Pieces, our goal!

An account of Dirty Bayes’ experience of the Miskatonic University game, in which I attempt to describe every puzzle we saw.

What is this?!

The Miskatonic University Game was an instance of “The Game,” a weekend-long road rally puzzle hunt that took place in the greater Boston area on August 16th through 18th, 2019. Our team Dirty Bayes was one of thirty-four teams that participated in this event, which lasted for nearly forty-eight hours (with one break overnight on Friday) and sent us driving all around the North Shore of Massachusetts. We encountered some challenges, solved many puzzles, lost a lot of sleep, but had a ton of fun. This writeup is an account of our team’s experience.

Note that this writeup spoils every puzzle we saw in the event; if for some reason you’d like to remain unspoiled, read no further. I’ve also noted each puzzle answer that I remember, but a few may be missing.

Preparing and Arriving

Gloucester Fraternity Club

Our five-person team began this event as a group of three, because a couple of our teammates had a Friday night wedding to attend. Luckily, the wedding wasn’t far from Gloucester, so we planned to convene the full team after the wedding, but before we drove up to our AirBnB near Newburyport—we’d decide the exact location and timing based on our puzzle progress and where we ended up. It was also nice that the three of us who were there from the start were able to fit at a dinner table with all of Mystik Spiral, a team made up mostly of our Mystery Hunt teammates from Setec Astronomy. We enjoyed catching up with them over pizza and salad.

The business portion of the kickoff began at 6pm sharp, with the Miskatonic University Game organizers going over safety rules and introducing us to the web app we’d use to learn where to go, receive hints, check answers, and follow along with the story. They also introduced our first location: Hammond Castle! After some explanation of where and how it might be possible to get an Internet connection while at the castle (since we’d need that for hints and for checking answers, and it’s in a pretty dead area for cell service), we were given the start code for the game, and we were off.

Hammond Castle (Friday night)

We arrived at the castle around 6:45pm, and after crossing over the drawbridge we were welcomed to the “student activities fair” by the university’s Dean Herring, who handed us our first set of four puzzles. We entered our start code upstairs before venturing down into the cellular dead zone to get the lay of the land and find a spot to examine what we’d been given. It was crowded with other teams, but eventually we found a quiet spot with a counter in a kitchen area (technically staff only, but they allowed us to stay there) to spread out the puzzles and start working on them. We decided to initially stick together rather than parallelizing, since there were only three of us.

The Organ Puzzle

The first puzzle we dug into consisted of a few pages of sheet music. One page was of the official Miskatonic University song, while the other pages contained a somewhat different-looking song with different lyrics. We noticed that the individual measures of music were actually identical between the two songs, but in a different order. We reordered the alternate song to match the order of the official song, and now the alternate lyrics described a set of instructions for a connect-the-dots drawing we could trace out based on the positions of labeled organ console controls.

We left our kitchen to find the castle’s organ console, and found a crowd of solvers around it already. We took some photos and returned to the kitchen to trace out the letters. With just B?A??, we supposed the answer might be BRASS, and then with BRA?? we guessed that answer correctly, saving us the time it’d take to reorder and transcribe the last two sets of lyrics.

Color Guard (part 1)

In the room immediately outside of the kitchen, a table was set up with two rows of flags along the long edges, and another set of flags along a short edge with the letters COLOR GUARD. The flags along the long edges had different colors and patterns (numbers of stars and stripes), and each had a crossword-style clue written on its pole. A printout of all the clues on a single page was also provided, so we grabbed one of those, wrote down the rest of the data, and then headed outside to find a spot to work on this. We found a table with a beautiful view overlooking the ocean.

We solved some of the clues quickly, but others were tough for us, especially because we hadn’t yet figured out any constraints on the answers other than that their lengths matched the numbers of stars on their flags. We thought that one side of clues might have alphabetically-ordered answers, but that ultimately turned out to be incorrect. After staring at this for a while without making significant progress, we switched to consider another puzzle.

Athletics

This puzzle took the form of a screenplay for some sort of promotional video for Miskatonic University athletics. There were a number of students introduced with strange names who played different sports, and somewhat obscure clues for what those sports might be, as well as some sections that seemed to imply orderings. We eventually started to identify the sports as the names of Olympic track and field events, and figured out that the students’ names could be combined to form transdeletions of the names of the events. This gave us a set of letters, and we ended up anagramming them (rather than figuring out whatever we were supposed to do with the ordering information in the screenplay) to get the answer PROJECTILE. This was the only puzzle of the four that didn’t rely on any environmental data from the castle.

It was now about 8:15pm, and it was getting pretty dark out, so we decided to head back inside, which turned out to be necessary to start the next puzzle anyway . . .

Humor Magazine

This puzzle contained a list of descriptions of art that could be found around the castle, along with clues about a type of object whose quantity could be counted within each piece of art. It was printed on two pages, so we split up to search the castle, with two of us taking the first page and the other taking the second. We spent quite a while searching for the artwork—usually it was pretty clear once we’d found the correct thing, but there was a lot of stuff in this castle! It was also sometimes tricky to get the counts exactly right, especially when there were quite a number of objects depicted (e.g. number of spear tips in a painting) or when the art was far away and hard to see in the dark. We correctly hypothesized that the counts should be converted to letters (using A=1, B=2, etc.) but we made a mistake by failing to number the pages before we separated them, and we had enough missing or wrong data that it wasn’t clear just from the letters we were seeing that we should switch the page order. We fumbled around a bit more before finally doing so and determining that we should ASK FOR HUMOR GUIDE.

We asked a member of the always-helpful Miskatonic University faculty for a humor guide, and were led back to the kitchen in which we’d set up camp earlier to be handed a page containing four jumble puzzles. The crowd of solvers had begun to thin out by now, so we had space to camp out on the stairs of the main castle hall to start working on these. Each jumble ended up solving to a phrase that was a homophone of a temperament of the ancient medical concept of humorism (e.g. “FLAG-MATIC” clued “phlegmatic”). Numbers scattered throughout some of the jumble boxes corresponded to an enumeration at the bottom of the puzzle, which became the instruction GROUP BY HUMOR AND INDEX. Each jumble image also contained a hidden number somewhere in the picture, which we determined could be used to index into nine-letter names that were associated with each art description from before.

Unfortunately, the three of us were not experts on the properties of the four temperaments, and spotty Internet access made it frustrating to research what we needed to learn about them to be able to group the art descriptions. We eventually got a fair amount of help from Dean Herring and Chancellor Red on how to identify which descriptions were associated with which temperaments, but our biggest problem was probably continuing to try to extract an answer in the original puzzle order, rather than reordering the description names by their associated temperaments. Once we finally attempted that, we narrowed down the possibilities enough to extract the phrase AMAZONBLANKVIDEO and solve the puzzle with the answer PRIME.

Color Guard (part 2)

It was getting quite late at this point (it was after 10pm) and we were told we’d only have a few more minutes to work before the site would close and we’d be skipped ahead past the first meta to set us up for tomorrow. On the bright side, though, we were soon surprised by our other two team members arriving at the castle! We hadn’t been in touch with them since earlier in the evening (due to poor cell service), but they assumed we were still working since we were unreachable, and they took a cab from the wedding to join us. They helped us solve a few more clues, and we received some hints from Dean Herring about how some large banners hanging from the ceiling of the castle’s main hall could be used as a reordering key to start matching up our answers, but ultimately we only confirmed one such pairing before we were ushered upstairs and given the answer HEADGEAR to enter as the site was closing.

Student Activities Fair Meta

We did get a quick explanation and peek at the meta for this round of puzzles, which involved a “facebook” wall of pictures of our fellow classmates along with their majors. A few weeks earlier, we’d been asked to send in photos of members of our team including their favorite numbers, and a few of those photos appeared on the wall. The trick to the meta was to only consider photos that matched the meta answers (e.g. only PRIME numbers, only BRASS instruments, etc.) and take the first letters of those students’ majors to spell the meta answer OCCULTURAL HERITAGE.

We were then brought outside to complete an interaction with a member of the fraternity ΕΩΔ, who informed us that we were actually legacy students, and gave us a box that had been left there for us by our alumnus relatives! It contained a very Lovecraftian tentacled plastic monster “artifact” token representing completion of this round of puzzles, and the box had enough space to hold a few more of these that we’d collect throughout the weekend.

Back at the Gloucester Fraternity Club, the organizers had supplied us with a large Miskatonic University seal magnet to apply to the side of our car, and as we returned to it, we noticed it had sprouted a tentacle in our absence. Things were starting to get creepy.

We finally started heading up to our AirBnB in Haverhill around 10:45pm, using the drive time to fill in our recently arrived team members about what they’d missed. When we arrived, we made a plan to leave the house early to grab breakfast in Newburyport in the morning before our 9am start time at the next site, and went to bed.

Newburyport (Saturday morning and early afternoon)

Our first location in Newburyport was the Masonic Lodge, where we were seated at tables in a large room to begin our first day of classes. The theme of this round was determining the textbooks we’d need for each class, and our first two classes were Horology and Basic Maths. The web app also had us scheduled for an Anatomy class at 11:30am. We initially split into two groups to work on the first two puzzles in parallel, but then joined forces on Horology after we solved Basic Maths first.

Basic Maths

This puzzle was projected as a slide presentation, where each slide contained one or more Greek letters and two Equation Analysis Test clues. The slides changed very quickly, so we recorded a video of the presentation on a phone, and solved from the video. We realized that each slide’s EAT clue numbers were in different bases, in increasing order, and noted the bases. This gave us a mapping between Greek letters and base numbers (which we noticed were all 26 or less), and then converting the bases to letters (using A=1, B=2, etc.) and ordering them by their associated Greek letters spelled the answer.

Horology

For this puzzle, we were provided with an exam booklet that had a time zone and an “error code” (made up of two three-letter words) written at the top of each page. We also noticed there was an analog clock in the room that was rapidly cycling through different hand positions (in groups of four, with a “break” where it reset to noon after every four positions). We grabbed a video of this one as well, but this took a while to record, as it had a much longer cycle than Basic Maths.

The times all had good hand positions for flag semaphore decoding, but initially the groups of four didn’t seem to be spelling anything useful, so we thought we’d need to shift the times by different time zones before decoding them using semaphore. This was quite a daunting task, though, as we didn’t know which time zones to associate with which hand positions. Eventually we realized we could read an instruction phrase directly from the semaphore if we ignored the groups of four, and it told us to match the groups of four up with the three-letter error codes from the booklet (leaving an extra letter per code), and then Caesar shift the remaining letters using the timezones. This spelled the answer, and we finished this puzzle around 10:15am.

Cartography

With over an hour to go until our Anatomy appointment, we were given a Cartography puzzle to work on in the meantime. This came as a deck of cards with map sections printed on one side and photos printed on the other. Each map section had a small numbered red circle (from 1 to 22) marking a location on the map, and the photo had a red circle obscuring a letter.

We pretty quickly realized that assembling the full map would require partially overlapping the cards at different angles, and we developed a system for taping the cards together as we found alignments while still allowing the cards to be flipped along one edge so we could look at the photos on the backs. Pretty soon, we had assembled a map of downtown Newburyport, and we headed outside to find these locations and determine the obscured letters.

We split into two groups, dividing the map in approximately half and tackling different sides of town in parallel. We really enjoyed this activity—it was a great way to see the town, and the scavenging presented just enough challenge without anything being too annoying to find. By 11am, both groups were finished, and we combined our letters over chat to extract the answer FOUNDATIONS OF FIELD WORK. This was good timing as our team was called early for Anatomy class.

Anatomy

We were invited onto the stage of the Masonic Lodge, behind the drawn curtains, to discover a giant copy of the board game Operation! We got to put on oven mitts and wield large wired metal tongs to extract small plastic items from holes in the board, without touching the sides (as it would buzz Operation-style, and we’d need to start that hole again). This was a funny and exciting reveal, and was ridiculous and fun to play—most of the objects weren’t too difficult to grab, but a few presented some challenge.

Once we had our items, we returned to our table to solve the puzzle. They gave us a printout that noted the body areas where each item originated, as well as two numbers associated with each item. We figured out that one number would indicate extraction order (they were consecutive) and figured the other would be for indexing. We considered a few incorrect theories related to chaining items together, finding letters in common between items and body parts, etc. before we broke in with the correct idea: find a word that’s both the second word of a two-word phrase with the body part, and the first word of a two-word phrase with the item, and index into that word (e.g. if we found a monkey at the elbow, we’d want the word GREASE, because it can make “elbow grease” and “grease monkey”). Taking these letters in order revealed the answer.

This was the last puzzle in the Masonic Lodge. We were sent to the waterfront next to pick up materials for the next puzzle, and were advised that this would be a good time to break for lunch, so we found a nearby sandwich shop and began looking at the puzzle materials as we sat down to eat.

Stained Glass Mosaics

With this puzzle, they gave us a photo of a stained glass map of the waterfront, with five locations marked, and a whole bunch of construction paper cut-outs with punched holes looped together into several different strands using yarn. There wasn’t much we could do with this while sitting and eating, other than start grouping the strands by color and speculating about what we might find when we followed the map.

What we found after lunch were some pillars with some very pretty stained glass art of ships. We were able to match the construction paper pieces to the glass pieces used to make the art, and this traced out paths on the art which looked like it could be forming letters (we realized later that if we connected where the holes were punched in the pieces, it was even more precise). We split up—part of our team went to gather photos of the art, while the rest of us grabbed a picnic table. We were able to present our photos to faculty to obtain paper copies of the art we could write on. Once we had these, it still took us a while to match all the strands, as many of the art pieces were fairly similar.

One of the strands was a bit different—it was made using red yarn instead of blue, and its construction paper colors referenced the art as marked on the map with colored dots, rather than individual glass pieces. We’d use this strand for answer extraction, and we guessed we’d want to take the letters from each pillar in top-to-bottom order, but we ran into a few problems before we were able to get this to work.

One problem was we’d managed to mix up the associations between the pillars and the extraction strand piece colors, because we’d assumed that the photos stayed in order when we uploaded them to our chat room, which they had not. Our second problem was an incorrect change in our approach to ordering the letters extracted from each piece of art—we’d noticed that all of the letters on a piece of art were made up of a unique number of points, so we thought we could order by that count, but it turned out that top-to-bottom was the correct order after all. We solved the puzzle once these two issues were resolved.

First Day of Classes meta

We were sent to the library next, where we received materials for a meta: the book covers for the textbooks whose titles were the answers to all the puzzles we’d just solved. We found some steps to sit on at the bank across the street to work on this.

Each cover had a dashed line printed on it, which we assumed meant we should cut along the line (a correct assumption in this case, but not always—more on that later). Now we had two irregularly-shaped cover pieces from each textbook, and we weren’t sure how to combine them. We tried arranging them side-by-side in various ways, and thought maybe we saw some animal shapes at some point, but this wasn’t really going anywhere. Eventually, we started to see some letter alignments of the title letters as we stacked the pieces on top of each other—and this quickly formed the meta answer, BEYOND THE MOUNTAINS OF CLASSWORK.

After submitting this, we had a brief interaction with a ΕΩΔ fraternity member sitting on a bench across the street, where we received another plastic meta artifact thingy and were sent on to our next location.

A Beaded Tiara

After a bathroom break and a quick stop at the van (which was parked near the next puzzle location), we met a faculty member by the Frog Pond in Newburyport. She shared with us a brief out-of-world (but welcome) history lesson, pointing out that the oldest still-active courthouse in the US was a nearby building. She then handed us a couple of printouts and a cardboard tiara with lots of beads threaded through it in various ways, instructing us to sit near any birdhouse we could find.

There were many other teams scattered around the pond, so we walked for a bit until we found a relatively isolated spot, and then split up as some of us did tiara bead analysis and others performed some birdhouse scouting. There were seven birdhouses in all (matching seven points on the tiara), but some were a bit hard to see, and it took us a few laps around the pond to nail down all the information we needed from them. Each birdhouse was signed with the name of the person (child?) who painted it, and these names matched enumerations that were encoded in beads on the tiara. These enumerations had associated color beads, and color pairs were encoded elsewhere on the tiara in beads to indicate straight lines we could connect from birdhouse to birdhouse across the pond.

The sheets of paper we were given had an outline that matched the shape of the tiara and the pond, but also had letters scattered throughout the middle. Extracting the letters along the straight lines made an instruction . . .  to do the same thing again with the tiara upside-down! Another round of line drawing instructed us to extract the remaining letters, which spelled the answer, NOWHERE.

It was around 3pm at this point, and finally it was time for us to get back into the van and drive to a new area. As we returned to the van, we noticed it had sprouted yet another tentacle magnet . . .  we’d end up with quite a few more of these before the end of the weekend.

Saturday late afternoon and evening

Shadowboxes

Our next location was another park just a short drive from Newburyport. Here, we found many other teams crowding around a table covered in fifty-three boxes containing cutouts of letters stacked on top of each other, with some numbers also written on the sides of the boxes. The letters within each box spelled a word, but since they were stacked up and we couldn’t open the boxes, some of the letters were obscured, and it wasn’t always easy to identify the word. We split up the data collection task, and then sat down on the grass nearby to analyze what we’d learned.

The words from the boxes could be connected together to form crossword-style clues, with the numbers on the sides of the boxes indicating the lengths of adjacent words in the clue. We were able to get a few to work with the data we had so far, but many of our word identifications were pretty far off, so we decided to have a second team member look at each box, and then tried to reach consensus—this helped correct many of our errors. Extracting the answer involved determining which letter of each clue answer would have been totally obscured if that answer word had been stacked in a box, and these letters spelled the thematically appropriate answer PANDORA.

Witch Cards

Now we were sent on our longest single drive of the day to the Witchcraft Victim’s Memorial in Danvers, where we arrived around 5pm. The monument here contained biographical information and quotes from victims of the Salem witch trials. We were given a set of trading cards that had the pictures and names of victims on one side, and enumerations of their quotes (with asterisks for letter extraction) on the other, along with a clue about what they were “accused of”.

The asterisked letters didn’t make much sense on their own, but we started to make progress when we realized that the “accused of” clues all clued words that were bookended by the letters in DEMON, DEVIL, SATAN, BAAL, etc. The asterisked letters then made up most of the rest of the word, leaving behind just a few letters that spelled an instruction to “unpeel” the cards!

Shortly after starting this puzzle, 3/5ths of our team took the van on a bathroom and Dunkin’ Donuts run, so there were only two of us around for this reveal. We were able to peel stickers off of the backs of each of the cards to reveal some white lines, which could be used to connect the cards in a layout. We tried to use this to reorder the letters we’d already extracted, but it didn’t spell anything good. Soon the Miskatonic University faculty stationed at the site came by to check on us, and wondered why we hadn’t peeled off the fronts of the cards . . .  great question, I have no idea why we didn’t think to do that. A letter was missing from most of the names on the peeled-off front sides, and taking these letters in white line order spelled the answer EPITAPH.

Dinner and Jigsaw

After we finished the witch cards puzzle, we were given a bag full of jigsaw puzzle pieces, and were advised to find somewhere to have dinner while working on this puzzle. We drove to a Qdoba in Danvers which worked out well for this (we met a few other teams passing through there as well).

As we assembled the puzzle, we noticed that it was made up of a few different disconnected sections. Each section was colored like a nautical flag, had the shape of a country, and had a hole in the middle in the shape of an aquatic animal. The nautical flags formed an ordering (from A to G) and the country names crashed with the animal names on a single letter, allowing us to extract the answer ECHELON.

Cow Tipping

The next stop was Richardson’s Ice Cream, where we settled down at a picnic table near the mini-golf course to dig into the Cow Tipping puzzle. It came with a set of instructions, three printouts of “holes” to play, and two very cute rectangular prisms representing cows. The instructions explained rules we should follow to roll the cow around on top of the printouts, with the goal of the cow falling into a hole after a certain number of moves. There were features we’d encounter along the way that would make or erase marks on the cow, and sensors at which we’d extract letters.

We figured out that the marks would form braille letters on the cow’s back, and together we worked through the first hole to extract the letters CHI. Knowing there were seven sensors total, we expected an answer like CHI????, and guessed CHICAGO, which was not correct. We split up working on the other holes, but as soon as we extracted the first letter from the third hole and had CHI??E?, we guessed the correct answer CHICKEN, and moved on.

Drawn and Quartered

It was around 8pm and starting to get dark by now, which contributed to the creepiness of the next location. We were sent to a parking lot, where we encountered a crime scene with a bunch of chalk outlines of dismembered body parts, marked with evidence letters. It turns out that the murder victims were the Scooby-Doo gang, and we had to figure out who had committed the crimes.

This puzzle was a low point for us in terms of solving acuity. Re-assembling the victims was an obvious first step, and at least a couple of our team members were able to match chalk outlines of arms and legs to Scooby-Doo characters without too much trouble (I certainly was not able to do this), but we struggled to make quick progress after that. One of our team members had the early insight that we might want to use the Sherlock Holmes dancing men cipher at some point (based on the detective theme, and knowing the poses of the body parts), but by the time we had gathered the data needed to apply it, we had completely forgotten about it. There was also some trickiness related to Scooby’s letters being used for two different word extractions (based on some of them being backwards) that also threw us off for a while.

After fumbling around with this for probably about a half-hour, we started to succeed in reading the names of famous male dancers (e.g. Fred Astaire) off of the evidence letters on the bodies, by always going around the bodies in the same way. Each was missing a letter, which provided an A through G ordering. Finally, we were indeed supposed to use the dancing men cipher to extract a letter from each body to get the answer, but the poses were also quite tricky to confidently identify. After a bunch of massaging of our data we determined the answer was HONESTY.

Have You Seen Me?

We drove to another parking lot to receive our next puzzle. There was nothing site-specific about this one—we were handed a milk carton, and proceeded to work on this puzzle inside of our van. Opening the milk carton produced a couple of copies of a diagramless crossword puzzle. There was a matching 15 × 15 grid on the outside of the carton, with just a few areas highlighted and a few words and numbers filled in—we left this aside to be used for answer extraction later, and started solving the crossword puzzle.

We didn’t have any experienced crossword constructors or diagramless solvers on our team, so this was also quite a challenge for us. We made some sort of error on our first attempt, and were convinced that there’d be a gimmick to this puzzle where some words would stick outside of the grid, until we came to our senses and realized that we had probably just messed up the diagram. A second attempt on a separate sheet of graph paper went a lot more smoothly, but we lost a lot of time here.

We got stuck on the answer extraction as well. We were pretty distracted by a picture on the carton featuring “Akeley’s Farm” and a cow being abducted by aliens, as LAKE and ETS crossed in the solved crossword grid . . .  we thought this might be cluing us to use the crosses somehow. One of our team members looked at the hints, and pointed the rest of us to the presence of EDWARD GOREY as an answer that might be suggesting something. We recalled his abecedarian Gashlycrumb Tinies, and realized that all of our answers that matched the highlighted areas on the carton grid referred to Gashlycrumb letters, allowing us to extract the answer INKBLOT.

It was now around 11pm, and a torrential rainstorm had started as we’d been sitting in the van working on this. We drove through the rain to the next location.

Independent Study meta

We quickly darted out of the van and through the rain into the next location, which was an old train car—this was a very cool spot. The seats were in groups of four (two facing two), which was a bit awkward for our group of five, but one person took this opportunity to take a nap in the van, so our four remaining solvers were able to sit together.

The meta handed to us was our “invitation” to the ΕΩΔ fraternity party. It was a list of sixteen nonsensical words, and we were supposed to use it to figure out the password needed to enter the party. We were also shown all of our previous answers from this round on the web app, as well as one answer to a puzzle we’d been skipped past at some point (STOMACH), for a total of eight answers. We stared at the list for a while before we started to figure out that our answers could be inserted into the middle of the nonsense words (with each of our answers ultimately used twice) to make two new words, leaving one or two leftover letters in the middle to extract. For example, NOWHERE could be inserted into PIASY to make PIANO W HERESY and extract a W.

This resulted in my personal favorite Lovecraftian pun answer of the entire game, KICK BACK WITH SOME OLD ONES.

Party Time (overnight)

The next location would be our “stay off the roads for the night” location—we arrived there at about 12:15am, and would be there until around 5am. One of our team members’ birthdays was on Sunday, so as we pulled into the parking lot just after midnight, we threw him a little birthday party just outside the van (complete with party hats and candles in a Twinkie) before we went inside.

We had a brief interaction with our ΕΩΔ contact, where we gave him the meta answer password and he gave us another plastic monster meta artifact and welcomed us to the party. It was pretty crowded with other teams when we got downstairs, so we started out sitting on the floor, but were able to grab a table after a few minutes (as teams were bouncing in and out working on puzzles that sent you out of the main room).

Beer Pong

This puzzle was given to us as a bunch of red Solo cups (of course) with multiple letters written on the bottom of each cup. There was also a blank space in the center of each cup bottom to fill in an additional letter, some shapes outlining a few of the letters, and some line patterns (straight, zig zag, squiggle) on the sides of the cups.

We figured out how to group the cups into two beer-pong-style triangles by matching the line patterns on the sides, and started to notice parts of some names of beer brands (e.g. Sierra Nevada) winding their way between cups, but we couldn’t quite pull together the rules for jumping between cup triangles until we were pushed some hints about how to use the outlined letters to connect words. We noted missing letters from the beer brands in the blank spaces, and used these to extract the answer VOLLEY.

Paddle Maze

The next puzzle took the form of another fraternity trope, a wooden paddle. This paddle had lines cut into it forming a maze, and several rotatable wooden pieces that could connect different segments of the maze. Once the maze was solved, each rotatable piece made a semaphore letter, and these letters taken in maze order spelled the answer.

Morse Code Cryptogram

We were sent just outside the building to a courtyard, where we encountered many different-colored sets of illuminated eyes hidden in the bushes, flashing messages at us in Morse code. Decoding the Morse code spelled some Lovecraftian words such as XOTH, NORTHOT, and RYLEH. We also collected a page of encrypted text, where the colors of the letters matched the eye colors. We went back inside with our data to decode the text.

We succeeded in applying cryptogram solvers to the monochromatic words in the text, giving us a (partial) substitution cipher key for each color. Decoding the Morse code words using these same keys spelled NATO alphabet words, which, when ordered in resistor color code order, spelled the answer.

The Silver Key

We were directed to a room upstairs to pick up our next puzzle. We were given an internally-locked wooden box with some pegs sticking out of the top of it, as well as some gears that could be fit onto the pegs. The top of the box also had a message written around the pegs in a circle: TO END THE LOOP WIND BACK ITS CENTER ONCE. This puzzle was themed around Lovecraft’s short story “The Silver Key,” and turned out to be about time travel.

Our first task was to fit the gears on to the pegs such that they all turned together, which wasn’t too tricky to do with only six gears. We also noticed that a line had been drawn looping across all of the gears, which we could align to give us an initial position for each gear. We then followed our instruction to turn the middle gear counter-clockwise by a single full rotation, and were delighted to then be able to open the box.

Inside the box were a few more items: a large gear with teeth on both the inside and outside (which fit around all of our existing gears), a small gear with a hole in the middle (which fit between our new large gear and an internally-toothed gear already anchored to the top of the box), nine slips of paper, and the eponymous silver key. We were able to fit the two new gears to the top of the box such that the smaller gear would indicate different letters of the instruction as we rotated the mechanism, and the key fit into a keyhole on the top and could be used to rotate the mechanism more easily and by specific numbers of key turns.

The nine slips of paper contained text with some instructions and data, which could be used to form a sort of logic and math problem that would determine the lengths of nine rotations needed to obtain the letters in the answer. The slips needed to be ordered by their first words, which matched the instruction from the top of the box. We worked out a system of equations to determine the rotation lengths, and after correcting one mathematical transcription mistake we made which caused a contradiction, found values that worked. We were able to correctly guess the answer after performing about half of the nine steps.

Rows Garden

We were set up in a room upstairs with a Rows Garden style crossword puzzle, which was built as a very impressively constructed “rose window” on which we could write with dry-erase markers. We were provided with a list of clues for the puzzle on a sheet of paper. The letters around the outside of the puzzle grid were given, and we’d fit answers into the grid starting from these outside letters and gradually work our way toward the center, which is where we’d extract the answer.

DNA Samples

This puzzle was presented as two areas of stuffed animals in jars that were lined up along some windowsills upstairs. The stuffed animals were stacked up in groups of three, and had a lot of missing eyes. We took some photos of the displays and then went back to the tables downstairs to solve.

We realized that this shape of two eyes wide by three stuffed animals tall could make Braille letters, and decoded the first area of stuffed animals to spell DNA, and the second area to spell USE CODONS. There were roughly four different colors of stuffed animal: pink, purple, green, and brown/beige. We determined that these four different colors could correspond to the four different nucleobases of DNA (A, G, C, T), and then groups of three nucleobases could form amino acids, which would each allow us to decode a single letter. We just needed to figure out which color represented which nucleobase.

It turned out that the DNA area of stuffed animals was the key—it encoded the letters D, N, and A in both Braille and in amino acids. Using this, we were able to disambiguate the stuffed animal color to nucleobase mapping, and read the answer to the puzzle in amino acids from the other set of stuffed animals.

Party Time meta (part 1)

It was around 4am at this point, and the overnight site was scheduled to close at 5am, so the organizers started to skip all teams ahead to the Party Time meta. This had a few consequences for us: it resulted in there being two puzzles at this site that we missed (and were given the answers to by the web app), and it also meant that the meta location, which was inside of a U-Haul truck in the parking lot, started to become very crowded with teams.

We sent one team member inside of the truck to capture as many photos and as much video as he could, so that we could attempt to solve from images, but many areas that we’d need to be able to see were obscured by members of other teams also standing in the truck. The interior walls of the truck were covered with maze-like lines that connected letters in eight different words. We assumed these words could be associated with our eight puzzle answers, and the maze would somehow form an answer extraction ordering, but before we spent time digging into that . . . 

The “Table”

. . . we’d been hearing rumblings from friends on other teams that there was a very cool “table puzzle” that we’d missed, and we wanted to see if we could at least take a look at it before it was shut down. We found the room upstairs that housed it, and luckily ran into Sarah (Miskatonic University Game head honcho) up there, who very graciously allowed our team to give it a try—at first, it sounded like we’d be paired up with another team that was already scheduled to do it, but when that team ended up skipping it, we had our own instance of it to play.

This puzzle was presented as a very mechanically impressive automated Ouija board. You would interact with it by speaking to it, and then it would respond by magnetically moving the indicator to spell out its response. It started by asking us to say seven-letter words, and as we offered words for its consideration, it would confirm it heard the word correctly by spelling it, and then indicate YES or NO a certain number of times. We figured out that this was a Mastermind game for guessing a seven-letter word—it was indicating with YES how many letters were in the correct position, and with NO how many letters in our word were also in its word, but in different positions.

We had some fun figuring out the rules and playing with this for a while, but since we’d already been exposed to the meta, one of our team members had already (briefly) seen the answer to this puzzle, and remembered it pretty quickly as we started narrowing down the possible words. He excused himself from guessing at that point, but it seemed kind of silly to keep playing pretending that we didn’t know the answer, so we soon offered the miraculously good guess of CHAFING so that we (and the folks running this puzzle) could move on. Before we left, the constructor of the puzzle showed us the internals, and explained a bit about how it worked, which I’m really glad we got to see. I was amused that their “voice recognition” system was not automated, but rather involved baby monitors and a human in the next room over manually entering our guesses.

Party Time meta (part 2)

Returning to the meta, one of our teammates noted that there was an aspect of confronting your fears present in the flavor of the puzzle, and we started looking up the scientific names of the fears of words in the puzzle (e.g. the fear of undressing is “dishabiliophobia”). We were able to connect these fear names with our round puzzle answers by forming two-word phrases that used the beginning of the fear names (e.g. CHAFING dish). Then, we could index into the round puzzle answers by the same number of letters as the meta maze indexed into its words - but our data on the meta maze was still pretty rough. We had it mapped out well enough to extract the first five letters correctly (ONEFL) but then it became messy and impossible to tell without returning to the truck what came next after that.

One of our team members proposed that this could be some sort of punny Lovecraftian take on ONE FLEW OVER THE CUCKOO’S NEST, and we quickly came up with ONE FLEW OVER CTHULHU’S NEST as an entertaining alliterative pun . . .  but alas, it was not a correct guess. A couple of us headed back to the truck (which by this point had emptied out quite a bit) to run through the extraction properly, but while we were in the middle of doing that, another teammate had the bright idea to try ONE FLEW OVER THE CTHULHU’S NEST, which was correct! So ultimately we did solve this puzzle from just ONEFL-.

Our time at this site ended with one of the best interactions of the entire event, as we spoke with a doctor who explained to us that this whole time we’d actually been hallucinating inside of an asylum, pointing out all of the ridiculous details we’d believed to be true (Chancellor Red and Dean Herring? Only $250 for university tuition in 2019?). He said we must decide whether to continue to indulge or to fight the delusions, and handed us our next puzzle, which was a Rorschach test.

Rorschach Test

We returned to the van, and started working on the Rorschach test puzzle there. There were four sheets covered in black shapes and dashed lines, and just as we’d done with the dashed lines on the book covers earlier, we got out our scissors and started cutting up the pages. Big mistake.

We decided that two things that would help us to make progress on this were a table and caffeine, and amazingly there was a nearby Starbucks open at 5am, so we drove over there and spread out the pieces on a table. After only making minor progress trying to fit and overlap the cut-out pieces together (we did somehow manage to spot the word MADNESS this way), we peeked at the hints for this puzzle, and they were all about folding the pages, not cutting them . . .  sigh. Luckily, we had labeled each piece based on the sheet it had come from, and we were able to reassemble the original sheets and tape them back together.

After some fumbling around with folding, we were able to see one word per sheet, and after trying some incorrect word orders (such as MADNESS LIES THIS WAY), we eventually submitted THIS WAY LIES MADNESS, which was accepted as a correct answer.

Breakfast

As we arrived at our next location (the “Sons of Italy” banquet hall in Winchester) we were ushered into a back room for an interaction with Dean Herring. He explained to us that everything we’d experienced was real after all, and that we would join him in summoning Cthulhu and bringing about the apocalypse, achieving “pieces: our goal!” We chanted this with him a few times, and he gave us some colored amulet-like necklaces to wear to protect us, and then sent us into the other room for breakfast.

Little Monsters

There were two puzzles we solved in this location, the first of which involved determining the relationship between features and names of monsters. There were a bunch of very cute crayon drawings of monsters posted around the room, with different numbers of tentacles, teeth, eyes, different colors, etc. and with very awkward-looking names. We were given a blank sheet of paper with just a name on it, and some crayons, and our task was to correctly draw the monster with that name.

We came up with some of the rules with relative ease (e.g. the number and position of teeth indicated numbers of vowels and consonants, hair/horns matched the number of ascenders and wings matched the number of descenders, etc.) but colors were a bit more complicated and took us a little longer to figure out, and we got stuck on the tentacles rule for quite a while. Once we submitted our drawing, we were given another sheet of paper with the opposite task: given the monster drawing on this page, come up with its name. It took us a few minutes to narrow it down before we submitted the correct answer.

Monster Bags and Boxes

This puzzle was given to us as four colored monster bags, each of which contained three smaller colored monster bags, each of which contained two smaller colored monster boxes, each of which contained a colored monster sticker and a crossword-style clue. We solved several of the clues (mostly, but not entirely, correctly), and determined that the color of the largest bag was correlated with the presence of a particular vowel in the answers of the clues within it, but we then got stuck for a while trying to derive more rules for the lower levels—we thought they might have to do with more than just letter presence, as we were reaching some contradictions when considering our solved answers so far.

We sort of sat here without making notable progress for quite a while, and it certainly didn’t help that it was around 8am at this point and our exhaustion was becoming quite a significant limiting factor. A few folks had been taking turns napping in the car during this breakfast stop, but those of us who were actively working on this puzzle hadn’t slept at all overnight, and it was really starting to show. Finally, one of our team members returned from a 30-minute nap and jolted us back into action—he got us to reconsider some of our clue answers, and focus more on the single box that contained the answer we’d extract, and this finally got us unstuck.

On our way out of this site, we were handed our next puzzle, which we were told we should try to solve during our drive to Boston (about 20 minutes away). There was no way this was happening—as soon as we got into the van and started driving, everyone except for the driver immediately passed out and slept for the entire trip.

Boston

As we parked in the parking garage in Boston, we started to look at the music box puzzle while still hanging out in the van.

Music Box

We were given a music box, four different strips with punched holes that could be played using the music box, and a USB thumb drive. The thumb drive contained recordings of several dozen Christmas songs that had their titles and lyrics rewritten to be about Lovecraftian subjects. We had to match these songs to the melodies played by the music box—some were easily-identifiable Christmas songs whose titles had not changed much, so they were easy to match, while others were larger departures from the original source material and required a little more trial and error to find. Each strip could play two different songs, depending on which direction we fed it into the music box. They also contained boxes that fit the Lovecraftian song titles, and when the titles for both songs were written into them, they matched at a single letter, which across the four strips spelled out the answer ROCK.

Labyrinth

We grabbed everything we thought we’d need for the rest of the game, and left our van for the last time to head toward the Greenway in Boston. Our first stop here was the labyrinth in the Armenian Heritage Park. We were given a magnetic whiteboard and a whole bunch of magnetic pieces with letters on them that could be assembled into the shape of the labyrinth. This took us a while to sort out, both because the path of the labyrinth itself was a bit complicated (we eventually pulled up an aerial photo of it on Google Maps, which helped), and because the magnetic pieces had subtle differences in curvature that made them a bit hard to place.

Even once we had the shape worked out, we still needed more help (from the always-helpful Dean Herring) to figure out the next steps—we were definitely running on fumes at this point and weren’t doing a great job of coming up with ideas to try. Reading through the labyrinth letters in order revealed names of famous people connected to one of four categories that were named surrounding the labyrinth (art, service, science, and commerce), with one extra letter per magnetic piece. These extra letters spelled the instruction phrase BEHEADED IN MATCHING QUAD. Reading in straight lines from the outside toward the center in each quad (e.g. not along the labyrinth path), we spotted names missing their first letter like INSTEIN, UGHES, etc. and the missing letters spelled the puzzle answer.

We walked a little bit down the Greenway to another stop, where we received our next two puzzles. These weren’t location-specific, so we decided to bring them into a nearby Panera Bread, and we split them up to work on them in parallel.

Bestiary Cards

This puzzle consisted of one double-sided card, thirty single-sided cards, and a piece of paper with diagrams and Lovecraftian creature names. The center of each card had either a diagram or another Lovecraftian name (with no overlap with those on the piece of paper). On the perimeter of each card were four constraints that dictated which cards could neighbor that one.

We slowly but surely matched cards in this way and found that we were making a 4 × 4 grid of cards, where one side had all of the names and the other had all of the diagrams. Names on the periphery of the 4 × 4 grid gave constraints for the diagrams on their flip sides and vice-versa, as they shared an edge. The diagram sides of the cards all had numbers, which indexed into the names and gave us an instruction to swap out the middle four cards in our grid. The constraints from the remaining cards allowed us to substitute in four of the diagrams and names from the piece of paper in their place, and repeating the extraction step again with these new cards gave the answer, ABUT. Which we of course tried to submit as TUBA first.

Fishy Phrases

This puzzle came with a bunch of mostly blank magnetic 3x3 tiles (with a variable number of cells blacked out per tile, and optionally one fish cell per tile), a list of “tile letters” indicating the sets of letters that would appear on each tile (but no hint as to order within the tile, or tile orientation), and a list of enumerations for “fishy” phrases (such as “big fish in a small pond” or “fishing for compliments”).

This was sort of a combination of the word puzzle of coming up with the fishy phrases based on their enumerations and the letter sequences given for the tiles, and the logic puzzle of fitting the phrases to the tiles and fitting the tiles together. The word “fish” was always represented by one of the fish cells on the tiles. The answer extraction involved reading four Braille letters from the grid, treating tiles with fish on them as marked and tiles without fish as unmarked, to get the answer POLE. We found this to be a challenging but ultimately very satisfying solve.

Apocalyptic Showdown

By the time we’d finished both of these puzzles, it was around noon, and time for us to be skipped directly to the endgame—we’d slowed down to the point where we missed two round puzzles and the meta for this round. We were directed to an area on the north side of Boston Common, where all of the teams were gathering for a final activity.

The teams had been divided into two large factions, based on whether they submitted THIS WAY LIES MADNESS or THAT WAY LIES MADNESS for the Rorschach test puzzle. One faction believed “peace is our goal” and wanted to stop Cthulhu from being summoned, and the other had “pieces, our goal” and wanted to see the world burn. The outcome would be determined by a giant game of “Cthulhu Says,” in which each faction tried to bring Rubik’s cubes from one side of the field to the other and build them into a mosaic picture representing their faction’s goal. A couple of our teammates jumped into the game for a little bit, but mostly we observed from the shade, until the “peace” faction was ultimately victorious.

Artifact Meta

After the game, (almost) every team was able to grab a “diploma” which contained the content for the meta-meta. Unfortunately, somehow there weren’t enough copies (perhaps some teams took more than one?), and the supply ran out before we were able to grab ours. We were eventually able to borrow one of our neighboring teams’ copies to work on the first part of this puzzle, but returned it before we discovered there was a second part we’d need it for again.

The puzzle contained outlines that matched the main bodies of our five plastic meta artifact tokens, and dots that enumerated the meta answers, several of which became covered by the ends of the artifacts’ tentacles when they were placed within the outlines on the page. Extracting these letters in left-to-right tentacle order spelled the name of one of the bestiary creatures we’d learned about from the bestiary cards puzzle.

We got stuck here for a while, before we realized that there was a second page of the meta we hadn’t yet received. This page contained a list where each item had a bestiary symbol matching one of the beast names we’d extracted in the first part, a number (which we assumed referred to tentacle number), and some arrows (which we assumed told us to shift to adjacent meta answer dots from where that tentacle ended). Without a copy of the diploma, though, we were unable to verify these assumptions, and left the site without completing the artifact meta.

Banquet

We made a quick stop at the van to drop off our stuff, and then walked over to the closing banquet at City Winery. We enjoyed some snacks and debriefed with our friends from other teams. Near the end, Sarah and the rest of the organizers hosted a brief Q&A, where we heard some interesting behind-the-scenes stories about the development of the event. Shortly after that, we left the banquet and returned to our van to clean it up, split up all the stuff we’d accumulated over the course of the weekend, and go our separate ways.

Final Thoughts

This was awesome. We were very impressed throughout with the quality and physicality of the puzzles—nearly everything we saw involved some sort of physical component, differentiating this event significantly from online and website-based puzzle hunts and making this a really unique experience. Puzzles that leveraged information from their sites really well (such as most of the Hammond Castle puzzles and the Beaded Tiara puzzle) were also special and unique to an event like this.

The “endurance” aspect was quite a challenge for us. In general, we were satisfied with our ability to mostly keep up with most of the content throughout the weekend, but we felt very drained on Sunday morning, and I can’t help but wonder if we would have better enjoyed that round, and experienced more of it, had we put some thought and planning into taking turns taking naps and being a bit better rested.

Huge thanks to the puzzle authors, organizers, and volunteers who ensured we had a great weekend and made this event a success!