun too. It’s all about variety IMO

So basically I think they nailed it–and I would totally understand if this became the norm for at least the areas of the hunt targeted towards small/medium teams.

My experience puzzling

After talking such high praise about the RP System and the parallelization of the Kingdom and Dimensions, you’d think I enjoyed going back and forth between all these elements of the hunt, but you would be wrong. As soon as we unlocked the first Dimension (Land of No Name) I decided to focus entirely on Dimension puzzles and never looked back.

While in my early years I yearned for runarounds and events, I find nowadays I just want to sit and do hard puzzles with my friends in a buzzing classroom. Luckily there were many about, and the crazy structures provided much enjoyment throughout the weekend.

The Kingdom puzzles I did work on in those first few hours were:

A Wanderer’s Colorlog ~ fun opening puzzle to look at
Crossloop ~ exactly the kind of word puzzle I want to be doing in the first hour (or any time really)
Do They Have Chemistry? ~ i heard whispers in the room of a “yaoi logic puzzle” and I was drawn like a moth to its flame. But to be honest, this is the kind of puzzle that I think works great in Mystery Hunt specifically: it will delight a small group of people but maybe not hit with everyone.
Central Precinct and Lockup ~ I am a simple man: I see the @Cryptics role in our discord server, I join

At that point we unlocked The Land of No Name. This was an insanely cool round and I spend a fair amount of time on it, mostly eschewing the Hyperbolic Space. I loved both the ability to prioritize across many puzzles and see where we can/should make progress, and also the feeling of progression as we gained letters and more became available. I experienced my highest highs and lowest lows all hunt long in this round, so here’s the parts that stood out to me the most:

Six Shifts ~ many other writeups have mentioned this, but this was probably my favorite puzzle in the entire hunt. This solve was just so satisfying and is something that only really makes sense within the context of this round. The ability to have puzzles like this was a really fun feature of this round structure. For the record, our letters when we solved was {J,M,Z}. I was riding that high all day…
Imaginary Factorizations ~ I felt this was another unique solve to this round. While it was certainly solvable from the beginning, we managed to grok the title from a few letters, and discovering the title basically told us how to start and we solved the puzzle from there. “Figure out the title and then you can solve” was a new thing to me!
Odd One Out ~ In general, having to decipher text (like crossword clues) in this round was a fun feature. This was a great use of that, because once you start to break in you also start to realize how much of a shitpost this puzzle is, in a good way.
MIT Dropquotes ~ Another fun word puzzle where the round gimmick just makes it unique without any extra work from the constructor. I was surprised at how solvable this was with only ~half of the letters, since there’s extra constraints in the columns being alphabetical order. That said, it still took a fair amount of time to actually finish them, but it was a novel experience to literally have the puzzle get easier over time, as other teammates finished other puzzles in the round. I realize that’s the whole point of the round, but it just felt very direct in this puzzle, since there wasn’t really an aha we were waiting on.
Bars of Music ~ this is a fun idea that nerdsniped me, but unfortunately the errata kind of ruined it. I’m completely sympathetic that the website change came out of nowhere and blindsided the constructor, but I also unfortunately lost more than 2 hours on a puzzle that was ultimately impossible before the errata. The final extraction was just vague enough in how to get the IDs that it was plausible we were just missing the correct method

At this point I started to look at…

The Alphabet (Land of No Name meta)

After getting the initial constraint on the answers, teammates managed to sidesolve two puzzles. After that we were left with two: Skilldrasil and Not A QR Code. At this point, everyone was discussing their own theories on what the numbers in the meta puzzle would mean exactly (was it the lengths of A-Z expansions? was it merely an index into the final expansions?). I was fairly convinced they were the actual lengths and took to python to prove it to myself.

I started Friday night but ended up spending a lot of Saturday morning in a jupyter notebook. After playing with the mechanic of the meta, I managed to convince myself that they were the actual lengths, but were out of order. This was a really fun conclusion to reach! It did feel kinda like actual data science to just be trying things and building intuition for what this system tended to do.

At that point, I tried to dictionary attack our last two answers: unfortunately I over-pruned on potential answer lengths and didn’t find the right answers. At that point I went and tried to solve Skilldrasil through sheer force and willpower. It took two hours, but I did manage to miraculously see the hidden message we needed to complete the forward solve. This was quite exciting!

One minute later I got the answer to Not A Qr Code from my program–I’m sure from HQ this looked like two backsolves after finally getting my dictionary attack to work, but Skilldrasil was a true forward solve! Even better, I already had most of the programming done to generate the paths for the meta, since I had done so in my data-science-exploratory phase that morning.

Unfortunately, I kept assuming that once we had the paths, ‘it would be obvious what to do next’. Maybe the paths would end at one of the highlighted letters. Maybe it would enclose exactly one of them. This was wrong–it was not obvious (at least to me). From the entirety of 2:30pm until about 10:30pm, I worked on this meta. Looking at as many aspects of all 26 paths as I could, and continuously onboarding people to the puzzle to see if they could have the golden insight. It wasn’t until Cardinality graciously gave us a hint that we managed to finally fell this giant.

Now look, I can’t complain too much. I certainly put myself in this bind: I could have put my foot down and just abandoned the meta until Sunday when everyone had fresh eyes, and when a hint was more inevitable. But I persisted, and it’s at least a pretty memorable part of hunt to me. I wish that I spend more of Saturday doing literally any other puzzles, but I also could have just done that? No one was forcing me. Sometimes, the enjoyment is in getting stuck more than you’ve every felt stuck, but somehow getting out of it. I may have also been motivated off of willpower-ing my way through our Skilldrasil forward-solve–I thought the same grit would work here!

Looking back, I think all the meta needed was a little more flavor text to clue the final extraction method. I’m also not sure how hard this meta needed to be, since the main enjoyment of the round was the round gimmick itself. As I said, this round gave me some of the biggest highlights of this hunt (or really any hunt) and also some of the biggest lowlights. That’s a good thing! It was a salient experience and I’ll definitely remember this round more than most other hunt rounds. The idea was just too good

The rest of hunt

I dabbled in the rest of the rounds afterwards. Here were some highlights:

The Puzzle that Cannot be Named (Mosaics) ~ very fun, very challenging cryptics. very fun to discuss as a committee
Lines (Mosaics meta) ~ This is a great idea and it was fun to do and it was completely fair, but the rules are a little weird conceptually, because IRL I’m pretty sure the thin line is a stronger connection that the circle. Anyone in NYC would tell you that the A and L trains also intersect at 14th St-8th Ave (and therefore don’t have a unique intersection), but they don’t in this puzzle
All of the Terminus Puzzles ~ I think some puzzles that don’t fit the traditional hunt mold but are fun and breezy are always welcome. They just often have to be in their own round to make sense: but Mystery Hunt is a perfect place to spend a round on short ‘puzzles’ like this. The “Can you really call it a crossword if…” puzzles were more normal, but the Trends and Layers puzzles were just fun jaunts.
Point of View ~ one of the rare puzzles that spawns an inside joke for us. There were so many picture of Arthur pointing in our sheet, and he always has “an original point of view”. Also after opening this puzzle, people were actually confused because some people were already zoomed in on their browser and just seeing different pictures.

After that we did the endgame and had a great time. We ended up using a few hints to push ourselves through some of our last puzzles so we could make sure we finished on time, but we only really started doing hints Sunday afternoon (besides the impromptu The Alphabet hint we got late Saturday night)

I had a great time solving with all of my friends on NES and I still can’t believe that we’ve managed to become a ‘real’ team that actually has the potential to finish a whole Mystery Hunt! Who knows what else we have the potential to do now