Joe's Photoblog

Mystery Hunt Wrap-up 2016

Well, the 2016 Mystery Hunt is in the books. As usual, I’ll warn you: Here there be spoilers! This is pretty long, so to summarize, it was fun.

This was my fifth year competing with team Palindrome. As is our tradition, we pick an interesting palindrome for our specific team name. This is usually done via voting about a month before and usually requires a second round to pick the winner. But this year, we had an overwhelming winner on the first vote: “Dammit I’m Mad”. It makes me wonder how we went so long without using it. The team did have a ton of fun answering the phone, “Dammit I’m Mad. How may I help you?”

Anyway, I’ve spoken about the basic format of the Mystery Hunt before, but for the sake of completeness, it’s a collection of about 150 puzzles that have no instructions. If there’s no instructions, then how do you solve them? Well, you typically look at the information you’re given. Sometimes it’s pictures of people or places, other times it’s crossword clues. There’s usually some “aha” moment (although from past experience with my time, I would call these “ooooohhhhhhh” moments) and you end up piecing things together and eventually get an answer, which is a word or phrase.

There are also rounds that have a handful of puzzles in them, and you put them all together for a metapuzzle. These are a huge deal to solve and we try our best to solve them with as few answers as possible. In a Mystery Hunt sized extravaganza, there are also meta-meta puzzles and typically the “meta-meta-meta-puzzle” is the runaround, where the goal is to find a coin hidden somewhere on campus.

Kickoff began Friday afternoon welcoming us to the MIT Muttstery Hunt. They made it sound like the hunt had a Dog Show theme. A few minutes in, something clearly went wrong and it seemed like we were apart of someone’s dream. As we would later discover, the Hunt had an Inception Theme. The idea was that there was some uber-dreamer somewhere on campus who and ten levels of dreams between, down to the Muttstery Hunt. Each round was themed after famous dreamers: Rip Van Winkle, Sleeping Beauty, Opus, King Arthur, Ysera, Dreamtime, Pam Ewing, Randolph Carter, Sleeping Gypsy and Endymion. I’ll get to the endgame later, but first, I want to talk about some of the puzzles.

In the Dog Show round, I started off on Ladder Dogs. This was a series of cryptic clues involving dogs. Cryptics are so popular on our team, it’s usually hard to work on these because we solve on a Google spreadsheet and the clues get filled in within minutes. We found that the clues broke into triples such that the second word was a transaddition of the first word and the third word was a transaddition of the second word (eg HIATUS, SHIATZU, AZIMUTHS). Sorting by the second letter added gave the answer in the first letters.

I worked on two puzzles in the Rip Van Winkle round. The first, Indecipherable was a Viginere cipher encrypted by a sequence of letters that was the melody to a popular song. The second was a very cool ASCII art puzzle. We got a list of four messages with numbers followed by a very long string of ASCII. Factoring the numbers yielded four prime numbers, up to about 139. Putting the message on the largest width gave an ASCII image hidden in the middle of the text with “CUT HERE” on the box surrounding it. For instance, for one, we got a picture of the polar ice caps. Deleting that section and putting on the second largest width gave us a picture of ice. Continuing gave us a picture of the Po river. And the last width gave us a picture of baseball caps (or just “caps”). Removing the last three strings from the first string leaves the letters LAR. Doing this for the four pictures gave the phrase WALLAROO (cross-breed of a Kangaroo and a Wallaby).

I also worked on two puzzles in the Sleeping Beauty round. Fire Swamp was an interesting one where you basically solve a bunch of Ken Ken puzzles and fit them into a grid in an overlapping manner. The final step was pretty cool, too, where the unclued boxes made either plus or minus signs! There was also one called Trubled Monicas, which is one of my favorite types of puzzle. As an example, you get a string CHARACTERIZEEEBEETHEEPRESENCEEOEEFROZEEEWATEEE. Here the clue is saying “Characterized by the presence of frozen water” but with the last letter of every word changed to an E. The answer is ICED, but you change it to ICEE as in the clue. Anyway, the hook was that these were all items from pop culture that misspell words in their names (eg Def Leppard, Blu-Ray). Another one I didn’t work on, but was at least somewhat interesting was “identify, SORT, index, solve“. These are the basic four steps of lots of puzzles, but in this one SORT was by far the worst. You had pictures of things that ended with “ian” or “ion” so they sounded like numbers. These were said to be equal to various large numbers. To give some context, the number of particles in the known universe was the third smallest. In fact, even Graham’s number, which I believe at one point was the largest number to ever appear in a serious math article, is the seventh largest in this list.

The only puzzle I really worked on in the Opus round was Crushed Petals which I thought was fairly straightforward. The King Arthur round proved to be one of two major stopping points for us. We had solved a fair number of the puzzles and knew the metapuzzle had something to do with the periodic table, but it took us way too long to realize it was the circular periodic table we should be using. Oddly enough, with four new elements being announced a few weeks ago, I read up way too much on the periodic table and actually saw that table. The main puzzle I contributed to here was Click to Win. By this point, I was looking for a chance to stretch my legs, so the first part of it was a detailed hunt around campus. Questions like find the second number on the serial number on this lamp post. We did them all and got the numbers 1 through 8 in some order. If you go to the link above, it takes you to a very annoying website with lots of ads to click. Each ad actually references some building on campus close to one of the 8 spots. This took us quite a while to realize. There was one more puzzle that had an amusing story. Follow the Pipes was basically another hunt around campus. You can read the solution on the page, but the group working it ended up with a few letters missing from their answer. What they were supposed to call in was ONLINE PORN (note: the puzzle did not feature any of this, they just needed the phrase for the metapuzzle). Instead, they called in BOVINE PORN. Minutes later we got a call from Hunt HQ saying that, no, that was not the answer. They were laughing. Just to be funny, later in the hunt, on a puzzle that had a sheep theme, we called in OVINE PORN.

For the Ysera round, I mainly worked with a group doing Vagabond Tours which involved getting pictures all over campus. Each one had to include two of ten items. Since someone with a DSLR camera was on the list and since I had one, I was in most of the photos. I also took one photo, with the camera, where I jumped in the air while doing a peace sign. We got in a strange conversation with a student or professor in front of a bathroom (where we had to take a picture). I didn’t really do anything in the Dreamtime round.

In the Pam Ewing round, I worked on one puzzle that I thought was incredible cool, we just never got it. World’s Longest Diagramless utilizes one of my favorite types of puzzles, the Diagramless Crossword. Here you get clues, but not the grid, so you have to figure out where the black squares go. This had a list of about 90,000 crossword clues. The top part solved like a regular diagramless, then had clues for AARDVARK, ABATING, ABET, ABDICATING, ABANADON, ABASE and ABREAST. Well, I was proud to be the one with the “aha” here as this is the order of the words that Yakko sings when he sings the dictionary.

In the episode, they cut away then come back a few times and we found that those words were all clued in the order of the song. The key to the puzzle, which we missed, was that for a stretch of those clues, the first letters spelled out a phrase. My favorite part of this puzzle is that in the song, Yakko messes up at the F’s and says, “Frankincense, Frankish, Frankishish, Shoot, Yada Yada, Flambe”. They clue all of these words (Frankishish being “Kind-of sort-of relating to an ancient Germanic tribe”).

I didn’t do much in the Randolph Carter round, but I will mention two puzzles. The Gibbous, Non-Euclidean Program involved fixing code that was written in Fortran 66. The Eldritch, Gambrel Scavenger Hunt was where having a large team really hurt us. The scavenger hunt usually involves just find a bunch of random things and that’s kind of what it was this year, but as you see, we have to cover a lot of area in a grid, finding things that fit three adjectives. It was incredibly hard because we have a large team, so we had to find more stuff. Fortunately, we had an awesome woman on our team who did all this for us.

In the Sleeping Gypsy round, I only worked on Sculptural Disassembly which was about Hanayama puzzles. I had given my brother one of them for Christmas a few years ago and recognized the general theme of the puzzle. We identified most of them, then I believe I started working on another puzzle before this got solved.

The Endymion Round proved our downfall, and was likely why we didn’t win the hunt. Of the 12 puzzles, we had solved 11 before we figured out the ordering of the answers for the metapuzzle, which had me kicking myself. We likely could’ve solved this from 6 or 7 puzzles, opening the runaround that much sooner. Each puzzle had a quote associated with it, supposedly from Endymion, but really from another source. Turns out the first names of the real speakers were all men who had walked on the moon, something I know a LOT about. Anyway, this round proved to have some of the nastiest puzzles of the hunt. It Is What It Is was about a kids show called Doc McStuffins, where she diagnoses her toys with fictional diseases. There were drawings of her toys, but the diseases were linked with the wrong toy. We quickly figured out that that would give us an ordering, but then we got stuck on the extraction (which turned out to be the page number from the original list). This puzzle falls into what we call STDPA (Solve the Damn Puzzle Already). We had spreadsheet after spreadsheet of things that might be relevant, including things like voice actors names. This was the only puzzle we didn’t solve that round. Orbits was probably my favorite of the hunt. It had a 135 MB text file where each line had an 8-letter string, sorted alphabetically. After lots of looking at the data, we found that if you look at every row 1 mod 7, you have a Y. 2 mod 7 you have an M, and so on. The string was YMCAYMC. On other cuts we found BINGO repeated four times, then BINO. We also found TROUBLE, DIVORCE and MICKEY MOUSE fairly quickly. These were all songs that spell out something. While solving this puzzle, it brought together two factions of our team in an interesting way. Lots of people on the team are the programming nerd type while others are the Broadway/theater nerd type. We asked for them to give us some other songs where they spell out something and we found DIVORCE, OKLAHOMA and METHODOFLOVE. Each of these was repeated a number of times, then a letter was dropped. After some massaging, this gave the answer.

I also had an interesting time on The Man in the Moon a cryptic crossword by Mark Halpin. Like I said above, lots of people on our team like cryptics and are very good at them, so it’s hard to get to work on one at the Hunt. But this happened to come out at about 3 AM, so there was a smaller group of people there and I got to work on it and see all the clues get solved by hand! There were several “aha”s here. It was pretty clear off the bat that most clues needed a letter changed to another letter. It turned out that a handful of other clues had two straight definitions and neither of the answers fit in the grid, but their overlap did, eg AUNTIES and UNITIES give UNITES. The letters that get dropped, when sorted correctly, gave the phrase CRTCHAKAPOWERIZE. We misparsed the clue which took forever to figure out. It’s supposed to be “CROTCH AKA, POWDERIZE”. The leads to GROIN and GRIND as answer, the overlap of which is GRIN.

Time Suck proved to be one of the evilest puzzles of the hunt. It was a 10-hour YouTube video featuring the Spongmonkeys from those Quiznos commercials of the early 2000s.

While describing this puzzle to a coworker he said, “Thanks a lot. I had forgotten that those things existed.” It’s basically grotesque creatures that sing really, really poorly. My team watched for about 2 minutes and discovered that one of them sang a horrible song that went on repeat and at 1:59, the one playing the guitar has a word bubble pop up with a letter. One person used the scroller on YouTube to go through the video and found the letters TSA, which we submitted and was not the answer. It was later discovered that the message actually said TIMESTAMPS AS MINOR PLANET NUMBERS. The letters only popped up for a single second and it was hard to catch while scrolling. One member of our team dedicated several hours to rewatching the video and finding approximate locations of each time stamp. He watched it very sped up and was sometimes off on the exact timestamp by 30 minutes, so the team had to refine it. It turns out that every time a letter appeared, the second Spongmonkey screams a number, which I hadn’t noticed because I had it on mute (the music was really, really annoying).

Solving all of the metapuzzles for these rounds unlocked the Limbo round. This featured ten short puzzles that were callbacks to prior puzzles, one per sleeper round. We operated on a false hypothesis that really killed our progress for a long time. Specifically, the idea was that we had to solve the first puzzle in order to solve the second one, and so on, then we would use the kicks to come back out. In fairness, that is sort of what the graphic implies. It turns out that all of the puzzle stand alone, but we spent a ton of time working on the Rip Van Winkle fragment. There was Morse Code that hinted that 11 names of people on the cover of Sgt Pepper’s Lonely Heart Club Band were hidden in the round. We only found 10 of them, but a few were in question. It turns out, we did have the right idea, namely to use the rows of the cover and convert to Morse Code based on where the people were.

While solving this we heard that another team, Setec Astronomy (a reference to the movie Sneakers), had found the coin. This was quite surprising to my team. Setec had won three hunts in rapid succession in the early 2000s and was sick of running it, so they instituted a rule: Solve a puzzle, take a drink. The hope was that slight inebriation would keep them from accidentally winning. I know several people on Setec and was told that they actually kind of wanted to write again, so they tried this year.

Hunt HQ announced that they were only going to be open for another 90 minutes. The Hunt had already lasted 53 hours, which appeared to be longer than expected. I, for one, didn’t mind the length of the Hunt, mainly because the last two finished early Sunday morning and it didn’t appear like this one was heading into Monday like the 2013 Hunt. Much of our team left, but we started making progress on the other puzzles. The Sleeping Beauty Fragment referenced “FONETIC FUTHARK”, a very old alphabet. It turned out the puzzle “Blocks” had Futhark symbols hidden in it! We got the solution and quickly made progress on others. For the Opus fragment [link] someone found the two words listed (RADIO and ACTIVE) in the Rose Garden puzzle. For the Pam Ewing fragment it referenced striking oil in between the rocks and calcite. Thos two words were in the solution to the diagramless crossword mentioned above and the letters “OIL” appeared all around. Eventually we had enough to guess the meta puzzle answer and unlock the runaround, about 15 minutes before HQ closed.

The host team graciously let us do the runaround. We had to send 10 teams of three to 10 locations on campus. HQ would send us a word to sync us and within 5 minutes, we had to send them a photo of us holding up the word and acting out the metapuzzle answers for each sleeper. I was among the last three in the room and we still had two locations to go. I went in a group of two to the place where we had to “LAUGH LIKE A CRAZY KOOKABURRA”. My friend ended up having to go, by himself, to another location and “KISS AURORA WITH GUSTO”. We were able to get these all done, and they sent us all on a mini-runaround, themed to the characters from the corresponding round. Because my friend was at a location by himself, we ended up doing both his and our runarounds. We met up with members of the host team and got pieces to a puzzle and were
told to go to Lobby 7. The rest of the team was there and we had four pieces to a puzzle. We quickly saw that it told us to go to The Alchemist art piece across the street.

We got there and some members from Luck basically said, “That’s the end of the hunt”. Setec had gotten there a few hours earlier and found the real coin, which was taped against the white sculpture with white tape.

Anyway, this was a superfun hunt and I look forward to what Setec has for us next year.

January 22, 2016 | Categories: Uncategorized | Leave A Comment »