2024 Mystery Hunt, Part 0: A Puzzle, an Anecdote, and some Meta-Thoughts (not Meta Thoughts)

(This is a post about the 2024 MIT Mystery Hunt, which happened this month. Puzzles and hints can currently be found here. This post contains spoilers for two puzzles, Boosted and Greek Girl Squad.)

This year’s Mystery Hunt ran much later than the average Mystery Hunt, but it also started quite a bit later than the average Mystery Hunt due to technical difficulties. Once we realized the website was probably not going to start producing interactive content imminently, I went to the front of the room, asked for a random word as a puzzle answer, and started leading the room in a puzzle construction jam session, remembering the time the first Galactic Puzzle Hunt apparently started to be written during Hunt downtime, though that was after that team finished a Hunt rather than before they started.

We finished cobbling together a fairly approachable puzzle and typed it up, and I was just trying to figure out how to print it out and slip it under the door of some of our hallmates (Unseen, Teammate, Swarm, Frumious… big Hunt party in Building 2 this year!) when the Hunt site woke up and we suddenly had other things to prioritize.

Here’s the puzzle we wrote in that time, edited slightly because I just found a mistake that probably stemmed from my bad handwriting. It hasn’t been tested or objectively fact-checked, which all actual puzzlehunt puzzles should be, but I recommend solving or trying to solve this before reading the rest of this post if you are so inclined.

PEELS, a hastily written puzzle by Setec Astronomy

• “Benefactor, gaze into this crystal!”
• Female student at UCSD or Miami
• Frightening back end
• Gate with a keypad lock
• Metallically strengthened concrete magnifying glass
• Price to buy Christmas trees
• Slightly prefers tarts over other desserts

Here’s some spoiler space:

Have you solved it or at least made enough progress to get the gimmick? Great.

Now imagine writing that puzzle as we did, scanning through the first puzzles to appear in the actual Hunt, and seeing the flavortext for the puzzle Boosted. Boosted was our first team solve and, based on the formatting of the solve graph (which was posted on Discord but for some reason not included in the wrap-up… it’s below if you haven’t seen it), I believe it might have been the second solve in the Hunt. But that doesn’t seem fair given that we had just written almost the same puzzle. We might have been first if I hadn’t spent some time searching HQ for hidden cameras.

Moving on…

I normally write a pretty in-depth breakdown of Mystery Hunt. One could argue that with my infrequent posting, the Hunt report is the only consistent reason this blog still exists (other than to attract undeserved Bravo Award nominations; thanks to whoever threw me in there, you’re literally too kind). I have never been the type of person who runs out of stuff to say about Mystery Hunt, a driving force in my life since 1998.

So it’s strange that I can’t figure out what else I want to say about this year’s Hunt and in what format. If I were to write about this Hunt with no additional context, I’d basically be rewriting what I said last year (and to a lesser extent the year before). This Hunt was much too long for my tastes, and while it did not feel like we were given quite as many free answers as last year, it did seem like we were strongly encouraged to use the hint system pretty early. Once Hunt becomes about strategizing where to use hints and free answers, I no longer enjoy it because I feel like I’m being rewarded more for that strategy and less for puzzle-solving.

I also may have had an atypical experience because I spent more than half of my Sunday working on what seemed like a really engaging logic puzzle but eventually took multiple people all day and at least four hints to solve, partially due to our breaking the logic due to a rule misinterpretation (our fault) and partially due to the puzzle having far too many sequential ahas (not our fault). As a result, after 2pm on Sunday, I looked at every potential new puzzle with a sense of PTSD; if I work on this, am I going to be sucked into it for eight hours and miss out on everything else at the time? Instead I spent most of the rest of Sunday staring at metas, even more so than usual.

Most of the puzzles (and metas) I did work on and finish were quite good and well-designed, and even elements of the puzzle I’m ranting about were really cool, so it’s unfair to judge the entire Hunt based on that one single-puzzle experience. I honestly had a blast from Friday afternoon to Saturday afternoon, although then my attitude started to turn Saturday night as I started to realize the likely scope of the Hunt, combined with some annoyances from work and home, and then the Sunday slog just kind of pushed me over the edge. Nearing the end of the Hunt definitely would have given me a second wind, but we came nowhere near the end of the Hunt, so I stayed pretty cranky until after wrap-up.

One question that I’ve been asking myself and other people since Monday is this: Would the average Mystery Hunter (or given the diversity of teams, does the average Mystery Hunter within each solver experience demographic) prefer (a) guaranteeing puzzle content through the entire weekend, which apparently now includes Sunday night/Monday morning, or (b) achieving a sense of completion, whether that means finishing the entire Hunt or some sort of meaningful midgame. I am definitely in the (b) camp, and given my Hunt history, midgame isn’t enough for me. I realize that every Hunt goes unfinished by the vast majority of teams, but our team finished FIRST in 2016 and 2018, and in recent years despite minimal change in personnel, we’ve been multiple rounds away from the target of finishing, even with Sunday night tacked on to the Hunt period and the ability to hint and/or nuke lots of puzzles we don’t want to deal with. It feels like the solver goalposts are moving, especially if you’re not on a gigantic team (or Unicode Equivalence, who are alarmingly effective for a reasonably sized team).

But maybe I’m in the minority in prioritizing a completable Hunt over maximizing content. I thought it was a consensus opinion that Teammate’s hunt had too much content–remember that they ultimately decided to cut enough content on the fly to be its own online hunt, and still needed to give out free answers like candy to get teams through the rest–and I don’t understand why a team would follow up on that by INCREASING the number of puzzles, unless they either can’t say no to constructors or think the content flood is a good thing. So maybe there is an increasing proportion of the Hunt community that believes that having more puzzles is inherently better. Whether or not that’s the case, it also seems like an increasing number of competitive teams are publicly stating that it would be impossible to their team to write Hunt, and some have (allegedly) intentionally stopped solving near the end of Hunt to avoid it. At the current scale, that impossibility might be true. But no one should be expected to write a Hunt at the current scale, and reining things in would be a quality-of-life improvement both for the constructors and for me as a solver. Maybe it wouldn’t be for other solvers. Please debate this in the comments!

Given the nature of recent Hunts and the size of the team writing this year, I expected this Hunt to be too long by my standards, and in considering the intersection of teams likely to win a too-long Hunt and teams I have faith in to do something about scope sprint (I usually say scope creep but let’s be real here), Death and Mayhem was in the dead center of my Venn diagram, so I’m really happy that they won. I personally hope they make the changes in Hunt that I want, but they should only make them if they’re also what the community wants. If they’re not, it may just be that Mystery Hunt is finally passing me by. Which makes me sad.

If I get over that sadness and find some time, I may be back with another post about specific puzzles and metas I enjoyed, and some thoughts on story and structure. But next week I start my first semester back at work after a sabbatical, so I might just see you again this summer when the chaos calms down.

Challenge, Frustration, and Balancing User Expectations: Level99, Boda Borg, Celeste, and Mao

Bear with me… this essay is going to be a stretch for a puzzlehunt blog. (But hey, posting anything is a stretch for this particular blog, amiright?) Some of the ideas are relevant to puzzlehunts, but I’m mainly going to talk about a video game, a card game, and two escape-room-adjacent Boston-area entertainment complexes. There will be the mildest of spoilers; I will explicitly avoid talking about game content, but I may use some vague references to specific L99 and BB rooms (for example, I might truthfully say that the third phase of Bollkoll at Boda Borg is delightful).

Celeste, a critically acclaimed platformer video game, came back to XBox Game Pass this summer. I’m on leave this semester and have thus had some more time for video games in the summer and fall, so I have Game Pass and decided to give Celeste a whirl.

I finished the main part of the game today, and during that process, I died a lot. (The game tracks how many times you fail, and I think I racked up a four-digit number of fatalities.) This had the potential to be very frustrating, but the game designers paid attention to user experience. When you do die, the reset of the game is almost instantaneous… no Mario flying upside down with a music sting for the 99th time. (Years ago I played Super Meat Boy, another killer platform that benefitted from the same rapid-reload.) And every time you traverse a screen, the game auto-saves, so when you advance in the linear sequence of hurdles, you know you don’t have to worry about the last part again.

Some of those hurdles require a handful of tricky maneuvers in a row, and it did reach a level of tedium a few times. There were at least half a dozen boards where I said to myself that I was going to try three more times before giving up. About half of those times, I actually succeeded within three tries, and in several others I made enough additional progress in one of those attempts that I kept going. This game wants you to keep going. This is explicit in the game’s story, but it’s also implicit in the game design; somehow whenever it was brutally hard, it gave me just enough encouragement to keep going. (I recently found out there is an assist mode for people who want to experience the game without the difficulty, though the given difficulty turned out to be just right for me.) It was after one of these “got it on one of my last three tries” incidents that I started revisiting some thoughts I already had this summer about Level99 and Boda Borg.

If you haven’t lived in or visited Boston, you may not know what Boda Borg and Level99 are, so let me try to describe them, starting with Boda Borg since it opened first (in 2015). Boda Borg is a Swedish “questing” franchise (their only American location is in Malden, MA) where you pay for access to a building full of doors that open to themed challenge rooms. Like a puzzlehunt puzzle, these rooms rarely come with instructions; your multi-person task might involve solving puzzles, pressing buttons at the right moments, or a ninja-warrior-style climb across the room without touching the ground. (Before you scoff at the physical aspects, I want to emphasize that I DISLOCATED MY SHOULDER on my first Boda Borg visit, in a quest called Jungle. I’ve been back, more cautiously, three times since.) Depending on whether you do what you’re supposed to do, you’ll be served a red failure screen and buzzer, a green screen and admission to the next room of the quest, or if you’ve completed the last room, access to an ink stamp to add to your quest card. Quest names range from the evocative (Pirates, Alcatraz, Spook House) to the abstract (Wumplefrump, Nostalgia, Tough Tougher Toughest).

The easiest way to describe Level99 (in Natick, MA, with a second location opening this year in Providence, which is super-exciting) is to steal a comment I saw on Reddit that called it “real life Mario Party.” The second easiest way is to start with the Boda Borg description above and emphasize some differences:

1. At Level99, all the quests (called rooms) are one room, and instead of succeeding or failing, you can earn 0, 1, 2, or 3 stars depending on your performance. Rooms are grouped into categories, so they have names like Space Void, Space Nebula, Retro Pinball, and Retro Vinyl.
2. Instead of stamping a card, you tap into each room with an electronic wristband which tracks your progress on all rooms (progress that is saved between visits).
3. In addition to the rooms, Level99 includes “hunts” (memory quizzes about the local art throughout the building) and “arena games,” one-on-one games such as a Pong climbing wall and a race to press colored panels in order.
4. Although one of the room categories is “mystery,” meaning part of the challenge is figuring out what to do, I personally don’t think any of the room objectives are particularly unclear after two or three attempts.
5. Boda Borg has a taco bar on the premises, and Level99 has a brewery. (Doing well in Level99 rooms earns currency you can exchange for free merch and/or free pretzels or french fries at the restaurant.)

Last month I went to Boda Borg for the first time since before the pandemic (fourth overall), and Level99 for the fifth or sixth time. I like Boda Borg. I LOVE Level99. The progress system might be enough to make the difference by itself; I’m a sucker for achievements and incremental progress, and just writing this makes me want to go back and increase my star count. (I desperately wish they would let you access your account outside the facility… if/when you do go, make sure to tap your icon after tapping into a room and before entering, because there’s a ton of account interactivity available that’s not immediately obvious.) But in terms of the quests/rooms themselves, the Level99 rooms are almost all designed with the care that Celeste is. Some of the Boda Borg quests are too, but some of them are designed like Mao. I’ll get to that in a bit.

As I mentioned earlier, two of the things that make the challenge of Celeste bearable are the frequent saving and speedy restart; these features ensure that while you may be doing a lot of repetition, it’s mostly productive repetition. Since L99’s version of quests are only one room, they avoid the undesirable experience of failing the third room and then having to repeat two previous rooms (possibly with some waiting around if another group is in the next room) before you can try again.

(One of the newest rooms, Wavelength Flux.0, is arguably an exception to this; we put a lot of effort into getting to the very last of a long series of challenges, and we ran out of time before we could figure out what to do with it. We’d now need to do a lot of the same work over again in order to possibly just get stuck again. I feel like this room would benefit from some sort of progress save, but that’s not consistent with the L99 structure. Starship Evacuate also has a sequence of tasks you have to repeat, but the tasks change more from attempt to attempt unless you have a group of six.)

But there’s another thing that kept me going in Celeste, which is that on every replay I knew what I was expected to do. Sometimes it took some thought to decide exactly where I should be dashing/climbing/hanging on, but the objective was generally clear. As I noted in my rundown of the differences between BB and L99, the objective is also generally clear at L99. BB, not so much. The peak of this is where I think BB stops working with their customers and instead works against them.

In at least two quests, Spook House and Potions, there is a room (notably) not the first, where you can passively fail; you walk or crawl in and encounter a buzzer and the failure screen. That’s fine, we’ll try it again… but my group has done each of these probably a dozen times and tried many different things with no additional progress. If every attempt was auto-loading as fast as a Celeste level, I might be more willing to keep spamming attempts. But both of these booby traps occur after a non-trivial physical activity, and repeating that activity just to get the same mysterious rejection is exasperating. And unlike a typical Level99 room, where a failure is often accompanied by an on-screen comment summarizing what went wrong, these two rooms have the same complete lack of feedback.

Just to keep this post somewhat relevant to the blog theme, puzzlehunt design has a similar pitfall, in that some puzzles call for a task or extraction that requires a lot of work with no confirmation that you’re doing the right thing until you’ve done the right thing. These are sometimes called “guess what I’m thinking” puzzles, since without guidance, you have to either read the constructor’s mind or fail. Modern puzzlehunts (at least the hardest ones) have undergone a troublesome trend where few if any solvers solve some of the puzzles without hints. I think some constructing teams don’t mind this, because they enjoy engaging with the solvers through hinting. But if everybody needs hints to progress, the hints should have been in the puzzle. Video games can have DLC, but DLC shouldn’t kick in before the main storyline concludes. Stepping off my soapbox.

Level99 and Boda Borg also have some form of hinting. Level99’s is more formal, as you can access hints from the pre-room tap-in interface. (I’ve only ever done this once, for Pirates Brig, which effectively told me, “No really, we want you to do the thing you think we want you to do.” I have zero stars on Pirates Brig.) Boda Borg has attendants who apparently sometimes give hints, I’m told. But when they first opened, I got the sense that this was not a thing. Even now, the BB decor includes slide shows with memes about how everyone at Boda Borg fails, laughing about how hard their quests are. I imagine this is meant to normalize failing, but what I’ve learned from Level99 and Celeste is that you normalize failing by minimizing the discomfort of the failure process. I think Boda Borg fails to do this in the rooms in question (and some others), and I think their design mentality violates my trust as a player. I don’t want to engage someone I don’t trust in a subjective hinting interaction.

So if these Boda Borg rooms don’t feel like playing Celeste, what game do they feel like? That’s where Mao comes in. I believe I discovered the card game Mao through math competitions where people played it after hours. It’s an Uno-style card game (played with a standard poker deck) where there are secret rules you’re meant to figure out while playing, and breaking these rules is penalized by being given extra cards and no explanation. In theory, this could be a good foundation for a mental challenge. In practice, every social group that I’ve ever played with Mao was a school-aged clique that used it as an excuse to laugh at newcomers. Make it through the initiation and you will share our secret information that we can use to torment others! Or you could just talk to people as equals and make friends. Sorry, I think I’m processing something from decades ago in real time.

Anyway, I don’t get the sense that Boda Borg wants every visitor to figure out Potions on their first visit… I’m not actually sure they want *every* visitor to ever figure it out. A wise constructor (I think Mark Gottlieb, but I may be misquoting and/or misattributing) described a puzzle as a battle of wits between the constructor and solver that the constructor intends to lose. When I got frustrated playing Celeste, I always got the sense that things were set up for me to eventually win, and that kept me going. A challenge where the player is not intended to win, like Mao or Spook House, might be fun for the challenge-maker, but if the challenge-maker keeps frustrating their audience, they’re going to run out of audience.

Now, I realize I’ve been slagging on Boda Borg throughout this essay, and I want to be clear that I still encourage you to go there, and there is lots of entertainment to be had. While some of their content seems unfair to me (and whether or not it’s actually unfair, the net result is unpleasant), I’ve really enjoyed others. In my most recent trip, some of the new quests we completed and enjoyed were Wumplefrump, Access Denied, and especially Eye of the Storm, which for me benefits from the multi-room model in a way other quests don’t. But I went to Boda Borg once this summer, and Level99 three times, and I think it’s likely to be that way in the future. Level99 feels fun and sometimes sadistic, and Boda Borg feels sadistic and sometimes fun. I would recommend trying both. And also Celeste. But not Mao.

I have one more observation to tie things together, but it involves significant Celeste spoilers, which I said I’d stay away from. So if you intend to play that game and don’t want to know how it ends (for some definition of ending), step away from the blog.

All clear?

Okay. Here we go.

The main storyline of Celeste ends with Chapter 7 and an epilogue. Then an unexpected Chapter 8 kicks off. After a few screens teasing a different dash mechanic, I encountered an impenetrable wall. Based on my experiences in previous levels, I assumed there would be a straightforward way to move through, but nothing worked. So I went back a screen to a character I’d just talked to and talked to them again, at which point my character asked about the wall, and was told something cryptic about retracing my steps through the entire game.

So that’s when my Level99 game turned abruptly into a Boda Borg game. It’s also when I stopped playing.

Microsoft and Brown

Hi all! It’s a pleasant Friday in April, and Brown is about to commence Spring Weekend, which makes me suspect that my office hours are going to be very quiet. Maybe I’ll be proven wrong [Edit: I was], but it seems like a good time to check in about some recent and future puzzling.

First and foremost, let me provide a plug for the upcoming Microsoft Puzzlehunt, which is online and open to the public next weekend (May 6-7). I’ve never participated in MSPH for real, but over the last few years I’ve had opportunities to test puzzles for the event. Usually due to the crowdsourced nature of the event (and because I’m often not available for long stretches of time), this mostly feels like grabbing individual puzzles from a big stack. But this is the first MSPH in a long time that’s been entirely written by a cohesive group, and I was also able to carve out more continuous time, and I have to say I *really* enjoyed it. I will obviously spoil nothing, but it felt on the nose in terms of difficulty relative to team size, and the puzzles, structure, and theming were all delightful. As I understand it, this was originally going to be on-site and closed to partially-Microsoft teams as is tradition, but it had to switch to online which led to the open registration the event has now. Take advantage of this! It’s worth your time.

It’s not often that in-person puzzlehunts come to me, but I was really psyched that there was a Brown Puzzle Hunt this year! And written by over a dozen current puzzlehunt-interested students, which is not a demographic I realized currently exists in my workplace! Unlike years past, that demographic doesn’t seem to exist in the math department; I tried to wrangle some grad students to form/join a team, but nobody bit. But talking about the event on Facebook, Wil Zambole, Jen McTeague, Dee Williams, and Andrew Esten all expressed interest in coming to campus from various definitions of out of town (Chicago, for example, is more of a trek to Providence than Boston is, not that it’s a competition) so I ended up on an adult ringer team. Which felt a little weird for a hunt on home territory… In fact, at first I didn’t know whether I should be on a team, since maybe this was for students? But then a shadowy figure delivered me a personalized invitation at the end of my multivariable calc course one day(!), and I figured that’s as clear a signal I was going to get that I was welcome.

I really didn’t know what to expect from the event itself, and the initial description suggested it would be BAPHLesque, so even though the event was Saturday/Sunday, I only made plans to be there on Saturday (with Simon home on weekends, that’s likely the most I was going to contribute anyway). It was not BAPHLesque from my perspective… if anything I’d say it was close to Puzzle Potluck? There were definitely some puzzles, even in the first round, that seemed crunchy for an event happening on a campus where puzzlehunts are not the norm, and some very technical (but nice!) metapuzzles. I hope this event happens again, and personally, I’d like to see some more newbie-friendly content in the first round; after that if you want to ramp things up, that’s cool. But in some ways, this felt more geared toward the ravenous internet audience than the undergrads that I’d love to see addicted to puzzling.

Despite my nitpicking about calibration, I enjoyed this hunt a lot as well, and my team was a lot of fun to solve with. And maybe it’s a combination of my preexisting awareness of Blueno and my predilection for dark themes, but I noticed and appreciated the round art more than in most hunts. In terms of individual puzzles, the most memorable ones for me were Financial Crimes, which is one of the more immersive logic puzzles I’ve ever solved (information in classic logic puzzles tends to feel a bit clean and contrived, and this was… not that), Remixers (very up my alley), and the Lamp meta. I was also pleased to see Community Photos, having written the similar-in-spirit Alma Mater in the 2002 Mystery Hunt. Which was over two decades ago? Excuse me while I send myself out to sea on an ice floe.

I have a few more things I’d planned to talk about, but I ended up spending the majority of these office hours actually doing my job (go figure!), so I’ll hit “Post” and try to find time to resume later.

2023 Mystery Hunt, Part 2: More is Less

(This is a recap/review of the 2023 MIT Mystery Hunt, which happened this month. Puzzles, solutions,and solving stats can currently be found here. This recap will contain spoilers.)

Before I start talking about length issues, after posting Part 1, I remembered something else I loved about the pre-AI portion of the Hunt; the automated vote-on-a-response team interactions were brilliant. They were a great way to enforce a team bonding experience and immediately advance the story without requiring the constructing team to invest live person-hours, and I think my team laughed out loud much more than we would in a typical live interaction. The idea was great and the writing was great. I’m torn because this felt really specific to this year’s theme and artistic design, but at the same time I want everyone to steal it.

Now.

In the preamble to a recent Hunt writeup, a member of Cardinality amusingly said, “I am not a titan of the community and I will not share anecdotes about how this puzzle reminds me of this meta back in the 1926 Mystery Hunt where they gave us 3 rocks which we had to bang together in the right way to make the correct fires.” Of course not! That’s MY job. So let’s start with a brief history (at least within my time spent with Mystery Hunt, which is up to 25 years now) of Hunts that were too f***ing long.

If you analyze Mystery Hunts that went long from at least one team’s perspective, you’ll find that they generally fall into one of two categories:

• Hunts where one or more metapuzzles end up being killers and block teams’ progress for disproportionate periods of time. Let’s call these mettlenecks (short for meta bottlenecks).
• Hunts where the act of solving puzzles to get to the metapuzzles was so overwhelming that the construction team has to modify Hunt procedures so that even the teams in contention get credit for them without solving them entirely on their own. I don’t have as catchy a name for these, but I’m going with forcefeeds.

I’ve been on the constructing end of my share of mettlenecks, including 2009 (Zyzzlvaria, where different teams got stuck on different one-meta-left situations, which made for an exciting finish), 2019 (Holiday Forest, where the last two metas were very hard to solve without almost all of the feeder answers), and 2005 (Normalville, where one team had nothing left to solve but one metapuzzle for 24 hours).

The interesting thing about a mettleneck is that often it gets bad reviews from the lead teams that spent a lot of time staring at the same meta (which is booooooring), but disproportionately good reviews from casual/middle-tier teams, because they’re able to proceed through a lot of the Hunt before a winner is announced, as there’s not much bottlenecking before the bottleneck. I do think this is still a negative result despite positive side effects, because while it’s good for a majority of teams, you don’t want anybody to work hard solving your Hunt and then end up frustrated.

Forcefeeds, on the other hand, seem to rotate into Mystery Hunt periodically. There have been a lot of jokes recently about years that end in threes, because three of the most notable forcefeeds are 2003 (The Matrix, which was ahead of its time in that it would NOT seem too long today), 2013 (Coin Heist), and now 2023 (MATE? Puzzle Factory? Relentless AI Assault?). One of the other most notorious examples was 2004 (Time Bandits), which managed to follow up Matrix with something even more forcefeedy. The same team that won in 2003 also won in 2013 and went in determined NOT to repeat mistakes of the past, and I feel that was a success. I was on the writing team for 2014 and not 2004, but I can’t take credit for the changes… the leadership of that team made very good choices, and I mostly served in an advisory role. I did insist that we keep our endgame operating in full for as many teams as possible once the coin was found, even though it was a pain in the ass, and I’m very glad we did so. 2008 (Murder Mystery) had some forcefeed elements as well, though I don’t remember the details as vividly.

The hallmark of a forcefeed Hunt is that at some point, the construction team realizes things are not proceeding on pace, and that something has to be done. In 2003/2004, that involved hinting puzzles liberally when teams reached even the smallest bit of resistance; in 2003 we were actually assigned a dedicated in-HQ hinter for an extended period of time. In 2013 and 2023, the constructing team took things one step further and gave out a large quantity of what I call “nukes,” the ability to get free answers for puzzles without any idea of how to solve them. I want to be clear that given the pace of both Hunts, this practice was probably necessary, since the Hunt would have gone well past Monday if teams were going to solve what they were expected to solve. But I want to highlight why you don’t want to find yourself in a situation where this is necessary.

When Setec first earned a couple of nukes, we met and talked about our strategies for using them, and on a related note, what we wanted to get out of Hunt. I said that my priority was solving metapuzzles and opening rounds, because that’s what I find exciting. Several other team members agreed. But cut to 24 hours later, and opening rounds wasn’t fun anymore, because we weren’t doing it by solving puzzles… we were doing it by giving up on puzzles and pressing buttons. When I saw the notifications that Ascent and Conjuri’s Quest, I had a “meh” feeling I’ve never had when opening new rounds before, because I didn’t feel like we’d earned access.

This is why I didn’t have fun with the AI rounds. It seemed pointless to work on a puzzle when we could just as easily flip a switch and disappear. At one point we opened Flooded Caves, which is a set of seventeen Cave logic puzzle variants. I love abstract logic puzzles, and one of our captains, Tanis, asked me if I was going to solve this, or if we should nuke it. We counted the puzzles and realized I’d probably spend the rest of Hunt solving it, or worse, I’d spend hours on it and then we’d give up and get the answer for free anyway. We nuked it immediately. We basically spent Sunday looking at puzzles, deciding whether they seemed approachable enough to bother with, and often deciding no. This wasn’t just us… the Hunt stats indicate that the entire Ascent ROUND (meta not included) had 18 successful solves, and 147 nukes. Teams didn’t complete this round. They took an elevator that went past it.

Even when we’re stuck on a meta I usually enjoy Mystery Hunt, but this year’s Sunday afternoon was the second time I remember legitimately not having fun. The other time was 2013, and with both data points in hand, I now assume the free answers are to blame. Solving feels pointless when puzzles are spontaneously combusting around you.

So what do I think Teammate could have done to avoid this? It’s easy for me or for anyone to criticize from afar, because writing Hunt is time-consuming and sometimes thankless work, and tuning/pacing Hunt is extremely difficult (as I stated above, I’ve been on teams that have messed it up, though more frequently through problematic metapuzzles). But we grow as a community by sharing insight, and I’ve been around long enough to have a lot of perspective, so here were some of my observations.

Many individual puzzles were bigger than they should have been. I already mentioned the seventeen caves above that caused me not to attempt to solve any of them. I also mentioned the number 147, which coincidentally happens to be the number of morals you had to identify in Moral of the Story, after finding a message in 147 typos. It’s unlikely that any single solver wants to do anything 147 times. Hunt is, of course, a team activity, so you probably won’t have a single solver doing it. But that still means you’re devoting multiple people to stare at a single puzzle for an extended period of time. When you have lots of puzzles that are really big (the one with the quiz bowl questions also comes to mind, which looked like a fun idea iterated way too many times for me to want to solve), you are spreading non-giant teams thin, and the vast majority of teams solving Hunt are non-giant.

Testing may have needed to take into account that most teams are not like Teammate. This is very similar to what I said in 2013, when I felt like Manic Sages wrote the perfect Hunt to be solved by Manic Sages… who were unfortunately the only team not solving that year. After the solutions were posted, I saw a lot of people reference the authors’ note for Terminal, which begins, “This puzzle being solvable at all was honestly a huge surprise to me.” That is a MASSIVE red flag, and the note goes on to explain that the puzzle was made harder because testsolvers got better and better at solving clues. I am curious how long this process took, and whether it was considered that in practice, this would only be one of many puzzles teams were contending with at once. For what it’s worth, we thought Terminal was a fun idea, and we had at least a dozen people put hours into trying to solve it. After expending those hours, we still had less than half the grid filled. We nuked it. Moral: Don’t make puzzles harder because your testers are surprisingly brilliant; your testers won last year’s Mystery Hunt. (There are similar “let’s make it harder” stories from 2004 and 2013, which is not a coincidence.)

Number of puzzles is not a good gauge of size/difficulty. The number of puzzles in Hunt has oscillated, but overall it’s grown close to linearly over the last few decades… that might be okay, because solving resources and ability are also growing. But the definition of what a puzzle is is also growing; the puzzles I wrote in 2000/2002/2005 would barely qualify in the modern era. And with really chunky online hunts like Galactic and Teammate and QoDE and Silph pushing boundaries throughout the year, people’s expectations for how long a puzzle can take and be reasonable are expanding. The problem with this is that if the number of puzzles is O(n), and the size of a puzzle is O(n), the total size of the Hunt is actually O(n^2), which is a terrifying rate of growth. The team that won has been reported publicly to have over 160 people, though someone on the team reported that they had about 170 unique solvers, and only 120 of them were active solvers. ALL THREE OF THOSE NUMBERS ARE TOO BIG. And it is vital that TFKA…TTBNL does not write a Hunt with a team of that size in mind, because few of those teams exist (and in my opinion, none should).

The first part of the Hunt did not feel friendly to casual solvers. I was genuinely surprised at wrap-up when Teammate said one of their goals was to support casual teams, because I found the first round and meta to be much less accessible than the intro phases of recent Hunts. The first puzzle I worked on, Museum Rules, immediately subverted the expectation that the copy-to-clipboard feature would consistently work, and the aha was very challenging (and even once you got it you still had a bunch of nontrivial superimposing to do). Apples Plus Bananas required two ahas (you need PLUs and you need to get prime totals) and turned into a significant logic puzzle if you assumed those things, which you could not necessarily confirm before solving. I think these are tough puzzles as an experienced solver, and I can’t imagine how an MIT frosh who wandered in and wants to see what all the puzzling was about would navigate them. As for the meta, I recognized what to do with it quickly because I’ve solved many Anglers/Numberlinks. Most newbies haven’t. I encountered very few easy puzzles in this Hunt in general, and I was surprised that the ones I did find most approachable weren’t generally at the beginning.

The recap over at Fort & Forge predicts that my Part 2 thesis will be that Teammate should have cut Part 3 of the Hunt entirely. Actually that was my Part 1 thesis, so maybe I was too subtle about it? As it stands, if Part 1 and Part 2 were going to be what they ended up being, then yes, I think Part 3 should have been cut (or more specifically, replaced with a single round or more involved endgame that introduced and neatly resolved the multiple-AI story). I think if Teammate really wanted to have Part 3, all the parts needed to be smaller, both in terms of number of rounds, number of puzzles, and the size/complexity of the puzzles themselves. I get the desire to mess with structure… Zyzzlvaria had a second phase that was all about messing with structure, and while Holiday Forest only had one structural innovation, carrying it out to the extent we wanted caused us to include more puzzles than we should have. But there were lots of fun structural things in the Museum and Factory, and frankly, I found the Innovations and Factory Floor metas far more interesting than anything I saw in the AI answer format gimmicks or metas. But even if you really like Part 3, I don’t think it’s defensible to say that Parts 1, 2, and 3 all fit into this Hunt as is, because the stats show that teams didn’t solve Part 3.

At the beginning of this rant, I pointed out that forcefeeding is a pattern that occurred ten and twenty years ago. One team responded to it by making things even bigger (and arguably more poorly edited). Another team–okay, the same team ten years later–looked at the issues and intentionally tried to fix them. I’m a little worried that, because this year’s winners are a giant team, they will all want to contribute and will write something that can only be solved by giant teams. Giant Hunts are what cause first-time constructors to be scared of winning; they assume they also have to write something giant. You don’t. Setec has had a writing philosophy for years to “write for the middle.” During Wrap-Up, someone said Mystery Hunt is growing and growing, and so it needs money or it will die. Donating to Hunt is a good thing, but please understand that if it does not keep growing, it will not die. In fact, it could use some shrinking and breathing room.

Despite my criticisms, thank you to Teammate for making something creative and cool. I thought the last Teammate Hunt excelled, especially for an online puzzlehunt, in terms of cohesive puzzles, art, and story, and I found the same to be true for the portion of this Hunt I was able to enjoy before time intervened and puzzles started disintegrating all around us. I know from experience that I can have a lot of fun your puzzles and structures, and I hope you recover enough to write another Teammate Hunt soon. But since you effectively just wrote 1.5 Mystery Hunts, I won’t blame you if you need a break.

2023 Mystery Hunt, Part 1: Less is More

(This is a recap/review of the 2023 MIT Mystery Hunt, which happened this month. Puzzles, solutions,and solving stats can currently be found here. This recap will contain spoilers.)

Erratum on Tue 1/17/23, 4:36 PM EST: The narrator of this story may occasionally be unreliable.

Mystery Hunt was back on campus this year, and I applaud Teammate for the Herculean organizational effort that that must have been involved. I know from working at a university myself that over the course of the pandemic, a lot of administrative procedures have been interrupted and complicated by staff turnover and reluctance to reinstitute stuff that wasn’t necessary when classes were remote. All things considered, MIT is very supportive of Hunt (if you don’t think they are, name another college that yields this much classroom space to an event they don’t directly supervise that doesn’t directly give them money), but getting everything you want when running Hunt is still a delicate process in the best of years. This may have been the worst of years, but apart from clearing out of HQ overnight, things felt shockingly back to normal.

Jackie and I decided that Simon is still too little for on-campus cameos (particularly with XBB raging in New England; I still seem to be healthy thanks in part to Setec’s requiring everyone to test every morning, knock on wood). I essentially asked for Mystery Hunt for Christmas, so effectively I got a weekend pass to stay in Cambridge for the entire Hunt, and Jackie showed up on Saturday when we could get babysitting suport during the day. I’m hoping that next year we can find a way to have us both there longer, because I especially missed solving with her during the quiet morning shifts (though I also missed large chunks of those shifts anyway, for reasons I’ll describe).

I parked at the same Red Line garage I’d discovered as my go-to Hunt Weekend access point in 2019 or 2020, got to campus bright and early, and got lucky twice in that my hotel room at the Marriott was available early, and I was able to find a kiosk quickly to print a physical alum ID. Unfortunately, neither my mobile ID or physical alum ID was able to open a single door all weekend. I’ve got to get that looked at before next year.

Setec has used a non-classroom Stata HQ for years, but since 2020, our connection has been severed, so we were in the classroom HQ pool for the first time in ages. Even if we’d still had our previous home base, maintenance work in Stata would have displaced us anyway. I had some trepidation about how much the change in location would faze us, and the answer was surprisingly not much. There was a classroom next to us reserved for The Team… To Be Named Later, which they apparently decided not to use, and which they even more apparently (by now) didn’t need. If our classrooms had been too crowded, I assume we would have hijacked their room once it became clear they weren’t using it, but since we didn’t, I guess we were ultimately fine. We also had an overnight HQ booked at the Marriott, which was very helpful (we had arranged for a suite, which flooded shortly before Hunt(!), and the staff at the Marriott was gracious enough to find us some meeting room space and waive some of the rules that had caused us to choose the suite in the first place). That worked out nicely and definitely soothed some of the sting of having to relocate a couple of times per day.

When Hunt runs smoothly, I tend to focus on metas whenever possible. Maybe that’s just an internal yearning for glory, but I also just like getting big ahas and solving around things. When we were solving 2021 and 2022 remotely mostly in Zoom breakout rooms, it was often hard to find the people who were thinking about metas. One of the best things about being on campus again was being able to hover near the front of the room and get involved with the meta brainstorming easily. We only solved two metapuzzles on Friday, but I ended up doing a lot of work for both of them on paper. Our first meta fell after someone found the restaurant connections for Atrium, and then I solved the Waiterlink (which probably could have been done in a spreadsheet, but it was faster just to draw lines).

And I also ended up being the pencil guy for a small group collaboratively labeling molecules on the Science meta. The latter is in a category of metas I often don’t appreciate, since it’s much more of a self-contained puzzle that happens to have the feeder answers as givens than a final stage bringing the answers you’ve gathered to life. But among metas of that ilk, I thought the Science meta was really nice; the examples given did a good job gradually revealing and confirming what color meant, what order meant, what interactions were happening, and so forth. Visually it reminded me of the Feynman meta from 2013 (maybe not the best meta collection to emulate), but I found this solve path much smoother. I usually try to be one of the team members that goes to sleep early and wakes up for the morning shift. We started making Science progress when I was about to leave, and it kept me up for at least another hour and a half.

I also typed STOP at Mate earlier in the day, but not until we reached the point where puzzles weren’t loading. I had noticed much earlier that there were weird square sequences but assumed it would come up in the Science meta (which wasn’t open at that point), and when this opened a new level, I was very pissed at myself for not trying anything an hour before I did.

Hunt is not all metapuzzles, so for the record, some of the early puzzles I enjoyed working on near the beginning of things included Inscription, Museum Rules, Natural Transformation, Scicabulary (someone mentioned “Leakfast,” and I said, “Hey, I think I wrote that Shinteki Puzzle of the Month years ago!”), and One of the Puzzles of All Time, which was morbing great.

I wasn’t as consistently involved with metas on Saturday; Art was solved entirely while I stepped out of the room for five minutes, and I never really got what was going on with World History. I did help get the aha for Natural History, but I was mostly pleased I could guess the appropriate wrestler who’d held the KOD 6-Man Title without looking him up. Fun non-metas I worked on (some of these may not have been on Saturday) included Baking Bread, Redacted Recipes, and This Puzzle Is Just Another Regular Cryptic. And Think Fast was neat, though we got incredibly stuck on the final puzzle, even after winning the game step and the in-person testing step. Sometimes things that have three parts should only have two. Foreshadowing.

And one more time I did not get to sleep as intended, becasue around dinner time someone figured out that the factory gizmos altered Innovation puzzles. This was a fantastic mechanism, and once we opened the meta, I got very sucked into it. Unfortunately, we made some spreadsheet errors when solving the wordplay bit and placed old and new letters in inconsistent columns, and so we didn’t see the first phrase we should have extracted. This means that instead of the likely solve path (submit that phrase, be told your gizmos are wrong, fix the gizmos) we followed the not-recommended Setec solve path (don’t see a phrase, figure out all the gizmos, spend a long time trying to extract an answer from the gizmo positions, eventually fix the spreadsheet and submit, never seeing a message telling you your gizmos are wrong). The bright side of this is that while analyzing the gizmos, I drew a graph almost exactly like the one we ended up being given in the Factory Floor meta (same orientation and everything!). We’d also already written down the Factory answer-gizmo sequence, so we had everything in place to solve the Factory meta; that kept me up a solid three hours later than desired, but as a result, we did have the fastest unlock-to-solve speed of any team in the Hunt for the Factory Floor meta.

Sunday morning, after a decent but probably insufficient amount of rest, I came back to HQ and soon we opened the Basement meta, despite having opened the Office round first. We caught onto the recycling and plastics code theme quickly, and after a few false starts at how to approach things, we managed to solve the meta and discover the drives in the basement. After that, we focused on the Office, used a few free answer chits and solves to open the meta, and solved it in a big group, following a correct hypothesis that got much better once Josh Oratz proposed that SOLID YELLOW was a pool ball rather than a naval flag. Now having gathered all the key elements of the story beat (Mate’s having a rough time! There are mysterious hard drives in the basement! There are other weird AIs about!), we solved the four reactivation puzzles and initiated a final runaround in which we used the new AIs to improve the factory and brought our Hunt in for a gentle landing on Sunday afternoon.

Yeah…

…except the end of that last paragraph isn’t true. What actually happened is I woke up Sunday to be told that there was an entire third level of the Hunt we hadn’t unlocked yet. We did solve both the Basement and Office roughly as described, although we were given Reactivation to solve before we finished the Office (apparently because fewer metas were now being required for that step), which made finishing the Office very anticlimactic from a story perspective, and possibly unnecessary for advancement. But more importantly, completing Reactivation didn’t trigger the endgame. Instead it opened FOUR MORE ROUNDS of puzzles, many of which felt long for the sake of being long, arcane for the sake of being arcane, and I’m sorry to say, poorly edited (though probably not for the sake of being poorly edited).

I am speaking purely for me, and not for Setec or for the Hunt community at large. But for me personally, the AI rounds took what could have been a really great tightly constructed Hunt, and ruined it. While I have considered trying to retroactively brainwash myself into believing we finished Sunday afternoon, I’m not going to do that, both because I don’t have access to the right medications, and because I think there are some vitally important lessons to be learned by the Hunt community from this year’s Hunt. I apologize in advance that this will probably involve some harsh criticism. I again want to emphasize that I really liked most of what we encountered in the Museum and Factory. And also Dispel the Bees. But I value quality over quantity (most of the time), so in the next post, we need to talk about why more can sometimes be less. After I calm down a bit.

Seven Months Later…

I realize it’s getting repetitive for me to start every post with an apology that it’s been so long since I posted, and an excuse that it’s hard to find time to write thoughtful posts when you have both a job and a child. I’d like to lead off with something more interesting, but it’s hard to find time to write thoughtful introductory paragraphs when you have both a job and a child. (Hats off to those of you who have already been doing this for years. I can see why most/all of you don’t have puzzlehunt blogs.)

Last time I posted here, it was “Part 1” of my Mystery Hunt recap, and you’ll notice this isn’t Part 2. I had intended to spend Part 2 talking about specific puzzles, and I had a list of the puzzles I was going to write about. But then in a conversation about the absence of Mark Halpin’s Labor Day puzzle suite last year (and how much I missed it), someone said something along the lines of, “Well, you should definitely solve Trickster Tales!” So I printed that puzzle and decided I should solve it before writing Part 2… and then I didn’t solve it… and then I still didn’t solve it… and I still haven’t solved it seven months later, in a future where half the puzzle titles I jotted down are now unfamiliar. So Mark Halpin stole your Part 2 post, folks. I’ll just say there were a bunch of fun puzzles I solved when I was able to focus on solving, and I hope we figure out a plan for 2023 that allows us to maintain that focus for longer periods of time.

Rather than the expected Mystery Hunt follow-up post, I’m checking in here to touch on a variety of topics, both so I can weigh in on them, and also so that the next time I post, I don’t have the pressure of writing my first post in seven months. Unless I don’t post anything until March. Which is entirely likely.

* These days I’m not finding much/any time to construct puzzles or write articles about them, but I’m still solving everything I can! For Huntinality 2.0, we expanded our smallish team from last year to a larger Mystik-Spiral-Alice-Shrugged mashup team, and that’s becoming a fun dynamic to solve under once a year. I attempted to solve both Edric hunts (Truzzle and Puzzle) mostly solo, and I enjoyed portions of them but didn’t make it to the end of either; in a way, having them so close together made neither demand as much focus, and I lost momentum at some point during both. I did finish REDDOTHunt (with a Mystik Spiral team under Mystsick Spir-Ill, since I was getting over a cold), and that included Manga, which might be my favorite hunt puzzle I’ve encountered this year. I mostly came here just to link to it, and I was dismayed to find that the REDDOT site seems to be down, so that doesn’t appear to be possible. But if you can find it, you should really solve it. It’s very artfully presented, and it has fun stuff to do right from the beginning, a fantastic aha, AND a jaw-dropping finale.

* I’m really looking forward to two upcoming hunts that are on my wavelength in terms of difficulty. The Galactic Puzzle Hunt is happening on the second weekend that my family is visiting Cape Cod. Since a fiendishly difficult puzzlehunt is more my jam than beach days, I’ve received clearance to drop everything and focus on GPH on Friday night, though Saturday I’ll lose a lot of time to packing/travel. If Saturday is necessary… after all, it’s a one-hour exam, what could go wrong? And then the weekend after (while Galactic is still technically in progress, though I hope Killer Chicken Bones will be done by then), it’s Cross Purposes, the return of Mark Halpin’s Labor Day puzzles after a one-year hiatus. You may remember Mark as the guy who stole your Part 2 post, but solve his puzzles anyway… These suites look P&A-length at a glance, but the puzzles themselves tend to put up much more of a fight. This should be a nice last hurrah before I disappear back into teaching for the fall semester. (I’d also be looking forward to Teammate Hunt, the sequel to my favorite online hunt of last year, but I hear they’re doing it in January this year.)

* Speaking of P&A, in the last issue Foggy mentioned that he intends to wrap up the magazine some time after its 100th issue. I understand and respect that decision… I can’t even write a blog post every few months, so it continues to boggle my mind that Foggy can produce a one-round hunt PLUS extra puzzles every two months like clockwork, along with a Puzzle Boat arguably longer than all of the issues for that year combined. I can understand the desire to focus on Puzzle Boat and/or do something else to keep things fresh. Of course, after repeatedly wondering when my top ten streak (which goes back to Issue 1) will come to an end, now I feel an intense urge to make it through the end of the run. We’ll see.

* Finally, there’s been news in MIT circles recently that MIT is moving away from their philosophy of being an “open campus.” For anyone who hasn’t visited the Institute (my undergrad alma mater), the academic buildings, many of which are interconnected, are open by default to public visitors, so anybody with a pulse could walk down the Infinite Corridor (which in some cases is the most effective way to get from some points to others in Cambridge, especially if it’s raining). During the COVID era, access was much more controlled, presumably to facilitate a vaccinated campus. I’d assumed that card access was simply added to locked doors, but I’ve now been told by friends who have been on campus that there are turnstiles, so a person can’t simply escort another random friend into a locked building. I bring all of this up because MIT has quietly announced that they are permanently shifting to this “closed campus” model; even after COVID, if that’s a thing, the main campus buildings seem like they will continue to be turnstiled and limited to people with MIT credentials and registered visitors.

This philosophical shift is significant for the MIT community as a whole (I know a petition was in draft form last time I checked, and I intend to sign it and circulate it), but in the context of this blog, the obvious issue is relevance to Mystery Hunt. At press time I have no idea if the 2023 Hunt is intended to be in person, and I had already heard many people speculate that it might never be again, but certainly this decision complicates a theoretical future Hunt that would occur on campus and be open to all. There are a lot of unknowns that affect potential access logistics. For example: (1) Do alumni IDs (which exist) grant campus access? (2) There is a process for students/faculty to register visitors… how many visitors can they register and for what purpose? (3) There is a process for events to register participants (or more specifically, give participants the opportunity to register themselves)… how many participants can they register and for what purpose? It seems clear that MIT is now going to be aware of exactly who is participating in an on-campus Hunt, and who they will allow is an open question.

With all that in mind, I’ve seen a few social media posts that this is likely the death of Mystery Hunt (or more diplomatically, of “Hunt as we know it”). I get a bit triggered by those posts, because when I was a Mystery Hunt existed in a form that would have had little trouble operating under the Institute’s radar. This is not to imply it was a secret, but it would likely not have set off any alarm bells. Hunt is MUCH larger now, both in terms of puzzle scope and participants, but that doesn’t mean that if the modern version becomes infeasible, it can’t go back to its roots. Maybe Mystery Hunt will always be online. Maybe it will get much smaller. Maybe the administration will give some latitude and let some people, or a lot of people, in for the weekend. We don’t really know, but my vote, regardless of how it affects me in the end, is that people let the MIT Puzzle Club determine what’s best for Hunt. It’s easy to snipe from the outside, especially if you’ve never constructed, but Mystery Hunt is and continues to be an MIT event, and the only reason it exists and receives MIT resources is the service it provides to the people already on campus. If you want to start your own non-MIT event, be my guest (lots of people have online over the last few years), and if you want to keep participating in Hunt if the opportunity arises, and that opportunity arises, by all means participate. But if you’re in a conversation with me, and you consider yourself to have ownership of the event because you’ve been an invited guest for so long, expect me to raise my voice.

Teammate is the third construction team in a row to have to captain a Hunt in these treacherous waters… I think the first two did an admirable job (even if online Mystery Hunts have turned out not to be my cup of tea) and I have a lot of faith in this year’s team to continue the streak.

2022 MIT Mystery Hunt, Part 1: The Big Picture and the Small Child

(This is a recap/review of the 2022 MIT Mystery Hunt, which happened this month. Puzzles and solutions can currently be found here. This recap may contain spoilers.)

My Mystery Hunt posts this year are going to be weird, because this was a weird Hunt… not that Palindrome’s Hunt was weird (apart from the sense that the past two virtual Hunts have been weird in the grand scheme of Mystery Hunts), but that my personal Hunt experience was weird. Forgive me if the first few paragraphs of this sound like they belong in a parenting blog… but becoming a parent is something many people go through and thus many puzzlehunt enthusiasts go through (plenty of my Setec teammates have kids of their own), and so maybe this will still be relatable to some of you.

My son Simon, who’s a little over ten months old now, started day care the Monday before Hunt. We are extremely fortunate in that we didn’t have to bring him in until now; my wife and I are both college faculty, and when our son was born, she had the remainder of the spring off from teaching, and then I was off in the fall. We are less fortunate in that just before day care started, he started experiencing separation anxiety for the first time… after 9+ months of barely being willing to be held, he now frequently grabs our hands and wants to be picked up. Predictably, he has not been fond of day care so far, and as a result he’s been cranky even at home and even less comfortable being away from us, even at home.

Our Mystery Hunt plan was to have a couple of squads of helpful relatives (Jackie’s mom, and my brother and his wife) babysit in shifts, with Simon staying in the house where all of the appropriate baby resources were, so that Jackie and I could do as much hunting as possible downstairs. As you can imagine from the context above, babysitting did not go as well as we hoped, and he generally screamed when we both left the room. On top of that he didn’t sleep particularly well. And while Jackie very nobly and generously took on more of the baby-wrangling than normal, she couldn’t do it all, so there was much less of us both solving than I would have liked, and also more time off from solving myself than I would have liked.

With all that in mind, while I undoubtedly still spent most of my weekend puzzle-solving, I was almost certainly more distracted than I’ve been at any other Mystery Hunt, and it all just felt… strange. For me, Hunt has always been a delicate balance of getting food fast and finding just enough sleep to be rested and maximize my puzzle contributions to my team, and when you start poking holes in both the sleep and solving, especially at times you can’t predict, both suffer. Not only did I feel less useful to my team than usual, but I also felt like a complete physical wreck by Sunday evening between the sleep deprivation and the hunching over a laptop in a cold basement. Optimistically this is peak difficulty, as next year I really hope we’re back at MIT, and even if we’re not, we’ll probably be in a better position to have Simon cared for at a different location for a few days. But given the confluence, I definitely don’t feel like I gave this Hunt the full attention it deserved, and my experience was unavoidably tainted as a result.

But other than that, Mrs. Lincoln, how was the play? Quite good, as far as I can tell. Much like Plane 1 Plane 2 Plane 3 Galactic Trendsetters Plane 4 Plane 6 Plane 5 (come on, you had one job) last year, I think Palindrome created and ran an excellent Hunt despite difficult constraints, though also like last year, I have my own nitpicks. In this post, I’ll run through some of the positives and negatives of the Hunt as a whole (keeping in mind, as always, that criticism shouldn’t be parsed as a lack of gratitude, but rather a conversation starter to help Hunt continue to evolve), and I plan to write a second post about individual puzzles that I found memorable.

As I do when I solicit feedback for teaching presentations, let’s start with the good stuff.

PROS

* The puzzles! I thought the puzzles I encountered tended to be of very high quality, in terms of their creativity, their cleanness, and their fairness. I confess to being one of many people who wondered if an NPL-heavy construction team would write a more “old school” Hunt dominated by crosswords and PDFable puzzles, in contrast to the robust multimedia/interactive content that has made recent Mystery Hunts and many recent online hunts vivid and exciting compared to the Mystery Hunts I started on. (The definition of a Hunt “puzzle” has changed intensely over the last two decades; most of the puzzles I wrote for 2000 would barely qualify as one step of a modern puzzle.) I had zero reason to worry, as puzzles came in a variety of formats and called on a variety of skills, and they felt like they came from a team that has been paying attention to the greater puzzlehunt landscape, not just Mystery Hunt. When I was able to sit down and solve things, I frequently had a blast doing so.

* I am a sucker for themed rounds (and have been ever since 2004 when ATTORNEY first wrote a Hunt where the presentation of the rounds really made them feel distinct from each other) and I adored the selection of round themes, the incorporation of those themes into puzzles, and the exquisite art design that made it easy to remember which round every puzzle lived in. I didn’t get quite as much of a rush from the Investigation or Ministry, but once the chunky part of the Hunt opened up and I got to see both Lake Eerie and Noirleans, I had that rush of wanting to solve enough to see what big reveal was coming next, and that’s what I want out of a puzzlehunt that lasts more than a day.

* Speaking of Pen Station, I really liked that the structure and scope of the rest of the Hunt was fairly clear when we got there. The MTA-styled line board made it clear how many rounds we should expect and helped to unite these disparate zones into a shared universe. When Plot Device opened, it was clear that this “round” was different, but not immediately clear how, and that was a nice balance of clarity and mystery.

* Also in the realm of balance, I thought the balance of event rewards was on the nose. It’s been tradition in many Hunts (including ones I’ve worked on) for the events to be a requirement for Hunt completion, but my calculus students can tell you that required things are not automatically fun, and especially in an online event, attendance may not be easy. I liked that this years events were not mandatory, but that they gave just enough free answer credit to feel rewarding without breaking the rest of the system.

* If you’re reading this, there’s a really good chance you’ve already watched the reward video for Recipeoria. If you haven’t, then either (a) solve Shopping List, the Recipeoria meta, since it’s a fun puzzle and the reward video partially spoils in, or (b) just watch the video. I didn’t watch it during Hunt (more on that later), but afterward, Erin Rhode told me to watch it immediately, and my jaw DROPPED. Hell of a get, Palindrome.

CONS

* I’m going to devote the most detailed criticism to the appropriately themed elephant in the room (and most frequently cited issue during our Slack post-mortem): The Ministry metapuzzle, and more specifically the fact that the majority of the Hunt was gated behind solving it. This left Setec in the fairly unpleasant position of unlocking no new puzzles, regardless of other progress, for hours on Friday night. I think we had just informed Palindrome that we weren’t having much fun (an hour or so into literally only having one puzzle open, because we’d solved everything else), when we figured out what we were missing sans hints and were finally able to open some exciting new stuff.

Now, let me be clear that I’m not objecting to the metapuzzle itself… I think it was a great puzzle! I personally think it was underclued, but only in that nothing in the puzzle clearly tells you you should be going back to reprocess the original 25 answers; since those had been used up, we spent most of our wheel-spinning time staring at the flavor text, mural, and five meta answers, which is not enough to do anything. I think a 25×5 grid would have done wonders to let us know what inputs we should be considering, particular in a meta intended to be solved by as many teams as possible. My objection is to the gating. Any puzzle in the Hunt might stump teams and create a bottleneck. If that bottleneck occurs at the end of the Hunt, it’s still not ideal, but at least teams have had the opportunity to see everything they might be interested in. A gigantic chunk of the puzzles, story, and art live in Pen Station, and a lot of teams never gained access to that area at all. I think this is a problem, and it’s one that’s arisen and has been solved before.

Who’s ready for Story Time with Dan Katz, Bay Bay? (That was a wrestling reference.) Back in 2009, when dinosaurs roamed the earth, The Evil Midnight Bombers What Bomb At Midnight similarly gated Outer Zyzzlvaria (more than half the Hunt, divided into rounds with weird-ass structures) behind the board game reconstruction puzzle that served as the capstone to Inner Zyzzlvaria. We did this for all the reasons Palindrome probably did… we wanted to focus teams on the task at hand until they completed it, the story called for teams not to discover Outer Zyzzlvaria until they had a reason to, et cetera. Except when running the Hunt, we realized Friday night that lots of teams were getting stuck there, and that, in particular, unless we gave casual teams enough hand-holding to get through a very daunting puzzle, they weren’t ever going to make it to Outer Zyzzlvaria. So we improvised. We tweaked the plot a little bit, our tech master pulled some levers, and boom. Solve enough puzzles, you can get to Outer Zyzzlvaria. The recent Setec hunts took this approach into account: each had a big midgame set piece with a capstone interaction and story reveal, but if you didn’t complete it for some reason, you could still see other stuff.

Playing devil’s advocate against myself, I know a common criticism of the 2019 Hunt was that for many casual teams, the puzzles constituted a molasses flood of their own; we explicitly decided that we would eventually release every puzzle to every team no matter what, and since our number of puzzles was arguably too high, some teams were unquestionably overwhelmed. So is it better to hide the hard stuff from teams until they earn it? Or to give them potential access to everything, but encourage them to explore the shallow end first? Personally, I’d rather give teams the choice. According to the wrap-up, 60 teams solved Fruit Around, which is a great stat for a midgame accomplishment, but with the gating, that means (N-60) didn’t even know there was a Pen Station until after the Hunt. And I imagine a disproportionate number of those fenced-in teams were casual student teams, which doesn’t sit well with me.

* On a closely related note, except for Where The Wild Things Are, the puzzles involving physical objects were buried waaaaay too deep. Since the wrap-up stats graph only showed the top nine teams (more team stats and submission logs, please, Palindrome) I have no idea how we placed compared to the pack; I’m sure we weren’t top ten, but I bet we were top twenty? And we didn’t unlock the second puzzle requiring a box until just before 6pm (using up our manuscrip because we didn’t think we could after 6). If we were anywhere near the front of the pack, that means the vast majority of teams paid shipping for objects they never got to use (including all the teams who never left the Ministry).

* Compared to recent Hunts, I thought there was a surprising lack of clarity about when Hunt would “end.” Unless I missed it, there was no advance “HQ will run until Time X, or when the coin is found, whichever is later” announcement, so given that last year’s virtual Hunt ran through Monday, I mentally prepared myself for this one to do the same. Then when the “coin is found” announcement came with a message that HQ would close at 6, it was a harsh blow; we were at least seven metapuzzles away, and it was clear that we were going to land farther from finishing than in any recent year. I pushed hard until 6 (and I believe we solved two more metas but got stuck with 10/13 on Heartford) and planned to solve casually into the night, but the fatigue I talked about earlier knocked me out and I had to jump ship. Then at wrap-up, I was really surprised to hear that some teams finished after 6. Wait, you could still finish after HQ closed? That’s a great feature, but it was not at all clear to us. A lot of our team dropped out at 6 if not before, and if we knew that Hunt was still “officially” going, I think we might have approached things very differently.

I do think what ended up happening, a combination of human interactions ending Sunday evening (provided the coin is found by then) but advancement still active until Monday morning, is a really nice compromise that could become an excellent tradition if it remains feasible once Hunt is on campus again; when Death and Mayhem kept Hunt rolling until Monday, I remember thinking that was very generous but should not become an expectation. Ultimately, the constructors should be able to decide when they want to close up and sleep. But they should be transparent about what can be done when, and as far in advance as possible.

* I thought last year’s Hunt was too long. I thought this year’s Hunt was too long. See last year’s posts for relevant arguments. But hey, nine teams finished and the winners found the coin at a fairly reasonable time, so maybe this is becoming an “if the music’s too loud, you’re too old” scenario. I do still think that as long as Hunt stays long, teams will feel like they need to be giant to compete (though UNICODE EQUIVALENCE coming in 4th with only 30 people might make them the new Evil Midnight), and if teams are giant, construction teams will think they need to write a huge Hunt to appease them, and this process can be extrapolated to pinpoint the heat death of the universe.

* Finally, this was a point of contention when I brought it up with Setec, but I thought that structurally the videos may have been too skippable. I consider myself a person that cares about the plot of Hunt, and I didn’t feel tuned into it (I didn’t realize Tock was a significant character until he started popping up in some flavortext). I didn’t skip videos because they were bad… the ones I watched were lovely… but there were a lot of them, and they didn’t feel like they needed to be watched, so I solved puzzles instead. I think both this Hunt and the Galactic Hunt could potentially have benefitted from a less is more approach to both story and interactions/videos. Since Hunt is above all a puzzle competition, I think it’s really helpful for a puzzlehunt to have a small number of story beats that are aggressively highlighted so people are tuned into them. If you want to have a reward of some sort for every metapuzzle, I think that’s fine, and many people will enjoy them, but I would really like to be pointed at the story elements that matter most. But I may be in a unique middle ground on this; most people I talk to either don’t care much about interactions at all or want as many of them as possible, so perhaps it’s weird that I just want a few that I can’t easily avoid.

As part of my job, I write a lot of feedback on people’s teaching, and what’s happened above often happens in those feedback e-mails; the criticism takes up way more space, just because there’s more to say about something that’s flawed and could be fixed then something that is already great. (This is why disassembled model kits have longer instruction manuals than, say, bottles of water, even though my son would need help with either of them.) So as I do in those e-mails, let me emphasize that just because there are more words devoted to the cons than the pros above does not mean that overall this was a bad event. I thought it was a really good Hunt that I didn’t get to participate in as much as I wanted for reasons, and that some casual teams maybe didn’t get to participate in as much as they wanted for other reasons. Sincere thanks to Palindrome for their talent and hard work, and I’m excited that next year’s Hunt will be written by teammate, the authors of the Puzzlvaria Puzzlehunt of the Year from 2021. Surely this is not a coincidence, so if you want to win next year’s Mystery Hunt, get to work on writing my favorite online puzzlehunt of 2022. Guess what the current leader is?

I could print all of this text again backwards to make this post a palindrome, but let’s just pretend I did and move on with our lives.

My Favorite Puzzlehunts of 2021

Disclaimer for all three of my year-end top ten lists: Descriptions of puzzles/hunts may include spoilers, especially as you approach the top of the list and I have more detailed gushing to do. Also, keep in mind that these are my personal favorites, which means they are shaped by (a) my personal content tastes, (b) the fact that I tend to like harder puzzles (which doesn’t mean easier puzzles are bad!!), and (c) the fact that I happened to solve or co-solve these particular puzzles, since hunts have swelled to the size that I can rarely pay attention to every puzzle my team solves.

A good blogger would add links to all the relevant puzzles and hunts, but I’ve already spent too much time writing these posts. Maybe I’ll add them later, and if not, I bet you can all use Google.

10. UMD Puzzlehunt

This is specifically the UMD Puzzlehunt from January; I believe they’ve had another since then that was on-site without the puzzles being archived online. I really hope they post the second event, because the first event was quite good, with a nice variety of polished meaty puzzles, a fun plot twist, and an excellent final metapuzzle.

9. Huntinality

Wah! It was exciting to see a fresh hunt by new contructors (early in the year before there started being multiple nontrivial events per month) with a fun theme. I might be a little biased here in that I too believe Waluigi has been unjustly snubbed by the Smash Brothers design team. As have Raz, Mike Jones, and Guybrush Threepwood.

8. Puzzle Boat 8

Recent Puzzle Boats have been different for me than the older ones due to COVID; Puzzle Boat was traditionally the yearly event where I convinced some Boston-area friends to schlep out to my house in the suburbs and solve in person, and we haven’t been able to do that in 2020 or 2021. In addition, the size of Puzzle Boat is becoming a bit less novel as online hunts become larger-scale, and the mostly-PDF format limits some of the interesting things that are being done with interactive puzzles. On the other hand, the overall structure of grouping puzzles into metas (which has been used in lots of other places since but was mostly innovated by PB) is still a really fun and daunting task, and it continues to be insane that Foggy Brume can create an event this large on his own, especially in a year when he’s also contributing to Mystery Hunt.

7. Puzzle Rojak

I honestly wasn’t expecting much from this event since I know almost nothing about the authors, but a mutual friend assured me that the constructors knew what they were doing. And that was very true! While a number of the puzzles we didn’t solve fell into my weakest puzzlehunt categories (I’m unlikely to touch a puzzle with a double-digit number of chessboards or one with languages in a non-Western alphabet with a ten foot pole), Go-Karts, Drop Tower, Sea Caves, and Shallow Waters were all delightful (one made my top ten list, and I was tempted to include others but wanted to control for recency bias), and I also really enjoyed how the final metapuzzles came together.

6. Puzzle Potluck

There have been a lot of great sets of interlocking puzzles this year, so as we approach the top of the list, you’ll find that in many cases, what makes the difference for me is a compelling theme with surprising story moments. I have minimal experience with Animal Crossing beyond memes, Smash Brothers appearances, and that heartbreaking story from years ago about the kid’s mother, but I definitely know Among Us, and I cannot say I was expecting to come across a brutally murdered Animal Crossing character. Mashing up the two social video games that gained the most traction during the pandemic almost feels like a no-brainer in retrospect, but I certainly didn’t see it coming, and I enjoyed the result. That plus a nice pairing gimmick made for a solid Potluck outing in 2021.

5. Silph Puzzle Hunt

While I don’t like to think of myself as a person who values style over substance, I have at times tried to delve through the Smogon Puzzle Hunt archives, and I generally haven’t made it very far because the puzzles are PDFs plunked into a folder system. I don’t know if people are getting tired of all of the online hunts that use forks of the gph-site hunt code, but I think it makes those hunts look polished and appealing, and it’s arguably spoiled me for anything more gritty. This year’s Silph Hunt was an attempt by the Smogon authors to port the event from a Pokemon message board to a mainstream audience, and I found the result very entertaining. Difficulties seemed nicely tuned for the first and second phases, metapuzzles were nice and thematic, RED SUS was aggravating in all the right ways, and I haven’t stopped hearing nice things about The Minimeta That Goes Wrong (though I only assisted in the final extraction, which was clever but bottlenecky). It is a testament to the quality of this hunt that I have almost zero (maybe less than zero) interest in Pokemon, and I still thought this was a great event.

4. QoDE

Shortly before QoDE, the constructors released a short teaser puzzle that revealed its Batman theme. Speaking as the author of the Learned League One-Day Special on Batman villains, I immediately decided the best way to carry out this theme would be one puzzle per ridiculous BatUniverse villain, and I was not disappointed.

One of the big changes to the online huntscape during the pandemic, besides just the obvious increase in content, has been the increase in content difficulty. For many years, there was a significant gap between the average puzzle difficulty in, say, the Mystery Hunt versus most online events, until Galactic Puzzle Hunt (particularly starting in its second year) really upped the ante for how Mystery Hunty an online event could be. This year I didn’t expect to get those sorts of puzzles in the window between Galactic and teammate, and QoDE dropped in unexpectedly with some awesomely challenging content. I had some issues with the presentation of the final metapuzzle, but all the other metas were great (including my favorite of the year, and a beautifully illustrated first-round capstone) and the metaswere preceded by consistently interesting puzzles with fun flavor from a universe I personally enjoyed. Na na na na na na na na good hunt.

• 3. MIT Mystery Hunt
• 2. Galactic Puzzle Hunt

(Please excuse the weird double bullet points above… WordPress was NOT psyched about two consecutive numbered items in reverse order for some reason.)

Look some planes Galactic Trendsetters what a coincidence some more planes should be extremely proud of themselves for delivering two high-quality large-scale events in a single year (I know when I saw they were doing a GPH this year, first my head exploded, and then I got really excited that it was during the summer when Jackie and I have more flexible schedules). However, I’m probably surprising some readers, and maybe some members of Galactic, by putting GPH above Mystery Hunt. This doesn’t mean GPH was an objectively better event; the Mystery Hunt was a phenomenal accomplishment this year, especially under unprecedented difficult circumstances.

But as I said in my Hunt recap, the remote nature made me feel more disconnected than usual from my team, and I didn’t get familiar enough with the avatar space quickly enough to feel like I could process that element of the Hunt. GPH, on the other hand, had more digestible length, and my team was small enough that I felt I could interact with them fully and take in most of the hunt. Does this mean Mystery Hunt has passed me by? I hope not, and I suspect I would have enjoyed the Mystery Hunt even more had I been in a room with the rest of the Setec crew. But at least in the form I experienced the two events, I got a more satisfying experience out of GPH. Yet I got an even MORE satisfying experience out of…

1. Matt & Emma’s Carnival Conundrum

Not to drive a wedge between sibling teams Galactic and teammate, but in 2021, I think teammatehunt was the highlight of my puzzlehunting year. The puzzles were high-quality and challenging (just as they were in the 2020 version) and while the pairs of puzzles that interacted with each other were sometimes frustrating, the connections that revealed themselves were often pleasing.

But I think the thing that propelled this event to #1 on my list was one of the most impressive story beats I’ve seen in any puzzlehunt I can remember. After last year’s Matt & Emma hunt, I commented that I was surprised at how wholesome the theme was and continued to be; I’m so used to puzzlehunt themes being darkly comic that I was certain the birthday party would turn into Five Nights At Freddy’s, rather than just a jaunt through a delightful fantasy realm. One of the effects of that wholesomeness is that I think of Matt & Emma as two sweet kids, and even though I don’t know much about them, I know they care about each other. So when the magician in this year’s hunt told each of them bluntly that not only was their sibling gone, but that they never existed… and suddenly the previously bright side design turned dark… I remember gasping audibly.

That said, the rest of the story, as told through silent images, didn’t work for me nearly as well. I got the general beats but didn’t feel story happening with every puzzle, and I didn’t need to. But that one big shift was so simply and artfully executed, that it gave me a real emotional investment in helping the main characters. That took what was already a great set of tough, well-written puzzles and elevated it to something more meaningful, at least for me. That was enough to bring it to the top of the medal stand.

Thank you to everybody who solved puzzlehunts and especially wrote puzzlehunts this year. 2021 was a roller coaster of a year, as COVID still ravaged the earth, which is awful in almost all ways, but it did facilitate more puzzlehunting opportunities. Meanwhile, I had a child, which is great in lots of ways, but it did facilitate less puzzlehunting time for me personally. I’m grateful to my wife for doing more than her fair share of Simon-wrangling on certain weekends to allow me to squeeze in as much solving as I could, and I’m grateful to this community for continuing to do cool things and support each other’s projects. Sorry for the sparse posting this year, but I hope to keep encountering many of you online and (fingers crossed) in person in the new year.

My Favorite Metapuzzles of 2021

Disclaimer for all three of my year-end top ten lists: Descriptions of puzzles/hunts may include spoilers, especially as you approach the top of the list and I have more detailed gushing to do. Also, keep in mind that these are my personal favorites, which means they are shaped by (a) my personal content tastes, (b) the fact that I tend to like harder puzzles (which doesn’t mean easier puzzles are bad!!), and (c) the fact that I happened to solve or co-solve these particular puzzles, since hunts have swelled to the size that I can rarely pay attention to every puzzle my team solves.

A good blogger would add links to all the relevant puzzles and hunts, but I’ve already spent too much time writing these posts. Maybe I’ll add them later, and if not, I bet you can all use Google.

10. Candela’s Mystic Correspondence (Silph Hunt, Level 51 and lydian)

The mechanic here is slick and well-executed, even if it doesn’t really break much new ground. What I really liked from a thematic perspective was using the text of “Fire and Ice” as the data set for the collision between the Pokemon fire and ice factions. This is a good time to point out that even though this was one of my top ten metas of the year, it’s still not even my top “fire and ice” meta of the year.

9. Infinite Corridor (Mystery Hunt, Jon Schneider)

I have to be honest: Even though this is ostensibly a list of my favorite metas, Infinite Corridor makes the list more for slack-jawed awe than for my remembering it as an enjoyable solving experience. (To quote one of my favorite test-solve comments I ever got one of my puzzles: “Do I like it? I don’t know, but I respect it.” Complicated metas are often not my jam, but I liked that this meta facilitated five puzzle families that each had both interesting mechanics and submetas to solve in their own right. Plus, we had a teammate who coded the Python script (the Infinite Corridor Simulator Simulator?) to pull the answer once we had enough submeta answers, and that teammate happened to not be around when we had enough data to finish, which meant I got to run the code and be super-excited when it actually spit out a thematic answer instead of just another word.

8. Wah Street Bets (Huntinality, Benji Nguyen, Dan Simon, and Ryan Liu)

While the finale to Huntinality was a little reminiscent of the cookie clicker in the second Galactic Puzzle Hunt, I liked this more because (a) there was a greater variety of stuff to do, and (b) I didn’t have to tensely watch my teammates click buttons just the right number of times so that we didn’t screw up and have to start over. Then the very last step did a great job drawing our attention to something we didn’t notice when solving the subpuzzles for a nice punchline. This felt more like a gentle victory lap than a difficult capstone to the event, but it served its purpose nicely.

7. Magical/Entanglement (Matt & Emma’s Carnival Conundrum, Herman Chau, Rachel Wei, and Patrick Xia)

The one puzzle my team never solved in the first round was Radio Noise, which contributes the most to the answers to this pair of metas… fortunately, the output is the most thematic, so it’s easiest to guess (though not as easy to use it to backsolve the puzzle) so we were able to intuit the right beginnings, and that arguably made the solve even more satisfying. This was also a nice teaser for the main gimmick of the rest of the hunt, without giving it away outright.

6. Final Metapuzzle (Edric’s Truzzle Hunt; Edric Haleen)

This was the first of Edric’s events that I fully participated in, so I didn’t know what to expect. It was reminiscent of the old Australian hunts in a lot of ways (including the visual presentation). But those hunts frequently had underconstrained answers and inelegant and/or inscrutable metapuzzles. The Truzzle Hunt meta, on the other hand, required the answers to be ridiculously constrained in a way I did not notice at all until I had to, and with those constraints in place, very little shell was needed to produce an answer. It was very pleasing to discover that so much structure had been hidden the whole time in plain sight.

5. Student Center (Mystery Hunt; Jon Schneider, Rahul Sridhar, Anderson Wang)

The Student Center meta is another metapuzzle I already talked about in my Mystery Hunt writeup. It’s high on the list due to its use of one big aha, a bunch of fun little mini-ahas, and a lot of logic-poking to work out what goes where once you know what’s happening. It’s also very self-confirming for a large metapuzzle; once you think you know how to interpret a club, the data does a good job letting you know if you’re right or not. You could also argue that the Student Center meta incidentally includes its submetas, which means it includes both Random Hall and MacGregor Hall, which were both awesome. In extremely different ways.

4. Telescope (Galactic Puzzle Hunt; Jon Schneider with contributions from Brian Chen, Colin Lu, Rahul Sridhar, and Anderson Wang)

It’s been a while, so I might be misremembering, but early on when we had 36 dots and the pictures of eyes, I quickly wanted this to tie into the eye chart meta from the beginning of the hunt (which had seemed awkwardly underconstrained at the time). The payoff here is really nice; I love the epicness of revealing that the pyramid-shaped thing you were looking at in Round 1 is actually a nice orderly square if you just view it from the right angle. My team almost certainly didn’t build the square as efficiently as we could have, but it was satisfying to do so, and even though my teammates worked out the right interpretation shortly after I fell asleep, I was still pleased to see how it worked in the end.

3. Bulletin Board (Puzzle Potluck; Curtis Liu, Darren Yin, and Rajeev Nayak)

You’re just cruising along, thinking, “Okay, I guess these visually hinted transformations are kind of cute, and I can apply them to each of the answers, but I’m not sure what to do now…” And then you realize that these psychopaths (no offense) managed to find answers that could each take on two transformations each. I’m giving this a very high ranking for the wow factor of that mechanic, the appropriateness of this meta for a hunt in which you had to pair up puzzles and chain together locations, and a solid thematic punchline to tie it all up.

2. Hall of Mirrors (Matt & Emma’s Carnival Conundrum; Jacqui Fashimpaur, Liam Thomas, and Samuel Yeom)

Metapuzzles often fall into one of two categories: the type where there’s one big aha, and once you figure out the gimmick, you can apply it and instantly grab an answer (nothing wrong with that); and the type where the answers are inputs to a multi-step puzzle that requires a lot of effort, to the point where the metapuzzle is itself just another puzzle, but one you have to solve (maybe something wrong with that, but that’s in the eye of the beholder). Hall of Mirrors landed in a nice space between these, where there were multiple ahas and some gruntwork to do, but the puzzle presentation guided us through it fairly and the properties of the answers were important enough to make this not feel like a standalone puzzle. When I really wanted a mirror to change the vowels of words, and a teammate pointed out how that was embedded in the design, I was floored. Solving this puzzle was a series of “Wouldn’t it be cool if it worked like this?” moments, where it almost always worked just like that.

1. Firefly & Mr Freeze (QoDE; Scott Handelman and Jonah Ostroff, featuring Adam Maresca)

I did promise you another temperature extremes meta, and this one’s all the way at the top of the list. This had an elegant presentation, it was fun to solve, and it relies on a really clever data set (big spoilers coming here): groups that not only have two clear members associated with hot and cold, but that also have a member with one H and no C’s, and another with vice versa. When solving this meta, we spotted the first of these properties first, and I was delighted when the second popped out of the answers we had to work with. If you’re an elegance nitpicker, you could argue that the plus/minus numbers in the shell make the construction slightly easier, but there’s so much neat stuff going on in an organized fashion here that I have no objections to shifting the positions to make the whole thing work.

This meta also gets bonus points because it admitted an answer that fit a Batman puzzle (in a Batman Hunt) so thematically that I was 100% sure the meta was written around forcing that phrase into the answer pool. I was subsequently informed that this was not the case, and it has caused me to reevaluate my own views on creationism ever since.

My Favorite Feeder (Nonmeta) Puzzles of 2021

Disclaimer for all three of my year-end top ten lists: Descriptions of puzzles/hunts may include spoilers, especially as you approach the top of the list and I have more detailed gushing to do. Also, keep in mind that these are my personal favorites, which means they are shaped by (a) my personal content tastes, (b) the fact that I tend to like harder puzzles (which doesn’t mean easier puzzles are bad!!), and (c) the fact that I happened to solve or co-solve these particular puzzles, since hunts have swelled to the size that I can rarely pay attention to every puzzle my team solves.

A good blogger would add links to all the relevant puzzles and hunts, but I’ve already spent too much time writing these posts. Maybe I’ll add them later, and if not, I bet you can all use Google.

10. Intersections (Galactic Puzzle Hunt; Brian Chen)

While I’ve sung for a whole bunch of hunt puzzles I’ve written, it’s pretty rare that I solve one by singing. But that turned out to be the best strategy for Intersections; once I knew what I was trying to do, I found myself singing duets with the puzzle author to find the unison notes. This could have been an unpleasant experience, except the alt-melodies were both written and performed excellently, so I thought we sounded quite pretty together. (I just checked with my wife, who was sitting next to me at the time, to see if it actually sounded good. She does not remember.)

9. Cafe Five (MIT Mystery Hunt; Nathan Pinsker and Josh Alman)

I won’t say too much here about Mystery Hunt puzzles since I already posted about my favorites in January, but if you want the opportunity to show off by speed-solving all sorts of different puzzles, Cafe Five is for you! (Barbie’s Murder Party at the House on the Hill is also for you, but that was from last year.) It’s also the sort of puzzle that requires so much of the team to pay attention to it that it’s instantly part of the shared Mystery Hunt consciousness. I expect “L. Rafael Reif approaches the cafe” to be slang for “this puzzle suddenly got much harder” for years to come.

8. The Meta Meta Meta… Puzzle (Galactic Puzzle Hunt; Yannick Yao)

The Author’s Notes for this puzzle mention the Infinite Corridor meta from the MIT Mystery Hunt, and this felt very much like a more approachable version of that epic puzzle recursion. There are a lot of levels of discoveries to make, and it was fun to chip away at it with my teammates, particularly while it was still in “blurry” mode.

7. Go-Karts (Puzzle Rojak; aki and Jonathan)

I admittedly didn’t carry out some of the more tedious word-searchy steps of this puzzle (props to my teammates Scott and Jenn for doing most of the grunt work), so I might be biased via only seeing the flashy parts. But there was something very satisfying about seeing all the messages fall out of the puzzle, and then to hear characters pop up in the sound clips at just the right time (once we knew to expect them), and on top of all that an unexpectedly jaw-dropping extraction step. It also helps that I like Mario Kart. I’m-a-gonna win!

6. Fun With Sudoku (MIT Mystery Hunt; Josh Alman and Mitchell Lee)

This is another one I posted about after Mystery Hunt. Logic puzzles plus combinatorics equals A-OK in my book, and this resulted in a particularly satisfying team solve.

5. All That’s Left To Do is Extract (Matt & Emma’s Carnival Conundrum; Ivan Wang, Alex Irpan, and Rachel Wei)

It’s always fun when a puzzle takes you to things that were hidden in plain sight in some corner of the hunt website (well, I should say it’s fun as long as you’re guided there in a fair way). ATLTDIE takes you on a pretty exhaustive tour of the teammate website and does some fairly bonkers things along the way, from hiding data in what you thought was your own image file, to embedding Braille in the otherwise innocent looking favicon (a term I learned from this puzzle), to making the website respond to the Konami code.

4. Make Your Own Math Quiz (Galactic Puzzle Hunt; Jakob Weisblat and Josh Alman)

It wouldn’t be a GPH without a Make Your Own Something Or Other, and having made quite a few math quizzes in my time, this felt like home. The concept of students complaining that questions aren’t similar enough is actually common in some situations, and so the interpretation here is pretty funny. And from a puzzling perspective, figuring out how to make the constraints all work at the same time was satisfying, with a great extraction punchline right at the end.

3. Qatalog (QoDE; Jonah Ostroff)

Chatting (and solving!) with puzzlehunters on many teams this year has brought at least one thing sharply into focus… I use OneLook more than a lot of people, and I use Nutrimatic less than a lot of people. I had also encountered Qat, but I hadn’t made much use of it because the syntax is fairly intmidating. Qatalog simultaneously served as a really effective Qat tutorial, and an ingenious array of puzzly ways to utilize said unusual syntax, with the most thematically appropriate final step that you could ask for. (It’s nice when your puzzle about Website X ends with Website X spitting out your answer verbatim.) “Minimeta” puzzles are in vogue right now and are almost always appreciated (The Minimeta That Goes Wrong was likely a favorite of many people, but I just wasn’t around for most of the solve), but I loved this one in particular for both its creative variations on a theme and for its ability to actually teach me something that I’ve used since.

2. Meta-Eval Times/Pin the Tail (Matt & Emma’s Carnival Conundrum; Andrew He, Katie Dunn, Catherine Wu, and Patrick Xia)

I’m considering this entangled pair of puzzles from this year’s teammatehunt as a single entry. Puzzles with lots of parallelizable mini-ahas are always fun, and the first stage of Meta-Eval Times delivered in that respect; meanwhile, after I spent several years of my life mapping Pin The Tail, I was thrilled that one of my teammates recognized it as an Among Us board. But by far the highlight of the puzzle, after spending way too much time trying to satisfy all the advanced Meta-Eval constraints at once, was remembering how actions in Remember to Hydrate had affected Mystery Manor and having the realization, “When we kill crewmates in Pin The Tail, does anyone die in Meta-Eval?” Not only is this a wonderfully unexpected effect, there’s also something delightful about the idea of writing a metapuzzle for picky test-solvers and assassinating half of them to get your puzzle into post-production. No offense to anyone that I’ve written Hunt with. You’re all wonderful, and it’s been long enough that I do not want to murder any of you.

1. Divide and Conquer (Galactic Puzzle Hunt; Lewis Chen and Anderson Wang)

It was difficult to rank most of the items in my three “best of” lists, except for this puzzle, which frankly never budged from the #1 slot for non-meta-puzzles since I finished solving it. Back in the early days of Setec Mystery Hunt writing, there was a house rule that a puzzle shouldn’t have more than one aha. Puzzles were much much much simpler then, and good puzzles these days often have multiple steps; the true art to this is giving the solver enough help to progress but enough resistance to feel challenged. And boy, was this challenging… I think I devoted a full afternoon to it, with Jackie’s help.

Having quarter-specific approval/rejection for the example puzzles was the perfect amount of assistance to help me identify regional mechanics. And I suspect I would have found the puzzle overwhelming if we’d had to solve every bit of it, but getting two of the mini-puzzles and recognizing the four-letter pictures was enough to crack the letter-to-variant relationship and solve just enough of the other feeders to complete the correspondence grid. Then we were able to crack enough of the big grid to get the clue phrase without the whole picture; thankfully after calling in the one example I knew of that could answer the clue phrase, someone on my team knew another correct one. I’m the opposite of a Star Wars fan, but I’ll still mine it for a metaphor; this puzzle is like an imposing Death Star that seems impossible to overcome from a distance, but there are enough (intentional!) cracks in the armor, that with enough skill I was able to fly in, drop a bomb, and break the thing wide open. I haven’t had a more satisfying solve this year, and probably not in the last few years either.