2020 MIT Mystery Hunt, Part 2: Some Puzzles I Remember

(This is a recap/review of the 2020 MIT Mystery Hunt, which happened this month. Puzzles and solutions can currently be found here. This recap may contain spoilers.)

[Before I start, if you’re seeing this shortly after I post it on Sunday, you may have time to drop by Left Out’s AMA on the Mystery Hunt subreddit. And even if you’re seeing this later, I’m sure the Q&A there will still be interesting to look at after the fact!]

It’s more standard for me to start by talking about structure, since that tends to be my favorite thing about a new Mystery Hunt, but I expect I’ll have more complex things to say about that, so I’m going to start with a rundown of some puzzles I remember solving (or not) and have something to say about. They don’t necessarily represent my favorite puzzles that I did, and certainly not my favorite in the Hunt as a whole since I still haven’t seen everything!

The Ferris of Them All

This was my first instance of following Setec’s planned “casual approach” to this year’s Hunt; this released very early, and while abstract logic puzzles often get devoured very quickly on our team, Jackie and I decided to print out our own copies regardless of what everybody else was doing. (The team explicitly mentioned solving puzzles in parallel as a way to make sure we didn’t finish first… in practice, when one group finished a puzzle, everyone else usually stopped, and if a group didn’t let other folks know what they were doing, feelings sometimes got hurt, so we’re still working out the kinks.)

Some steps of the puzzle were harder than others; I think on Step 2 we realized too late that looking ahead to Step 3 would have eliminated some options, and so on Step 3, I used some constraints from Step 4. Jackie and I mostly worked independently and then combined forces when we got stuck. If I’d known we were going to be struggling to finish the Hunt a couple of days later, I would have approached this (and probably some other early puzzles) very differently.

Star Maps

Star Maps was probably the puzzle I put the most time into solving, and it’s another case where if we were trying to win, I probably wouldn’t have focused it for so long; but it was very satisfying, so I’m glad I stuck with it. I joined the puzzle after the jigsaw had been assembled, at which point we needed to decide where to place the red dot pieces. We briefly pursued a red herring–surely a puzzle called “Star Maps” with fifty nodes would have something to do with connectivity of the 50 states–but Alaska and Hawaii eliminated that option, and as we got some dots placed that seemed to match up, we realized each edge seemed to pass through one graph. The more dots we placed, the easier remaining ones were to place, which led to a nice solve flow. We didn’t do ourselves any favors by setting the puzzle up in a room with poor lighting; between that and the small printing, we eventually had to use flashlights to finish solving.

For the graph identifications, we ID’d about 15 out of 50 properties (I’ve never taken a course in graph theory but I used to dabble for fun), and once it occurred to us to assume the 5’s at the ends were all “graph,” that gave us enough letters to wheel-of-fortune the first 40 out of 50 letters of the clue phrase. For the last ten, we got stuck with T???PL?????, and even once we guessed that the PL?? might be PLUS, we didn’t know what the T??? could be. With some more ID’s, we got to TW??PLUSO??, still perplexed by the four letters before PLUS; finally somebody suggested TWON (it was actually K, and I still don’t know what a snark is), and we were able to draw the answer. It was a loooong process, but it got more approachable a bit at a time, and it was very rewarding to make it to the end.

Looking at the solution, I can tell you we never wrote the properties on the intact sheet, and therefore I had no idea until now that the properties were alphabetized on the sheet. And don’t get me started about the easier-to-read erratum that came out after we solved the whole thing.

Peak Adventure (and Ox and Chairiot Races and Penny Park Guide, etc.)

I’m a fan of the sort of puzzle that is actually a little suite of variety puzzles; it provides a lot of ahas, and if any of them are unusually hard to approach, you can usually solve them from a large subset of solutions. (This should not be surprising… I like puzzlehunts enough to have a poorly updated blog about them, so of course I’m going to like puzzles that are little puzzlehunts.) I’d say the only real drawback to this kind of puzzle is that it’s time-consuming to write, so I have no idea how Left Out wedged a whole bunch of them into their Hunt.

The one I interacted with the most successfully was Peak Adventure. Clicking around the puzzles in Cascade Bay at a time that most of our power solvers were hammering away at the Safari metas, I found a nice little collection of a dozen mountain-themed puzzles. Checking our solving spreadsheet, I saw people had spreadsheetified a lot of data, but only about four of them had been solved, so I started bouncing around completing partial solves and getting a few ahas myself. In the last stage, Jackie and Josh Oratz and I figured out which three winter weather words could be combined with nonconsecutive substrings of our answers (well done by the author in making the set enumerations similar so that you needed most answers to start placing things confidently), and we worked out a final answer.

Ox was the puzzle of this type I heard the most positive feedback about on social media; I didn’t look at this during the Hunt, but Jackie and I cosolved a bunch of the subpuzzles after Hunt (I’d already spoiled myself on the extraction) and they were fun and creative, though I was sad that the sound file link was broken. There were at least two more mini-suites that proved too hard for us overall; I was sitting next to the group working on Penny Park Guide, and there was a lot of cool stuff going on, but in particular there was a lot of it, and by the time they solved it (still forward, I believe!) I believe we had solved the metapuzzle for that round and had all the other answers. I did work on Chairiot Races, and after hours of solving and many people looking, I don’t think we solved half of the subpuzzles… which was a shame since reading the solution after the fact, the end step seemed really cool. I think the pieces of this last puzzle suite were probably miscalibrated in terms of difficulty, and I’d be curious to hear how many teams solved it as intended.

Spaghetti Western

I almost included this in the mini-suite catchall section, but this one was inherently different, since the individual puzzles overlapped by four inputs, which means the main first goal was to solve for those inputs by comparing ALL of the minipuzzles, and then eventually completing each one independently. I’ve certainly played Spaghetti on Eric Berlin’s social media before (in fact, I was unexpectedly the author of the weirdest example “solution” when he presented it as an exercise at a recent NPL Con), so I appreciated the reference, and I thought the flow of solving was a lot of fun… just like in Star Maps, making progress in one area made everything else a little bit easier, so there was a gradual and satisfying acceleration toward the final answer.

In Space, No One Can Hear You Sing

I believe I was the first on our team to open this puzzle and see that I should submit the word KARAOKE and come do a thing, and if you know me at all, you know you don’t have to tell me to do this twice! We did, on the other hand, have a hard time finding a second person to go… eventually when leadership asked the room, “Seriously, we need someone to go to this thing with Dan,” Jenn Braun said, “Okay, I’ll do it. What is it?” When somebody volunteers for karaoke without realizing it’s karaoke, it’s useful if that person happens to be, like me, a former president of the MIT Musical Theater Guild.

We were happy to see some of our good friends from Attorney (I’m calling them that from now on, see Part 1), and since we’d been told HQ would be able to (and was strongly encouraged to) watch our karaoke game show, I Slacked the room that there would be Rhode/Anand content, and Tanis was very pleased. Erin unfortunately got stuck with a set of songs she didn’t recognize (Rich Bragg had to tell her to stop reading the entire list, as it made the puzzle potentially easier, at least if you hadn’t spotted the aha), and so from now on I will assume that Lady Gaga’s “Diamondheart” is an avant-garde spoken word number. For some reason, I tried to pick my song very quickly once the options came up, thinking that would help(?), and I ended up doing a mediocre down-one-octave rendition of “Feel It Still.” I only actually remember processing the middle two songs on my list of four, and choosing the one I did because I liked it much more than the other. In retrospect, based on the puzzle mechanic, I think “X’s and O’s” by Elle King was likely on my list, and I regret not choosing it because I think it suits my vocal range better. None of this was the point of the puzzle, of course.

Weakest Carouselink

This was the other game show I volunteered to go play, shortly after I woke up Saturday morning when few of our team members were awake. Apparently few teams in general were awake, as we (Todd McClary and I) ended up playing a game designed for eight people with only four. The organizers handled this by giving us four “ghost” players, which made things… anticlimactic? No, antibeginningic, since in the first four rounds, no matter what we answered right or wrong, a ghost was eliminated each time. I survived to the last round of two people, in which we tied on questions answered (I think zero each), and I got eliminated simply by virtue of order placement. Again, as above, not the point of the puzzle.

The point of the puzzle was to take the correct answers to the questions and link them into chains. Todd and I had fun trying to link the answers partially from memory and partially from the cards on our way back to HQ, so when we got there we already knew half of what we needed to do. We got stuck on extraction, but Scott Purdy walked by and noticed there was consistently one overlap that was shorter… I thought this was a bit of a stretch until I processed that the short overlap was literally “the weakest link,” and we read off the clue phrase and answer shortly afterward.

Dance Party

I didn’t get involved in the HQ step of solving this, but I was recruited when the invite to the follow-up interaction heavily implied there would be DDR. As it happens, I didn’t actually play any of the DDR, as Philip Loh came with me, jumped on the dance pad first, and never really needed to leave.

The last step of the multi-phase interaction involved placing a gibberish handout on a turntable in front of a strobe light, which highlighted the answer letters in a very cool effect. Our experience here was slightly awkward because one of the two turntable setups had broken down right before we arrived, and so our handlers were extremely paranoid about our doing anything to break the one version that remained. This led to a frustrating dynamic where they were clearly encouraging us to touch the turntable, but not do this, or this, or this… I voiced my annoyance at this and was probably unfairly impatient, since I can imagine how panicky I would be running this station with two instances suddenly becoming one. Eventually they reminded us they’d told us to bring our handouts but took all the no-longer-relevant ones away from us, which was a nice way to narrow the search space and get us to the near effect (and, of course, clear the room so other teams could play).

Wolf and Rat

Both of these were fairly straightforward, and I co-solved each of them during a period where HQ was a bit slow and we needed more answers to plug into the Safari Adventure metas. Rat was one of the first puzzles I tackled Sunday morning, and a crostic was a good thing to wake up my tired brain without requiring too much creative thinking. Wolf appeared to be a food pyramid puzzle (somehow my eye picked out the relevant lines for that amidst the others) and turned out to be a food-and-two-other-pyramids-I’d-never-seen puzzle. We didn’t get the ordering mechanism at first, particularly since the first batch of letters we extracted was anagrammable without a mechanism, but the last step was cute once we grokked it.

Film Clips

I did nothing to help with this task-then-puzzle until the end. When I arrived in HQ on Sunday morning, the emoji meta had been solved in my absence (and worked much more satisfyingly than I’d expected when I went to sleep, since I’d been finding single-digit changes in the hex codes for the emojis) and there were a handful of backsolved emoji listed on the board next to our unsolved puzzles, one of which was an envelope. There was also an envelope marked FILM CLIPS taped to the board, which we’d received after making a movie trailer and which apparently no one particularly wanted to solve. I suggested that we try the envelope emoji for Film Clips, just in case the puzzle resolved to something like “What this puzzle came in.” We did, and it did.

Magic Railway

Or as we called it, the Harry Potter Konundrum. I have a fairly established rule not to solve konundrums (or generally immersive direction-following puzzles) for my Hunt teams, because I usually get annoyed by them. So it’s a testament to the quality of this one that after glancing at a bunch of the cause-and-effect instructions, I thought it actually looked fun, and I would happily participate as long as I could be Professor Ophidian. After all, my crossword grid nameplate on my office door does contain the hidden micro-bio, “Math Snape.” Anyway, I did enjoy the interplay between characters, but I was kind of disappointed when after two actions aboard the train, I got off the train, and I never got back on. I know from experience that writing these can be exhausting, but I would have been happy for this to go about twice as long… I felt like our group had finally found a good rhythm in interpreting the instructions, and then all of a sudden we were extracting stuff. But it’s rare for me to want there to be MORE stuff in someone else’s konundrum, so compliments to the chefs.

Water Slides

This was a chutes and ladders wordplay grid variant we solved in a group of four that was very straightforward to solve, given its placement. I want to bring it up to say three things about it: (1) Given that the constructors didn’t get to pick the locations of the chutes and ladders (they were taken from the original game board), the grid fill was very impressive to me. (2) While my teammates saw immediately that the player banter referenced the clued words, I initially thought they’d reference the pictures of kids doing things on the originaly published game board. This means I spent way too much time finding a hi-res version of that image, and also that I can attest that those kids are doing weird stuff. (3) In reference to this being a pretty easy puzzle near the end of the Hunt, after Hunt someone on Left Out mentioned that they’d put some easy puzzles right near the end in order to give people a smooth landing. If this was intentional, oy, please don’t do this, constructing teams. I get this principle in the design of, say, mixtapes or calculus exams, but in a Mystery Hunt where you have to unlock earlier puzzles to open later ones, the end of the Hunt is the last place you want to bottleneck teams. If Cascade Bay and Cactus Canyon had opened their puzzles in a different order, we might have had a fighting chance to reach the endgame.

TEAMWORK TIME: Various Subtitles

I was asleep for the first few of these, so I think my first participation was Tug of War, when only five or six of us were awake, and therefore at least three of our players had to be Philip running separate instances on his computer. Anyway, I think these were fun, and whenever they occurred at a time I was awake, somebody yelled out, “Teamwork Time!” and most of the team participated. I’m told these were similar in nature to the Whistle Stops from a recent Microsoft Puzzle Hunt, and whomever you want to credit for the innovation, I think it was a cool social element for Hunt. Based on some of the interactive features in recent GPH’s, I hope/suspect Six Planes Two Words will do something similarly inspired. Though if it could not require me to work out a set of English words that adds up alphanumerically to 851, that’d be swell.

First You Visit…

This parody/loving tribute to First You Visit Burkina Faso unlocked while I was asleep, and I woke up to see everybody poking Ben Monreal (the author of the original, who was remote this year) on Slack. Sadly, I’m not sure we ever solved this, and I don’t actually know if we even cracked the pop culture reference that I now see the puzzle revolved around. Given how significant the original was in the battle between the top two teams in 2019, I think sending it up this year was great, and the particular mechanic used was ingenious. I apologize to Mark and Gaby that you didn’t get to see our team solve this, as I’m sure it would have been the highlight of your entire weekend. (Nothing else significant happened to the two of you during Hunt, did it? Well, other than the release of the Radiohead Public Library, which I’m sure caused Mark’s head to explode.)

Next time: Structure and Big-Picture Design! What was different about this Hunt compared to other Hunts? What worked? What didn’t? You decide! (Well, it’s my blog, so I quasi-decide first, then I let y’all weigh in in the comments. Speaking of comments, what puzzles did you particularly like from this year? Opine below…)


17 thoughts on “2020 MIT Mystery Hunt, Part 2: Some Puzzles I Remember

  1. Our team was equally excited to get to watch you sing karaoke.

    I have not gone back to watch the video since a) I saw the good parts live and b) Laura Royden informed that my spoken word was “actually really bad, sorry.” On the other hand, she apparently knows Diamondheart (I guess it’s on Joanne) so she’s one up on me.

    Also, I was in the group on our team that did the Penny Park Guide and can confirm that it’s great, but yes, quite long. We forward solved it.(Galen went to the bathroom just as we did our last step [which was supposed to be the first step] and returned and immediately declared “Oh, it’s a [SPOILER!],” which the remaining 3-4 of us had failed to see in the 10 minutes we’d been staring at it… and then we got the final answer.) The same group then turned around to the blackboard and solved three metas. Which means, as a result, I missed out on three entire rounds of puzzles because I was working on the Guide.


    • My small team forward solved the brochure, but it took over a day and a few hints. We definitely needed all the subpuzzle solutions for it, so there were a lot of ahas needed. I was definitely appreciative of the hint system for it — I don’t think we could have solved it without it but it was still a worthwhile experience. Definitely a bit surprised that a puzzle considered challenging even by top teams was an opening puzzle!

      I would be curious to learn if the half page of semi-instructions in the puzzle was added after testing to make it solvable (the format of the sub-answers suggests that to me…)

      Agreed on Magic Railway — it was super fun acting out the konundrum in real life, and then suddenly it was done! (And the absolute last step was extra anticlimactic in that it was basically “fill in the blank that has been staring you in the face for a couple steps”.)


      • The how-to-use-this-guide section was in the very first draft, although some of the language in it was changed in subsequent drafts to make the “aha”s easier.


  2. I really enjoyed Film Clips, but that’s at least in part because I’m not a very skilled/experienced solver yet but I’m a pretty good writer, so getting a chance to write a script and then gather a bunch of my teammates into a hallway and get them to act it out made me feel like I actually contributed something.

    Though as far as specific puzzles go, I really want to know what was up with the fish puzzle on Midway! We never solved it, even after we eventually solved the Midway event itself, and the event solutions are not on the solution website. Incidentally, we found the Midway event to be extremely difficult and didn’t solve it until the next day when we eventually got a hint. I was one of the two people from my team to go to Midway and while I really enjoyed it, I found the puzzle part incredibly challenging, although once we did find solutions for several of the minipuzzles I was like “We were SUPER OVERTHINKING THAT, weren’t we.” (Particularly Upside-Down Croquet. I spent a lot of time staring into a thesaurus for that and it turned out to be entirely irrelevant.)


    • For the fish puzzle, you had to catch 9 different sizes of fish. All fish of the same size had the same crossword-y clue on it; the answer to each clue also happens to be the name of a fish (e.g., “Part of a birdhouse” = PERCH). Each fish had a printed “weight”. Solve the crossword-y clue, then index a letter into it by the fish weight. Sort the 9 fish by size and you have a 9-letter answer.

      Liked by 1 person

      • …oh! Okay. The thing we missed was the sizes and the weight not being congruent. That makes so much more sense now. Thank you!


  3. I really liked that the puzzle difficulty stayed consistent throughout. Made it so there was always something approachable for all solvers, plus it felt like there was steady progress rather than us being held up by a few things we just weren’t getting.

    As far as the guide, I thought it was brilliant but it took many different sets of eyes. And nobody seems to be able to do a valley fold as intended.


  4. Dan, I too am sad you didn’t choose “X’s and O’s”! After we’d run 10 or episodes of the show, it was clear that certain songs were being chosen WAY more often than others, so it was always a welcome surprise when someone chose a song that no one yet had chosen. In the end, “X’s and O’s” was one of 3 or 4 songs that was never performed on any episode of the show.


  5. The Emoji meta was a fascinating idea, as was the Safari round. Since you love meta puzzles and dissecting them, I’m very interested to hear your thoughts on them.

    This year had a lot of different elements to discuss: the wedding (I think that was a noteworthy highlight), the automated answer-checker versus calling in answers, the cool down on guessing, the hint interface and its unlocking mechanisms, the “hinteractions”, MIT’s crackdown on overnight policy, the lead teams getting bottlenecked for 6 hours on the pressed penny supermeta, the puzzle that had Oliver Smoot in it, and all of the physical puzzles.


  6. Like a lot of years, this year’s hunt was a bit of a blur. Possibly I was too brain-dead from the start. But as an overall impression I liked the amount of puzzles that involved physical things and activities and not just staring at computer screens. Here are a few of my shouts:

    Star Maps: As the one working on that beginning to end I agree it was really enjoyable even though I wouldn’t have considered it my kind of thing. Took a big chunk of my Saturday and involved a couple waves of solvers coming in and out, but progress was being made the whole time and it was satisfying in the end.

    Fortune Cookies: One I worked on for a while but didn’t finish. After solving the initial clues we tried a bunch of things and briefly came up with the right a-ha, we just weren’t applying it the right way and quickly abandoned it. Facepalmed after someone else solved it and I saw how it worked. Really nifty use of thematic hinting. I suspect the original idea was to try to put them into real fortune cookies (that’s what I would have tried before realizing just how impractical it would be).

    Marching Band: Most years it feels like there’s one puzzle put there just for me to discover; I really hope this is common for all MH players and not just me because it feels amazing. Summary of a conversation between Chris and me during Saturday dinner break: Me: “Whatcha working on?” Chris: “Marching Band” Me: “Saw that, assumed someone would solve it faster than me. How’d it go?” Chris: “We solved the marching bands puzzle but now we don’t know what to do. Some clues have dates and what look like scores, we think they might be football games, since it’s called Marching Band.” Me: “Did you look for The Best Damn Band In The Land?” Chris: “?” Me: “We’ll solve it after dinner.” Looking up one confirmed I was on the right track, and we worked it out with some assistance on the extraction. (I didn’t see the “Home of the Buckeyes” clue.) This one combined three of my favorite MH puzzle features: clever thematic hinting, source material that’s unexpected without being arcane, and encoding a message in an innovative way. Curious how many teams got this.


    • Re: Marching Bands

      A bunch of college football fans were pouring over this one for awhile and getting nowhere, assuming that the scores were game scores for the given date, but not getting it to work. Not being a big college football fan, I was not a part of this. Finally, Joel discovers “The Best Damn Band in the Land” moniker and pulls up some Ohio State marching band videos, but also doesn’t know what to do with that fact. Overhearing The Ohio State fight song causes me to curse Joel out and insist that the puzzle can’t work that way simply because Ohio State sucks. (Go Blue! Also, this is at about 3 am, for whatever that’s worth, when things get punchy.)

      Finally, a few minutes later, Galen and I are asked to fresh-eyes the puzzle. I immediately note that the grid is Ohio State colors (and we never saw the “Home of the Buckeyes” clue either because, like you, we came in after the grid was filled in) and groan. Galen pulls up a random marching band video from one of the dates in question and notes that they make a horn, the matching word. I sigh deeply and roll up my sleeves for some pain — over 30 minutes of pain, since we had to watch some of those videos for multiple minutes until the key formation happened. I pretty much complained the whole time. [Why the hell was The Ohio State doing “Are you smarter than a Michigan student?” during a game against San Diego State? I can only assume the answer is a deep-seated insecurity over knowing that, in fact, on average, none of the Ohio State students are smarter than Michigan students. And if they were to do that at a game with actual Michigan students to compete against, they’d just be embarrassed.]

      Galen, the non-sports guy who once made reference to the “Minnesota Dodgers” because he truly could not name a real baseball team, was the one who figured out what to do with the “scores” since he had no prior bias as to what a football score should look like.

      Beautiful puzzle, truly. Possibly one of my two favorites along with Refreshment Stand. And I fucking hated it.

      … also, no one has mentioned Refreshment Stand. That simultaneously had the most frustrating element (that somehow you were expected to know to call in a text phrase during the emoji round because it was actually an interaction request) but also one of the most satisfyingly beautiful endings I’ve ever experienced in a hunt puzzle. So beautiful that I forgive the frustration from the submission of the first step. Obviously spoilers abound, but this is the video of The US Census getting the final solution: https://photos.google.com/share/AF1QipNaWCHl106qo2NMrNNw2DP7MwWChzF4M-tf0WPh_pQYoRDlj9QsOoC5Xia7GSbagQ?key=SkpFaC1SN0lXSy1OWHFrUDZMamdJN0VnYmh0QXN3


      • Glad you liked both of those! For Refreshment Stand we had fully intended to display a message on the puzzle that mentioned submitting an intermediate phrase. Somehow that didn’t make it onto the live server and we didn’t realize it until mid-hunt. For certain teams (maybe yours?) we ended up either unblocking the submission cue or contacting teams that clearly were trying to submit juice emojis. And thanks for the video… that was great to watch the final payoff!


      • For us on Refreshment Stand, at the time that we had 3-4 possible juice-and-or-steroids emojis in the queue (after the more obvious ones had already been rejected) leading to an hour of delay, Jason noticed that one of the pages of rules suggested that if we felt we were being unfairly rate limited, we could e-mail you guys and ask for mercy. Not only did you unblock our queue at that point, but you gently suggested that perhaps we should just submit the phrase. Which we very much appreciated.

        However, that actually really worried me for other reasons. One of our more… colorful… team members is Seth Bisen-Hersh who is currently soliciting investors for his off-broadway musical, minimum share $5000. (https://lovequirks.com/ if you’re interested — and you’re welcome, Seth.) Seth wanted to call in the phrase immediately and we waved him off. He was being very persistent (which is usual with Seth) and I loudly declared “Seth, this is the emoji round. If the answer turns out to not be an emoji, I will invest $5000 in your musical.” Which got everyone to laugh and then Seth (temporarily) gave up on his insistence.

        But then we got your e-mail and I was worried I was out $5K. But, many witnesses agreed that I said that the *answer* to the puzzle had to be a non-emoji, so I am in fact off the hook.


  7. II&F unlocked Chairiot Races just as we were shutting down for the night. Our remotes knocked it out overnight and a local got the necessary final data on her way in.


  8. I should also say I love the paper airplane puzzle, having gone back and done it. I have always wanted to make an origami-based puzzle; this set the bar high. (The wrap-up showed some hilarious back solve videos people did. Spirit of St. Backsolve, indeed…)

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s