Damn, another detective mystery, so soon after last year—and quite the whirlwind, too.

As usual, spoilers follow, so if you don’t want to be spoiled on hunt content, consider averting your eyes. If I go in-depth on a puzzle, it will be in a dropdown, but puzzles that take ~1 paragraph will not be marked in this way. I will also be referring to several other people on my team; for the purposes of this write-up I will be referring to them by sona name. We are Literally Animal Farm, after all.

Waltz

(If you’re not interested in the Prelude, skip this section—just know that I rank the social hour before hunt about as highly as Hunt itself.)

Hunt is an… interesting affair. Despite this ostensibly being a convention for hunt puzzlers the world over, there’s not a lot of room for social interaction during the hunt itself—everyone’s too busy doing puzzles.

As such, I’m generally a pretty loud proponent of socializing before the hunt (and maybe after, if people aren’t too tired).

I was originally planning on arriving to campus two weeks before, to coincide with the start of IAP. However, some Events™ happened on campus that delayed my return by a week or so. Not that it affected my schedule any.

The Tuesday before Hunt is almost always a Puzzled Pint day¹, and there happened to be some puzzlers in town a week in advance, so that was the first thing I went to. I spent 30 minutes on puzzles and an hour waiting for food, because the Tasty Burger it was hosted at was not expecting this number of people to visit that evening.

Ah, puzzlers—HTTP Error 500-ing sites even in real life. I should have gone to the Shake Shack across the street.

Next came Thursday, the day before hunt, when most people are just starting to arrive. Some people had bunched around the Banana Lounge², so I went there to hang out for a bit, which mostly consisted of watching one of them play modded Minecraft.

I noticed that the table we were sitting at had a bag of letter tiles³ scattered across it. For fun, I decided to see if I could assemble a periodic table from the letters.

We ran out of a few letters pretty early and had to resort to “creative liberties”.

Periodic Table

I flew over to Le Meridien afterwards to visit some more people, chatting with a few of them before they headed off to a wedding. We then played cards with a few members of Cardinality. Before the game, one of them commented on my fursuit, thinking someone on their own team had spent funds on a mascot.

Not the only time I would be called a cardinal that weekend. But I’m getting ahead of myself.

We headed back to campus afterwards, since I didn’t feel like sitting at home for 3 hours. I solved a few past Puzzled Pints in the meantime.

Then came How to Hunt, which I had a semi-obligation to attend as a past member of (TTFKA){6}TTBNL⁴. John, of course, brought All The Energy to the conversation, which definitely made me crack up once or twice. (Also +1 to the cardinal counter from an audience member.)

As soon as it was done, I rushed over to the T Stop and rode it to the World Trade Center⁵.

I’ve been lamenting ANE landing on the same weekend as Mystery Hunt since the date was announced last year. Especially since visiting last year with all the puzzling furries was essentially The Highlight of my entire last year, and also because my fursuit came too late to get it to experience ANE in full.

The sad news is that the official ANE events are all Friday–Sunday, which is entirely within the confines of Hunt. The good news is that this makes Thursday the day when everyone arrives—we could make a Thursday night visit and absorb the atmosphere, which was the best we could do given the circumstances.

So a few of us did. I tried to make the most of it, and came out of the lobby-con-ing with a two-minute portrait sketch, some stickers, and a d20 for some reason⁶.

Magical places, these.

I woke up bright and early Friday morning—a significant change from my usual schedule, as always—and got dressed for the occasion.

Death & Mayhem (the runners of this year’s hunt, henceforth D&M) had sent out something earlier in the month: an invitation to a gala celebrating an engagement. It recommended bringing somewhat formal attire (cocktail dress, whatever that meant).

So naturally, I had to show up in my most expensive suit.

Most expensive suit

Of course, socializing doesn’t stop just because it’s Friday, so I hung around at the local Flour for a bit to meet with people, including a surprisingly large contingent of my team.

At 11, my team broke away to drop some things off at our room. Then we headed our separate ways for kickoff.

Kickoff is normally⁷ held at Kresge, the large auditorium just west of Mass Ave. This year, that wasn’t possible due to MIT starting construction in the area, so D&M split kickoff between three smaller auditorium-like rooms. Each one consisted of a small skit by live actors, different between locations, before playing a video for the rest.

To make a long story very, very short: the engagement party is interrupted because someone stole a diamond, the family calls on a private investigator to determine who stole it, the P.I. hires us, something something puzzles solve everything, also here’s a radio, back to your rooms you go.

So we did… and now we wait. 3… 2… 1…

Nocturne

…but first, a few things to note about the hunt’s structure this year. It was a choice-based unlock system, much like what Brown did last year; you got N keys and had to choose what to unlock from a pool of M ≥ N puzzles. I didn’t really know what to think of it, given I hadn’t been through those particular trenches myself (besides the facsimile MITPC got testsolving for Brown), but I thought it was a pretty interesting twist… up until we unlocked too many hard puzzles to spare keys for the easier ones.

I immediately jumped onto the puzzle titled Unreal Islands, mostly because someone said “Nurikabe” within earshot. I set up a conditional formatting thing and got to work.

Like five minutes into the solve, though, I noticed that one of the islands we got looked like an actual word, mentally filled in the missing letter, plugged the name into Wikipedia… and immediately clocked the theme.

Coincidentally, I had played Alchemia⁸ but two weeks earlier, and promptly used that knowledge to cheese the bottom half.

First solve, 13 minutes. Awesome.

I look around for other things to do. Apparently someone’s already hauled back a physical puzzle of 45 pieces. The people working on it explain the assembly to me—two parts go together down here, a third slides on the top.

I immediately call the assemblage out as a Brass Rat⁹, go on to explain how the rest of the puzzle data supports this, direct them to the site with all the data, and promptly walk away because I wanted to look at something else. Perks of being an MIT student.

When I came back, Sniper (seated next to me at the time) was working on… a copypasta? He had deduced that he needed to send the pasta to other Mystery Hunt teams, in order to collect data for the puzzle.

The problem? We were apparently the first ones on that puzzle. Every team he tried was very confused. He asks me to send one over to TTBNL (knowing I had just written with them last year).

Well, TTBNL is actually two teams now, but I know more people on the casual team, so… oh look new puzzles.

Conveniently, a second puzzle had just been solved, and it revealed an Einstein that I immediately slam a key into (because I have basically become the designated Einstein solver when the team did P&As). I do some small amount of initial investigation—transcribing the names and numbers and whatnot—before I get distracted by something else.

Someone asked an on-site person to knock three times on the radio.

Oh, right, I mentioned a radio in passing earlier. It was given during kick-off, supposedly as a party favor. Turns out it’s a puzzle right from the get-go, and the back of it was secretly a criss-cross grid, which was what prompted that request. But that wasn’t all.

Because it’s also a musical instrument. And the moment it became one is the moment I got sold on the radio.

We soon figured out three of the inputs, used them to mirror the radio’s rendering of Mary Had A Little Lamb… and immediately groaned when we realized the next song was just marginally less desirable to hear at this time.

We pull through the rest (after quite a lot of trouble with shining a phone’s flashlight into a pinhole-sized photodetector) and get the radio to spout out an answer. Whew.

Surely this wouldn’t be the only puzzle that used it, right? There’s too many numbers between π and 2π that haven’t been used…

Anyways, I walked back to my desk, sent our copypasta off to TTINNL¹⁰ like I promised (though a bit late to get the data), and returned to the Einstein, to find it mostly done… save for the fact that the other solvers had run into a contradiction. Welp, clean the slate—time for a do-over.

Spoilers for Drunkens and Flagons

Second run: we somehow run into a different contradiction. The third time around I got fed up with hitting contradictions everywhere and grabbed an online solver to make the consequence deductions for me. This time it passed (without needing to figure out the order beforehand). Apparently someone thought “tequila was the most-ordered spirit” meant “tequila has the largest number of orders” and forgot that it meant nothing relative to things like wine or ale. Too many cooks in the kitchen? Probably.

Anywho, the puzzle now signalled us to go to the Gala to pay off a tab. Conveniently, the Gala was set up on the Stata Center’s 4th floor, which was just a short elevator ride and a hallway away from our HQ—it was furnished to give the appearance of a bar, complete with bartenders, cash registers, and a receipt printer, just because they could.

Now, I was going to start with small talk and ramp up to the actual apologizing. But one of us decided they’d roll a persuasion check instead. It came up as a natural 20.

Well, now we have a receipt. We guessed the drink names on the receipt had to be wordplayed somehow, based on the comment on the bottom. It took twenty minutes of staring at the names to reason any amount of wordplay out of it, and then another two to realize the wordplay method was in the puzzle the whole time.

I soon hear that our team sent That Copypasta to D&M. Hm.

I also catch wind of a bird puzzle. Unfortunately, most of the bird things had been found by the time I found time to hop on, and all that was left was the half of the puzzle that involved non-bird things. I didn’t know that it was non-bird things (given their choice of flavortext was Donald Duck, of all things), and spent forty minutes trying to find actual fictional birds in the grid.

I found something else to occupy my attention eventually: we had unlocked two of the metas by this point. We also knew we were out of puzzles to unlock in the main round, so I started focusing on one of them.

Spoilers for The Casino

We were originally trying to match puzzles to metas based on whether the text given after a puzzle solve used the character’s name. I pretty quickly decided that the text was completely useless for meta-matching, since I found a relatively straightforward mechanic for matching an answer to a poker suit and half of the initially-matched answers did not do that. I then backpedalled on that statement immediately when I found where the number came from.

Great, we now have five cards and an extremely clear mechanic for matching them to the tables (which took some time to figure out). Surely this would be easy… but, well, missing two cards meant the solution was a lot less constrained. I eventually logicked out which two cards we needed to fill in the hole and promptly attempted to find a list of backsolve answers. I found one that fit the ace of hearts, but didn’t submit it to anything, instead trying to solve the meta with the fake answer.

I soon worked out an answer of FACE CARD SHARES.

Now, I didn’t know what a card share was (if anything), but seeing FACE CARD made it convincing enough a pun. Upon it coming back incorrect I decided my entire theory on extraction might have been a little shaky—after all, it relied on a mostly-inconsistent ordering of the letters in each pair. If only…¹¹

I still thought the answer seemed legitimate, though, so I chose to listen in on the puzzle I thought it was the best fit for, just to check. As it happened, the answer was slated to have thirteen letters, and the solvers there had just figured out extraction.

It was for the other card… which really just seemed to confirm the initial theory. Juro then filled in the rest of the pun for me.

I’m bad at words, what can I say. CARDSHARK hardly came to mind (though in my defense I’m more familiar with the non-eggcorned version, CARD SHARP).

D&M borrowed teammate’s modus of interactive cutscenes for these post-meta interactions, and we kept steering it onto the most… interesting dialogue option possible. With that out of the way, though, new round! I compiled a list of backsolve answers and moved on.

We then sunk almost all of our extant keys at the time on the five open puzzles in the new round, Background Check… not realizing that all of them were hard, and especially not realizing that this was probably the hardest round in the hunt.

Not knowing this, I took a brief look at He Shouldn’t Have Eaten The Apple, which had a… rather large amount of descriptions of historical sites. Supposedly. In reality they made no sense in at least two places.

One place I could handle—they gave potential replacements below the lines, I could see places where a few of them could fit, and the first letters of the words they replaced were particularly vowel-dense, which meant they would probably spell something. But even after this replacement they made very little sense—I could know, for example, that the Temple of Bel in Palmyra was involved somehow, but the description didn’t square with what I (or Wikipedia) knew about the Temple.

Too hard for one brain. I left it be.

We’re soon done with another meta, The Boardwalk. After being somewhat sad at the slight lack of keys for what we have now learned is a fish round 42 puzzles long, we unlock two of them. One of them is a physical puzzle. I gravitate towards it.

It’s a bag of popsicle sticks, complete with 40-something dad jokes. Cue groaning.

Not much to comment on besides—well, save for the fact that I wound up doing a dropquote on the sticks. I’d take one minute of asking someone for a sharpie over four minutes of transcription any day.

Well, that’s done. Back to staring at metas. Art Gallery had made some progress (mostly noting the colors) but was now stuck on extraction. So I looked at the other unsolved one (Jewelry Store), which had made no progress at all.

A lot of ideas were thrown around. A lot of the answers seemed to be incredibly specific things, like two answers related to bulls—maybe semantic chaining was on the table?

Several semantic sh-tposts¹² later, someone tries chaining orthographically instead. It works.

I blame the wraparound mechanic.

This made three unlocked rounds, which exacerbated our key problem slightly—our distribution was now something like 5 on Background Check, 2 on the new round (Paper Trail), 1 on Stakeout, and 1 still stuck in the first round.

How many did we start with? Also 9. The width has not increased.

I did note that one of the newly-unlocked puzzles claimed to be the World’s Largest Crossword; having missed World’s Largest Logic Puzzle two years ago, I was excited to take a crack at it.

Before I started, though, something on-site demanded my attention.

Three hours earlier, I noticed that we had unlocked a Fermi estimation challenge as a quasi-event-y thing. It said MIT knowledge was useful, but I considered myself pretty bad at actually doing Fermi estimation, so I didn’t go the first time.

Well, newsflash: MIT knowledge was very useful. Like, every question useful. And apparently a few of the questions were actually Geoguessr problems. Also the result you got back was overly abstract unless you got a very accurate number.

So I went to the room and guesstimated to the best of my ability. At least, I hope my guesses were saner than the people in the other corner who kept guessing multiples of π¹³ for an obviously integral number of things.

Turns out, I still suck at estimation. But I knocked three of the four Geoguessr problems out of the park. (They eventually got someone who was very good at album identification to clean up the data.)

Back to HQ!… still the same nine puzzles, so I just go back to what I was doing. By now people had finished off the crossword clues, and it was just about trying to fit them into the grid.

Spoilers for World's Largest Crossword

And just like WLLP, it was also secretly circuitry. Well, I say secretly, but it had overt “Zero” and “One” clues (which we hadn’t solved at that point), not to mention that the puzzle literally looked and acted like crisscrossing wires.

The grid was dissectable into sections. Someone had already categorized the larger sections into 36 types (even noting that two of the categories were direct opposites of each other), so I went for the smaller sections, which I surmised could be filled one of several ways depending on what was being carried.

We spent an hour trying to fill out the components, only figuring out what the 1 and 0 clues really were once we had filled one in. Another hour later, and I had all the components… and promptly discovered that the second layer was a Kakurasu.

Another half-hour of twiddling and we figured out what each of the major components did, and someone promptly solved the 6x6 border-drawing puzzle. Yep, definitely WLLP-coded.

I honestly feel like this puzzle could have done without the long and clunky flavortext. It works well enough without it (in fact I mostly disregarded it while solving), and not having it would serve to emphasize the grid (and whatever feelings it evokes) more.

While doing that though, another group pulled me aside to stare at more birds. (Another bird puzzle? In my Mystery Hunt?)

Of course, we gravitated to identifying the cards immediately, then looked at the songs, only to find that we… may have done that step in reverse. Identifying the songs would certainly have made the birds easier to find. We then got stuck on the extraction step for all of twenty minutes until the other bird in the group spotted the right thing.

I swing around all the unlocked puzzles again. By now we just have one meta left in the main round; clearing it would unlock another round. Not that that helps matters, since we ran ourselves out of keys ages ago (even with the three-key boost we got from the event), and metas didn’t give any… but regardless I opened the spreadsheet to take a gander.

It had been open for nearly seven hours by this point, and the other people on the puzzle had only just found out that using the values they got as ASCII gave them letter-like shapes. (Not letters—those have an immediate tell for which ASCII values they can occupy—just letter-like shapes. The dataset was probably too restrictive for just letters, but still.) And now everyone was prepared to spend another seven corkboarding theories for the sort order. (Notably, the letters didn’t anagram to anything good, so the hypothesis was a tossup between “missing a letter” and “non-answer nature”).

One look at the answers later… oh. There’s a length ordering that everyone missed somehow.

Fresh eyes save lives. Unfortunately we also timed ourselves out with our three million anagram/transaddition attempts, so we had to wait until 2330.

More on the fourth round later; we realized that we needed to start on the mid-hunt runaround basically immediately to finish it before the end-of-day period kicked in. So a large group of us went up to the Gala, had the radio tuned to the right station, and proceeded to get absolutely lost around the Building 13 segment of the chase.

Just for fun, I did this part in suit. I tripped over my own feet a few times. Worth it.

Eventually we hit the ending near the East Campus construction site. When we got back, I took the map we grabbed earlier and quickly marked off the clue locations and order on it. We then left, because it was half an hour to closing time.

1.5 hours and some patently stupid happenings that I do not wish to recount later, I was back in my dorm, heating dinner and hashing out the remainder of the metameta. We soon discovered the building mapping, but had some trouble getting it to stick (particularly for the Building 13 area, which we later realized was because the paths). Eventually we had the right arrangement, though the fact that the extraction reading in the expected forward direction started RACE… threw us off for a bit.

With that done, though, I went to catch my six.

Mazurka

I decided not to bring the feet this time. Partly for mobility, but mostly because I knew the weather was going to cause some puddles on the ground.

The overnight squad solved through the slow trickle of fish puzzles. They had also requested a hint on the one puzzle we had left in the intro round, On The Corner, in the hopes of eventually getting its key back.

I knew about the troubles with the puzzle, of course—they had called me in once or twice on the sticking point, because it seemed to be directly connected to MIT. And now I have the benefit of hindsight.

Spoilers for On The Corner

When we opened the puzzle yesterday, I was asked about the covers. I did realize from reading the rest of the puzzle that an MIT periodical was likely involved, and Technology Review specifically was namedropped (obliquely) in one of the subproblems.

I looked through the covers and couldn’t find a tight fit for either of the covers obscuring the first four problems. Someone else did find one of them, though—the Tech Review for this month. The other one, however, was elusive—none of the recent issues had an orange bar up top, and looking sufficiently far back revealed that none of them did.

As it turns out, it was the Alumni News magazine for that month. And the thing about that magazine is that it’s hosted on the same website as Tech Review, and one is given much more billing than the other. (Alumni News is like halfway down the front page, or stuck in a corner link for the issue page, and additionally goes completely unclued on the puzzle page unlike its more prominent cousin.)

The problem? The data we needed was specifically a subpage of That Other Magazine (requiring an even deeper dig than that). Which meant that first cover, and everything else that clued towards the Tech Review, was doing f-ck all for the puzzle. In fact, by implying there was something important in that issue when the opposite was true, it was entirely counterproductive to the solve path.

Props for making those puzzles a Mystery Hunt dataset under Dan Katz’s nose, but also clue your dataset better.

The hint proved unhelpful in actually directing, choosing to focus us on that first magazine. We wouldn’t solve it until 11.

Interestingly, we were also very close to backsolving it, having reduced the options to B_{6} FOREST. We distrusted the first sensible Nutrimatic option for being extremely out of left field for a puzzle about MIT when other options seemed like better fits for the meta, and tried every other sensible option instead. It was the first sensible Nutrimatic option.

While that was going on, though, I plugged a key into a puzzle whose title was in a Brahmic script. I’m a simple bird. I see title in interesting foreign script, I click.

Spoilers for ಕಾಬವದೋೀ್

And fortunately, I am enough of a keyboard nerd to instantly see the puzzle was about keyboards just from looking at the Hebrew text.

Exhibit A

The mechanic was quick to identify from there (a mechanic I was all too familiar with), and we soon had a cluephrase. (My custom keyboards are keyed to my own idiosyncracies on the QWERTY layout, so I had to pull up the keyboard settings previews on my Mac to get layouts that were actually used.)

Unfortunately, this cues directly into my first major screw-up: I misidentified the title’s text as Telugu, rather than Kannada¹⁴. So I had the wrong keyboard for a while. Eventually we fixed it, and I typed the result into Google Translate to ask its transliterator. “Ske-cha pyā-ja”, it says.

Three parses of that result and zero correct answers later, we get locked out for a bit. I go back to double-check everything. Someone submits a clearly incorrect answer in the meantime. I ask for a pause on answer submissions while I double-check everything. I do eventually find a few errors in what we wanted to type, and it comes back with a much clearer “ske-ch pyā-d”.

I go back to the puzzle to submit it. Someone had submitted nonsense in the meantime, and we were locked out for another few minutes.

Needless to say, I lost my shit. Enough to uncensor a swear word or two in Discord chat, and almost throw the bell we were using to celebrate solves against the floor (after submitting the Actually Correct answer).

I bounced around the puzzles again. I don’t really remember what I worked on during this time; I know at some point I led Juro over to the Borderline murals beyond 66 for the puzzle that wanted them.

When I got back, I was cued into a puzzle with one long poem and some other encodings that had already been cracked. I noticed that the big poem was actually a bunch of interleaved smaller poems, only to find that it clued a step that had already been done. Welp.

By the time that was done, we had unlocked the Chinatown meta. And for… obvious reasons… I got stuck on Geoguessr duty. I’d seen enough of main campus to get most of them with reasonable amounts of accuracy. The ones I didn’t… well…

$[2025-01-18 1255] The duality of Geoguessr: sometimes you know it's this part of the country based on the color of the sky and what car Google used but sometimes you can't get it any narrower than \"rural US road\".$

Eventually I’d piece together that some of them were in approximately the same location—the three at the Chord in Building 2 did enough to confirm that.

Some of us in fact went to the Chord, and found nothing of particular interest there.

Now stuck, I swapped onto a bunch of different puzzles again. First I tried A Map And A Shade (Or Four), which we knew the basic gist of (four color map theorem) but kept running into contradictions when trying to apply it to a world map, no matter what we tried. (I would later learn, after it had been solved by others, that it was a map of the US, to my dismay at not realizing this sooner.)

I then peeked at Absolutely Not Balderdash. Recognizing each set contained a surprisingly uniform distribution of rare letters, I guessed it was Scrabble-related and took a letter count of each to confirm this theory. I then answered a few clues and then realized those, too, were scrabble-distribution.

I set up a thing below the puzzle to track it and left it to the other word puzzle people. Occasionally I’d tab back in to fill a few words.

When that was done, I briefly took on Lab Scrabble: a short 10 minute affair while I waited for other things.

When we were done, a different thing came, in the form of Given Up.

Spoilers for Given Up and also Chinatown, interleaved

Of course, I had to take this one as the MITGuessr person. (This makes three puzzles I’ve had to do for this purpose now.) Fortunately relatively easy: I knew 6 of the 9 locations offhand, and several candidates for a 7th.

The last image piqued my interest, though, because it was a picture of the radio face. Immediately I asked around for the radio… which was noticeably absent from the room.

Turns out we had unlocked a different radio puzzle earlier—the one asking for a performance of the Star-Spangled Banner—and the resident radiohead (Grackle) had seemingly disappeared to some other room to practice.

Welp, leave a ping and put a pin in it for now; Chinatown had made some progress in the meantime (due to incidental airwave interception, no less), and we’d sent some people on campus to investigate. Apparently the relevant information was stashed under the benches? Who’d’ve guessed¹⁵.

Eventually the radio came back, and I took it to the first location (conveniently just outside our room). I flipped it to FM (as the puzzle indicated) and began to search… and did in fact hear something unexpected.

It immediately clicked.

I went back to grab my computer. This was going to be a long day.

When I got back I saw someone I knew doing the exact same puzzle; good to know we weren’t that far behind, say, ✈️✈️✈️ Galactic Trendsetters ✈️✈️✈️¹⁶.

I took note of the station I heard The Song on and moved on. This continued for a bit until I couldn’t get it to work in Building 13, moved outside to get the one outside Building 12, and wound up getting caught in the rain.

As I found a place to sit to record the data, the song was ending, and I expected it to start up again… only to be met with a recorded message.

So now I had to backtrack, listen through The Song (which could take up to three minutes!), and then listen for important details at the end of The Song. Which details? I didn’t know! I got unlucky with my cycle timing, so I had to endure the whole three minutes basically every time. If I were smart about this, I would have recorded the message with my phone, or brought a co-solver with me, but I was already juggling two devices and it simply didn’t occur to me that I also had a recorder on hand.

All told, I spent 2 hours on campus¹⁷, met with the Chinatown contingent twice, perched at the Eastman Lobby for a bit to clear my brain (and help out a bit with the Chinese characters in Chinatown, though they had that well in hand), and ended up having to backtrack a few times to catch the details I missed before I knew there were details. The Chinatown group actually found one of the images I couldn’t (the Glass Lab in the basement of Building 4), so I went for that, backtracked a bit more, then landed back at the room to take a bio break.

Even then, I still didn’t have a sort order. I knew the text was important somehow, but I had already spent 2 hours with the radio.

In all, it felt like a puzzle that didn’t respect solvers’ time. And I already have very strong feelings about puzzles that intentionally waste solvers’ time. Hell, even Hunt HQ seemed to encourage it, given the hint they gave on the puzzle was an uncharacteristically unhelpful “try looking at the puzzle”. Twice.

I got back to our room at around 7:00, very annoyed. Sniper commented that it was a “Never Gonna Give You Up Run Around”, emphasizing the last two words (which happened to be a lyric in that song).

I could only respond to that sting with “This puzzle let me down. Hard.” Which was essentially an invitation to get barraged with every other lyric in the chorus.

When I got back, I noticed that our team name was suddenly not Literally Animal Farm. Sniper told me not to worry about it (it was a puzzle). Moving right along, then!

I passed the radio back to Grackle for their performance. In the meantime I figured out the last location I needed for Given Up (after unintentionally barging into an HQ-only room—whoops!) and grabbed the radio immediately after the performance to take notes on the two I was missing.

I still couldn’t solve the puzzle by randomgramming, though—both because I recorded a bad letter somewhere along the way and because I didn’t account for duplicates.

I swapped back into Chinatown—by now all the things had been found, and it was just parsing a large mess of Chinese radicals and messages headed by bird names. (Three bird puzzles? In my Mystery Hunt?)

I start looking the birds up. A few of them have Chinese equivalents that looked promising (all featuring the 鳥 radical, of course). Alas, that pattern quickly fell apart.

A lot of spitballing later, I get tired of the discussion. The puzzle with the Japanese title just unlocked, so I join in on the group doing that one.

Spoilers for 皇帝の暗号

By the time I hopped in, they were mostly done with the identification, complete with the Iroha alphabetical order for Japanese. There did seem like a few inconsistencies with the answers we had, though I chose to ignore those for now; if one worked, we could tweak the others to fit.

We did eventually spot a few Caesar-pairs in the English words. A few refused to cooperate, though, and soon there was debate between organizing the words as a 12x2 and as an 8x3.

The Japanese words were short enough to make 8x3 the better possibility, though, and I soon demonstrated an example chain (while Grackle worked on automating a Japanese Caesar cipher).

We eventually got 7 out of 8 characters (only failing to identify a good word for “inferno” to complete the set), and figured out the puzzle was probably asking about the end of the Heisei Era. But what to put in as an answer?

We tried a few sensible things (2019 for the ending year, REIWA for the era that came after), neither of which worked. And we soon had to put this on hold, because…

In the meantime, we also unlocked a scavenger-hunt-like thing. We soon scheduled an interaction for ~9 PM, and met with the judges at the east end of the Infinite Corridor. We were told we had an infinite amount of time to do 14 tasks. Easy!

Except, just ten minutes later, we got a message, and promptly realized they meant to capitalize the word “Infinite”¹⁸.

We gathered at Lobby 10 to strategize. While there, Grackle and I checked the chat for 皇帝の暗号—someone was asking if anyone had just submitted a translation of the cluephrase as an answer. We said we were trying to parse it as a cluephrase. They said that the translation matched the enumeration at the bottom of the puzzle.

…wait, there was an enumeration?

blind birds

We soon finish strategizing and walk back to the room to work on puzzles in advance of the next appointment at 2300. I was kicking myself the whole way for not noticing. Grackle stopped by the piano in Building 26 and spent the time between scavenger hunts playing it.

Despite that minor setback, I enjoyed the puzzle as a whole, and it did basically everything right with respect to the use of the language. I couldn’t ask for more. (Why am I making this comment? …you’ll see.)

Off the language high, I tapped onto the Tunic puzzle, now that I have actually played the fox game and can help without spoiling myself. I’m also pretty quick at translation, so I solve the clues quickly, point the actual fox working on the puzzle in the direction of the manual… and then get immediately distracted, because it’s Scavenger Hunt o’clock.

The strategy was to pair up based on tasks, and speedrun going to the locations (since that particular speedrun was easier to do in 10 minutes). I got the dog task, and might have tried to sell the bit by making my “dog” play fetch with a water bottle.

This time we had plenty of time to spare, and got a certificate. Which I then solved in the time we took to get back.

The key went to a new puzzle, and the moment someone said the word “Lingo” the entire furry cluster on LAF had jumped on the puzzle.

…OK, we kind of knew a Lingo puzzle was coming just from incidental commentary. Between collective word knowledge, personal knowledge, and just cheesing the resistor-reading step (once I knew how it would work) by measuring ratios, it got torn apart in 40 minutes. In the meantime, the others had finished off He Shouldn’t Have Eaten the Apple, which resulted in a message being blared over the radio. Hearing an interesting historical tidbit, I looked into the answer, to realize it was the same as Bukhansan—a name I recognized from somewhere, but wasn’t sure where.

Back to Tunic—Ryker (aforementioned fox) had finished the manual step at this point. I started writing the answers out in the hexagon script. Some looked like they were a geometric transformation of another… or at least suspiciously similar to such.

It made sense. Not enough sense, though, given we couldn’t find matches for a few of them. It was almost certainly right, though, since it also created a mechanism to convert our nonsense syllables into sensical ones.

It was getting late, so we packed things up again. This time, when I got back, I chose to focus down the extraction on Chinatown. There’s a lot of unused information—one answer in each set of three is just the radical, for instance, while the other two have other things attached. There’s also the message text we haven’t used. Surely those things are important!

Bad language usage

…aaaaaaaaaAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA——

Romance

One ordering later, we had our second round done. I promptly went to sleep, because 4 AM, but also because I got annoyed at the Chinese puzzle being assembling Chinese characters entirely in service of an English translation. That’s an entirely separate essay, though, that I have not put to paper.

I got back to campus at 9:30, investigating on-site things for Sounds Like a Dodo to Me on the way to the team’s HQ (we never ended up actually solving it), and arriving just in time to see a physical puzzle I had been waiting for. I had seen other teams playing with their copies outside their rooms, just enough to have an idea of what the puzzle was about.

Good news: my instinct was right, and I even guessed the right idea for identifying the other locations. Some time afterwards I also recognized the PLANET B poster and went down there to check my idea (and was mildly amused by a different team using the third floor of the Infinite Corridor to lay out their copy of the same puzzle).

Bad news: I am sufficiently bad at guesstimating that my first attempt to find the Swedish location ended up with a search area the size of Luxembourg. (The actual location was 8 miles away, and very much not visible at the resolution I was using.)

In hindsight, probably not a good idea to start with the largest scale, in a puzzle where precision is everything. Live and learn.

As I was going to measure another one, though, Ryker came back with a hint for the Tunic puzzle we got stuck on last night. The moment the word “cube” left his mouth, I gave myself the hardest facepalm I could muster.

Somehow, somewhen, I forgot that the Tunic script was actually written on the edges of a cube. I had internalized it as a hexagon with spokes¹⁹ (the projection of a cube) for so long that it never occurred to me that the way it was designed was to reflect Tunic’s own isometric perspective.

Afterwards, I chose to refocus my attention on doing extractions. Taking on Beyond a Shadow of a Doubt, I picked up a tentative attempt at extracting something and confirmed it by finding the colors in the text. I had extracted “YOU PLAYED IT” and expected the answer to go further, so I started grinding through the last dropquote for more colors. Midway through, though, I saw the answer appear in the top left. Turns out that was it. Whoops.

In the meantime, Cardinality had found the coin. Which meant we needed some forward momentum if we wanted to cross the finish line by 2200.

Fortunately, that extraction was enough to get us to unlock the MITropolis meta. I refocused my attention on unlocking metapuzzles, and quickly extracted _land as well—enough to unlock the three metapuzzles in Background Check.

Then Grackle asks me to help with Cacciando Trio, which I had already gotten the gist of but hadn’t actually bothered with the actual transcription. Pulling it down to Audacity, however, meant I could skip live-transcribing the Morse. I then quickly did the Braille as a mostly-live transcription, before starting to brainstorm a method for the part now revealed to be notes in 26-EDO equal temperament.

While working on this, though, I get pulled aside by Juro—he wanted me to go print the 48 pages we currently had on the meta for Murder in MITropolis, which we now knew was structured vaguely like Cain’s Jawbone²⁰.

See, I’m also the token MIT student on the team, which means I’m one of like two people able to use the printers in the area²¹. Off to Stata I went, then—my original plan was the nearby EECS lounge, which I misremembered as having printers.

I had just gotten settled when the rest of the group finished Cacciando Trio, and Juro then bugs me for the next three pages. Fine, another five minutes.

20 minutes later, the same story (with a different puzzle, of course). Fortunately it’s the last set of pages, which meant I could start focusing down the tail end of Paper Trail.

Spoilers for the Paper Trail metas, and specifically Shells 4 and 5

By now, others have solved both Shells 1 and 3, and were now pretty confident in a Catch-22²² situation: the metas of this round could also take meta answers as feeders.

Our next attempted inroad was on Shell 5, which we had now deduced to be about lunar features (and specifically maria) after plugging in the answer for Shell 3. Everyone else starts brainstorming about maria that could fit the blanks; meanwhile, I was looking at the extraction at the bottom and attempting to deduce the answer knowing nothing but its letters’ topology.

I decided that BOAT was the only possible answer that fit the topologies; of course, one try in the answer checker was all it took to disprove that theory.

I swapped to Shell 4, which we knew was a Chinese Remainder Theorem problem, and which had by now gotten a tip to get the glass to spell PURE. Unfortunately we only had two answers for it, which left a lot of possibilities. I listed them out anyways.

Eventually we worked out the answer for Shell 5, which I promptly used to finish off Shell 4, which was then promptly used to backsolve Shell 6.

In the meantime, we had finally cracked Papa’s Bookcase, the meta for Illegal Search. Illegal Search itself was a creative concept—the round was structured like an escape room, complete with all the locks (and puzzles that needed to be solved for the answer to those locks). I didn’t manage to interact with the mechanic that much.

The Bookcase, as it turned out, was… also such a lock, revealing a whole other room with eight whole puzzles to solve.

I suppose this is revenge for the Nashville round Sunday surprise from last year.

I took a brief look at the new puzzles. None particularly caught my eye; I did take a look at Fechtbüch, given the heraldry terminology, but I soon guessed that the puzzle was just using heraldry as obfuscation for Something Else (an obscure sport of some kind?) and went back to Paper Trail.

Back to Paper Trail.

I soon found the right arrangement for Shell 7, leaving just 2 and 8 before the final. I had by now figured out how to parse 2’s instructions, but could only find one answer to slot into the meta. I tried to find good numbers to slot in, and deduced some properties of the Y numbers based on the goals.

However, Shell 8 was closer to completion, so I just hopped onto that, realized the extraction was spelling state nicknames, and backsolved Shell 2 before submitting the extraction. Shortly afterwards, I found the other answer to slot in… and it was an answer I hadn’t considered possible, having discounted L as a valid leetspeak letter²³.

I held off on helping the metameta for the moment while I did extraction work on Passage of Time, but swapped back on afterwards. By now it was 1930, two and a half hours to Hunt HQ closing. Basically every puzzle was either solved, was well on its way to being solved, or had a “clue” (free answer) used on it, so it was just meta duty. Given the Nashville surprise and our lack of progress on Background Check, we probably weren’t getting a runaround.

Back to Paper Trail.

Naturally, we thought the way the meta answers fed into each other would be relevant. How, though, would elude us for a bit.

Our initial theories revolved around word transformations. In particular, the book on the round page for Paper Trail was suspicious: eight bullet points for eight metas, and at least one of them looked like a wordplay transformation that could very well be used on two of the feeder answers! Surely this is right, right?

Well, news from Hunt HQ: it’s not right. In fact, the entire book is irrelevant to the meta.

Fish

Some more prodding eventually directed us to the actual interpretation, and we eventually chalked out a directed graph on the blackboard. Well, most of one, anyway—we backsolved Shell 6, so we had an incomplete picture of what fed into it.

Obviously, the answers wouldn’t fit directly. I did notice, however, that starting from the second letter might just work (justifying it as a first arrow leading into the shell from the feeder), and managed a decent part of a subgraph before… running into a contradiction.

OK, wipe the board, try again from a different starting point… aaaaand also contradiction. The two graphs I got both managed to overload Shell 7, and I couldn’t find a satisfactory explanation. Not to mention they both had the same arrows in a surprisingly large number of places…

As I was about to wipe the board and try again again, an emergency communiqué from the group working on Shell 6: they had found the last arrow. There was already an arrow pointing from 7: I expected the last 6 arrow to come from elsewhere, and completely redefine the structure of the graph in a way that would be unintuitive to derive from my current construction.

Against all odds, it was the double-up.

Completing the graph, and defining the trails to the hidden company (which I called “bailouts” for… no particular reason), was easier.

Paper Trail

Less than an hour left. But just five minutes after finishing off Paper Trail, we also solve the second Illegal Search meta. Of course, Nashville surprise (again, again), this time featuring alternate extractions from the five post-lock puzzles (flavored as shining a blacklight on the relevant objects, another escape room trope). I fast-tracked the second extraction for Jargon; the others fell in short order to either a cursory glance or the force of a magnifying glass thrown at Mach 2.

So by rule of three, this must be the last meta in the round, right? Maybe we can make this a four-round solve before Hunt HQ closes.

Unfortunately, even with the telephone message that we knew was coming for Chekhov’s gun reasons, we still didn’t manage to find the right books in time. I even attempted a brute-force approach (that ended up not panning out because I forgot it was indexing into the answers).

We started packing up. Fifteen minutes later, the group working on The Killer comes back downstairs, having managed to solve it during overtime.

Galop

It had started snowing outside.

I quite enjoy snowfall, having generally lived in places that do not get snow naturally²⁴. The first Mystery Hunt I had in person was the one we won on Monday morning, and it had just started to snow when I stepped out of Lobby 13 and started trudging back to Le Meridien with a box full of puzzle pieces.

This experience means I’ve also come to read the first snowfall during a Mystery Hunt as when “nature” expected the hunt to end. Of course, nature does as it wants, and the weekend of hunt is not subject to the whims of the CPW weather machine²⁵, but it still holds personal significance—essentially, the snow defines a shift to an “overtime” mentality towards hunt.

This was true in 2024, of course, as I’m sure at least three thousand hunters would be all too happy to remind me about. (The snow fell at 6PM on Sunday.)

Either way, I pulled out an umbrella to keep the snowfall on the suit to a minimum. Right as we got to Mass Ave, though, Grackle asked if I wanted to winter at their AirBnB.

I’ve lived alone in my dorm room for about as long as I’ve been on campus, so post-hunt socializing tended to be restricted to the post-wrapup period, when people were already booking it for a flight four hours later. So this was a social decision as much as it was a decision about spending five-or-so minutes less in snowy weather. It did mean I would be skipping on personal hygiene for a day, but a day is basically nothing.

While this would have been a good way to wind down, I could never truly get the overtime mentality out of my head. We were pretty close to closing out the hunt, after all—just two rounds and four metas. How hard could it be?

So while the rest of the group was listening to… something I can’t even begin to describe, I pulled out a computer and started hacking away at them, focusing on Papa’s Bookcase for now. Since Hunt HQ was closed, I had to make do with asking the nearest possible source of hints, which in this case meant Taph (a friend on ✈️✈️✈️GT✈️✈️✈️ who we had invited over).

Well, the linchpin in the solution was load-bearing flavortext—we somehow never read too deeply into the word “audible” (which did double-duty this time). OK, so how are we going to identify the audiobooks?

I spend several minutes throwing the answers into Audible hoping it would turn up something possibly usable. Of course, nothing turns up. Turns out the answers took another layer by instead being movies starring particular actors. Why this when every answer is a box-standard phrase and otherwise near-unrecognizable as a movie? Why this when one movie could have like a dozen stars? Questions I can’t really answer. But at least it left the puzzle one reasonably-sized complete search away from a solve.

Which just left Background Check. The team had found the main gimmick of literally checking the background and solved The Mark before overtime. Oversight (once I had seen the chat message explaining that we had to find helpmates-in-1 instead of an oversight on Black’s part) was relatively easy to complete, which left Grand Illusion and the metameta itself.

Spoilers for both

Taph had left by this point, so I asked some hunters who were online at the time—few and far between, since it was 3 AM—for a tip. One chose to cue me into the word “check” being important to the theme for all three. Which lined up for chess and poker, but I couldn’t think of something for this one. Baggage checks? Was this secretly about IATA codes? It did look like a map…

A different person then showed up to ask me about things I knew in the world that sounded like “check”.

I definitely needed that clue. It then took a bit more back-and-forth to confirm that, yes, I needed to overlay it on a map of Czechia—I tried this already, and could see the coordinates of the star relative to the gridlines lined up with the decimals in the coordinates of Prague, but that left most of the cities outside Czech borders. Including one far off in what I guessed to be Ukraine…

I didn’t actually overlay the map—I was calculating the coordinates by hand instead²⁶. So I guessed the Ukrainian one to be the formerly Czechoslovakian city of Užhorod, which prompted me to realize the extraction mechanic. (I then double-checked the coordinates to realize the city was actually Lviv.)

While I was doing the math and recounting the history on this, Grackle (who had stayed up to work with me) had read the metameta’s flavortext, instantly knew what to do with the image, and then did it for the two minimeta answers we had. I joined him for a bit to see if we could identify a console that worked with Grand Illusion. We couldn’t, so I finished Grand Illusion off (using the console list to check that the answer could work).

Parsing the new shells took some time. I tried Oversight first, but it didn’t look like the chessboard was nearly filled enough for the answers to work. I then looked at the Grand Illusion one, and we went back-and-forth over whether the dot we had found on Carter’s suit was a 9 or a 3. It definitely looked more like a 9, but we hadn’t found a 3 to contradict that theory, and in any case we didn’t have 9-letter answers to spare on that puzzle.

It was now 6-something AM. We were tired. I set an alarm for 10 and fell asleep on the couch.

When I woke up, a third person was offering hints. Another back-and-forth, and I finally had a reason why we didn’t have an answerlist that fit.

Because it was the wrong answerlist. Most of the answers were AKAs, and their other names fit the new shells.

I had noticed the property for some of them (having looked at Samgaksan when Adam’s message came blaring over the radio) but assumed the meta constructors needed the weird letters for extraction (reasoning that if they needed the semantics, they would just use the Wikipedia title). By the time the metas actually came around, I had forgotten this, and I didn’t really work on those metas enough at the time to devote headspace to the idea.

I took the opportunity to confirm that, yes, there were 9 letter-slots in both Oversight and Grand Illusion, and found a bunch of the alternate names.

It’s crazy (positive) that they made this work. But it’s also crazy (negative) that they didn’t really leave breadcrumbs on that particular trail—the most visible is the already-overloaded title and flavortext, and there’s a lot of words that do double-duty there.

By now, though, it was 10:30, which meant A. wrap-up was imminent and B. the group had to check out of the AirBnB. This would have to wait for afterwards. Meant no finishing, but I wasn’t expecting that at this point.

I sent a message to the chat for what was almost certainly the last major idea needed to solve the puzzle and headed off—first to Flour (where at least one person congratulated me for supposedly winning the hunt), then to the wrap-up room²⁷, then just outside the wrap-up room to hang out with people, then showing some people on my team to the rhythm game room in the basement of Walker Memorial. After saying my goodbyes, a stopover at the wrap-up room again to grab my umbrella (and having to decline another congratudolences from D&M), then a 20-minute long trek back to the dorm.

When I got back, the group had gotten some much clearer images, re-solved the Mark, and was now working on the Oversight. I jumped in to close it out with Grand Illusion, and the final answer.

End scene.

This hunt had a rather higher variance than my past hunt experiences. If my 2022 were a dromedary²⁸, a mostly uniform experience²⁹, 2025 would be a Bactrian: good experiences were generally extremely good, bad experiences tended to be extremely frustrating, and both were doled out in approximately equal measure (though take this with a big grain of salt; I didn’t particularly emphasize the kind of experience they were prioritizing). If their goal was to make a memorable hunt, they certainly succeeded at that.

It speaks to the effort, though: high-effort things tend to be more polarizing. Which… well, explains the existence of Discourse™ better than anything else I can think of.

Content warning: opinions

(For the last time, Adal, you should probably type “Discourse about Discourse”, because Discourse is bullshit. You’ve been over this before.)

It’s hard trying to think about this. Even ignoring the hunt I wrote for, the fun I have with hunt seems to be (mostly) positively correlated with the supposed length of the hunt³⁰—except really the fun I have with the hunt seems to be unrelated to length and more related to how much I liked the puzzle content (or, y’know, the limited slice of puzzle content a given person interacts with). Then again, “length of hunt” is a crazy variable on its own, defined entirely by the first team to finish. People will call it a Sunday noon hunt despite most finishing teams wrapping up closer to 10 PM, as if Cardinality didn’t just blow everyone else out of the water this time. And I know people who will specifically cite that timestamp as a reason for their positive hunt experience, as if the 100 other enjoyable things D&M actually put into the hunt experience have about the same worth as a single fucking number.

Again, one number is horridly reductive, especially such a random variable that is, ultimately, semantically vacuous. The emphasis on this number devalues the effort put into the hunt.

Like, obviously the writing team cares about The Number—they’re the ones interfacing with MIT for the most part, and the ones who have to deal with reserving space on campus. But we shouldn’t. What the writing team wants to do is their business and theirs alone. In general, I trust that they have the best of intentions for the future of hunt, and have designed accordingly. I can tell that D&M wanted a better in-person experience and a better MIT student experience, and I’m glad they could deliver that where we (and our size-1 MIT student contingent) couldn’t.

If you want to stump about it, go win the fucking hunt first, and express your opinions with the hunt you write.

The first lesson is I need to get better at reading flavortext.

The second lesson is unfortunately one I think could apply to puzzlehunts in general, and that's how aesthetics interacts with puzzles.

Death & Mayhem went all-in on aesthetics this year, especially the in-person aspects. This manifested in a few ways: most of my experience with this in-person was from my two visits to the Gala and the radio playing music and the occasional weather report³¹ in the background, but they also spruced up the website with environmental storytelling and stuff. Things like the puzzle icons in the hidden room in Papa’s Study tying Baby to the prologue or photos of Katrina showing her police badge, well before the respective meta answer spells it out. All in all, it seemed like they really wanted you to be invested in the plot (a reasonable reaction to the fact that the plot was a lot less present for 2024).

I did like what I saw; perhaps if I had chosen to spend more time at the Gala³² I’d have more to say on the positive impact. Unfortunately, and more saliently to my personal experience, added aesthetics also has the significantly more visible downside of introducing red herrings. Our experience on the metas, and particularly Chinatown and Paper Trail, were quite marred by this, especially given potentially-meaningful text that turned out to be nothing but set dressing. (A lot of our hints were just “this thing is not relevant to the puzzle”.)

Personally, I love when aesthetics plays into the puzzle in a pleasing way. But it’s an incredibly delicate balance to strike. I know firsthand from 2024: I tried to evoke a particular aesthetic in the design for Asphodel, but the aesthetic part caused an entirely-wrong initial assumption on the first testsolve. The puzzle that got left on the cutting-room floor had a similar issue, even when I made specific efforts to differentiate signal from noise—and it’s left a (figurative) scar on my right hemisphere that I’ve been struggling to bridge over. (There were other factors to these, too, and they certainly didn’t help—but I attribute these primarily to prioritizing aesthetics.)

If you can make it work, go for it! Just, uh. Be careful with what you leave strewn about. Especially consider a testsolve or two with all the components and art in place.

The third lesson is that red + bird = cardinal in basically everyone’s mind. At least four times I can recall from that weekend.

And since ANE scheduled on top of Mystery Hunt again, it’s time to confuse literally everyone by suiting at hunt next year. :>

Bar the occasional year where New Year’s Day falls on a Tuesday. ↩
A lounge just across from the 26-100 lecture hall, so called because some MIT people stocked it full of bananas for students to grab. Pun very much intended. ↩
A bananagrams bag, because of course it was. ↩
How-to-Hunt is generally hosted by the previous year’s winning team. ↩
The one in Boston, not the one in New York. ↩
I got pulled aside by some stranger because I satisfied the criteria for… some kind of scavenger hunt? Vaguely adventure-themed? I don’t remember the exact details. ↩
At least it’s been there between 2018 and 2024, the Big C notwithstanding. ↩
This Alchemia. One of the puzzles within is about ghost islands. ↩
MIT’s class ring, designed during your class’s late sophomore year. ↩
Team That Is Now Named Later, the casual branch of the TTBNL split. ↩
…yeah, we never noticed the cards on our side being overlapped differently depending on the order. A detail too subtle for discerning eyes. ↩
Literally, since one of the answers was about throwing bull dung. ↩
π is approximately the square root of 10, which would make it the optimum if you were actually trying to guess between two orders of magnitude. The numbers in this one were small enough that you probably needed something more precise. ↩
I can’t believe I didn’t recognize the swish on the top wasn’t the Telugu checkmark. To atone, I will now write ఠ_ఠ instead of ಠ_ಠ in all instances. (Yes, those are derived from the same Brahmi character.) ↩
OK, like two of the pictures showed Katrina putting something under a bench. None of the others really indicated much though, so we were quite lost on this. ↩
Though who knows? Unlock order was extremely non-linear in this hunt. On my walks around campus I saw other teams with a certain physical puzzle ages before we got the same puzzle. ↩
About .5 of which was just spent in the Eastman Lobby looking at progress on Chinatown, but I digress. ↩
Infinite, n. The length of time it takes to traverse the length of MIT’s Infinite Corridor. Not to be confused with ⊥IW’s Infinite Corridor, which is Actually (countably) Infinite. ↩
The character is divided into the “hexagon” (vowel) and the “spokes” (consonant) for translational purposes, which definitely didn’t help the association. ↩
A puzzle requiring ordering 100 pages into a cohesive narrative. Still need to solve this sometime. ↩
A privilege I will lose next year, given I will be out of my MEng by then. ↩
A round from Microsoft Puzzle Hunt 2016 with a similar mechanic, where every puzzle depends on other puzzles’ answers in a potentially cyclic manner. ↩
This despite the fact that L is the first letter in leet. ↩
The absurdity is that I remember seeing hail before ever seeing snow. ↩
A supposed phenomenon where Campus Preview Weekend (the weekend in April where we invite all the prefrosh over to MIT) is not subjected to standard Boston April weather. In practice, though, it’s not really all that effective ↩
At some point I should just learn how to deal with svgs so I don’t have to use an image editor for measurement tasks. ↩
Not much to write about here, though you might see an occasional spot of red near the front of the crowd (I had taken off most of the fursuit to avoid blocking the person behind me.) ↩
I don’t know why I’m attempting to plot the distribution of a continuous function. I guess I’m sampling from this function? ↩
Not to be confused with a uniform distribution. Though, mind that this was also a Big C year. ↩
The exception, of course, is also colored by primacy bias. ↩
Also a puzzle, of course. ↩
Hell, I even dressed the part for all three days just in case. ↩