(This is a recap/review of the 2017 Galactic Puzzle Hunt, which happened earlier this month. Puzzles and solutions can be found here. This recap may contain spoilers, but I’ll try to avoid them where possible, especially for puzzles I recommend.)
Back when I started solving Australian puzzlehunts, I usually played with The Sons of Tamarkin; we were perennial contenders against plugh, The Elite, and [pi] (aka Galactic Trendsetters, the team that generously brought us this hunt). A few years ago, I decided to join Killer Chicken Bones for SUMS, figuring that would be my annual break from my usual team… but in the meantime, some of the Sons of Tamarkin have gotten too busy to solve and others have found better teammates (for this hunt they totally outclassed us as Brown Herrings). So for now I’m with KCB for the immediate future, until we hit a hunt where enough original members are available to hit the team size cap and I get bumped.
The GPH is the second American “Australian rules” puzzlehunt I’m aware of, and the first intended for the public. (Last year some friends at Brown ran an event called CRUMS, which was aimed at the Brown community and doesn’t appear to be online anymore.) Maybe the American “flavor” made it more appealing, but it was definitely my favorite Aussie-style hunt I’ve participated in. Puzzles were generally solid and polished, and there was a very elegant meta structure at the end (and the meta was actually worth points… what a concept!).
On Day 1 and 2, we solved all of the puzzles before I went to sleep; on Day 3 we knew what we were doing on our last puzzle when I turned in, and everything was solved when I woke up. Things took a turn on Day 4; we only solved two puzzles before I went to sleep, and we still only had two the next morning. We hammered away at the rest and solved two of them before I had to leave for a wedding, but we went 24 hours without solving A Basic Puzzle, knocking ourselves out of contention for first place. We failed to solve one additional puzzle on Day 5 before overtime, but we were able to solve both puzzles fairly quickly with hints, so we finished first among the non-perfect teams (fifth overall).
Speaking of hints, being on this end of the yes/no system has changed my opinion of the oracular hint system, which we used in the Mystery Hunt, and which very few teams took advantage of. Galactic Puzzlesetters (sorry guys, I’m too lazy to figure out how to insert the airplanes, but rest assured I made the noises while typing) have posted a great wrap-up dissecting their own hunt, and they have some interesting thoughts on yes/no hints, but here’s mine: One yes/no hint is rarely useful, but a series of them (where you can ask follow-up questions) is potentially very useful. On both of the puzzles we solved in overtime, we figured out what we needed to do not by asking one question, but by asking several and narrowing down what we shouldn’t be thinking about. So in retrospect, I think we should probably have given out batches of questions in the Mystery Hunt, rather than singletons. (Though I again want to emphasize that earned hints were only intended to sort things out as needed for contender teams… we always intended to be generous to more casual teams, and that was instituted here with the “infinite hint” system, which is much tougher to execute in an online hunt!)
= = = = =
Puzzles I especially liked:
* Zero Space (Day 1)
Took a long time to get the last aha here (exactly how to interpret the first phrase you extract) but once we did, it was hard to believe we didn’t see it for so long. Very clever and elegant solve path.
* Very Fun Logic Puzzle (Day 2)
As advertised.
* How to Best Write an Essay (Day 2)
Completely missed the hidden message in this puzzle. We guessed the interpretation of W quickly (though we were surprised there was no accompanying S), and then recognizing what Y should be without the message was an extremely satisfying aha. I’m almost a little sad that we weren’t supposed to figure it out the way we did, though I get that someone unfamiliar with the artist would have no way in in that case.
* Drive (Day 5)
This mostly got solved while I was at Jenn’s wedding. When I got back, I found a mostly solved puzzle on the spreadsheet, at which point I said to myself, “What a cool data set. And what a cool way to put it into a grid! And it would also be neat if you extracted like this… Yup. Solved.”
* Everything on Day 6 (Day 6)
There was no advertised meta in this set, but given the weirdness of the Day 1 answers and the sixth day that usually doesn’t exist, I predicted the five puzzles on Day 6 would be five five-puzzle metas. I suppose it makes much more sense in a March Madness hunt to have a 2^n-puzzle final meta, but you can’t blame for failing to predict that 25 puzzles would lead to a 16-answer bracket!
For what it’s worth, I do think the “put these answers in a line and modify them step by step,” as seen in Duck Quonundrum, may have jumped the shark. I’ve seen it used in a few places, and I don’t think it’s ever been as cool as it was in MH2015. But given that these answers each had to work in two metas, I get that this is a flexible mechanism to achieve that goal.
The main mechanism for the final meta was something we immediately considered when we saw it… and dismissed, even after solving two conferences. The rest of the team put two and two together while Jackie and I were asleep, but looking over it after it was already solved, I thought everything was fair and satisfying.
Puzzles I especially disliked:
* Famous by Association (Day 4)
I was delighted by the puzzle quality for most of this hunt, but I thought there was a downturn in quality on Day 4. Famous by Association was probably the clunkiest puzzle for me; the matches weren’t clean enough for us to be confident about how the mechanic was intended to work, and even after we sort of knew what we were doing, we frequently had multiple options for some of the items, and we eventually chose things based on giving us good letters. It’s never good when solving individual pieces of a puzzle leads less to an “Aha!” and more to a “Well, maybe? I guess?”
* The Treasure of Apollo (Day 4)
Most of the issues I had with this puzzle are acknowledged in the posted solution. The overall gimmick of the puzzle is neat (though once again the matching wasn’t tremendously clean… we had our third and ninth characters switched for a long time), but there was a lot of extraneous data, and the way you’re expected to parse the data was inconsistent. I also see why they added the enumeration after the fact… given all of that ambiguity, that’s an incredibly indirect phrase to parse without spaces.
* A Basic Puzzle (Day 4)
This was one of our late solves. We immediately figured out what to do with the first line, but we tried a lot of similar approaches to the other lines with no success. Personally, I think the solution space was just too open here… you could do a lot of almost-right things without confirmation if you didn’t know what you were looking for, whereas with some narrowing via hinting, the rest of the team was able to polish this off while I was at the wedding.
* Unaligned (Day 5)
This is the other puzzle where we couldn’t figure out what to do without narrowing the solution space (though we should have, since Kristy suggested doing the right thing and I advised against pursuing it). My complaints about this are that (a) having two identical grids really makes you want to combine them in some way, and (b) confirmation here really relies on seeing that a bunch of three-letter strings are all words, which is something that could easily happen by accident. Our hint requests helped us determine which parts of the completed grids we could completely ignore, which helped a lot.
This was another puzzle where the posted solution revealed a hidden message we never found (two of them, in fact!).
Puzzles I simultaneously reluctantly loved and lovingly despised:
* Scramble for the Stars (Day 3)
This will be a bit more spoilery than the responses above, because I want to get into details. Somebody counted the number of clues and suggested constellations, and after trying a few things, I suggested the puzzle might work exactly how it turned out it did work, and said I really hoped it didn’t, because that sounded like it would be a pain in the ass. Then it occurred to me that this would only work if there was an even number of letters… I added them up and got an odd number, and I breathed a sigh of relief.
We had STRAIGHTLINE instead of STRAIGHTLINES.
Once we made that correction, I feared again that the puzzle would work the way I thought it did. I remembered from my ceiling growing up that Draco was the largest constellation (my incredibly awesome parents not only put a bunch of glow in the dark stars on my ceiling one day while I was at school, they took the time to make an accurate star map with constellations), and Draco didn’t have the right number of neighbors. But then I looked up the actual largest constellation and smallest constellation, and looked up their numbers of neighbors. They matched. Crap.
So, finally believing my original idea was right, four of us spent a looong time scraping the adjacency data from Wikipedia by hand (I considered doing it on a printed-out star map, but it was way too small to see anything). Once we actually got that data down and figured out how to keep track of assignments, the logic puzzle portion was exquisitely elegant. As the posted solution suggested, we started with rare letters and the necessary degrees of their neighboring nodes, and then things got easier and easier. This would have been a nightmare if we weren’t sure of most of the answers, but the clues were clear enough that we had all but two right. One of them, LOAD, was our fault… but having a clue for NOSTALGIA that could just as easily clue NOSTALGIC as just mean.
I went to sleep while Brent and Kristy were still assigning letters, and my only idea for extraction was to alphabetize the constellations and read all the letters in that order… but constraining all 88 letters seemed like an impossible construction feat. Putting the answer phrase on the zodiac was genius.
So as it turns out, this was a magnificent puzzle, and I have tremendous respect for the constructors. But at the same time, I stand by my initial prediction… solving it was a pain in the ass. 🙂
* X-Ray Fish (Day 3)
My only hate for this puzzle came from the sound clip being really annoying after 50 times, and from it being very difficult to count stuff in the video (even pausing it). I did like the overall mechanism, and knowing which part of the song to focus on due to playing it too many times in Rock Band. (Me: “I think those sounds are in the actual song, aren’t they?” Someone I will not name: “No, definitely not.”) And if you haven’t already, be sure to check out the posted solution for this puzzle.
= = = = =
As it happens, part of the reason the Galactic Puzzle Hunt happened (according to the wrap-up) is that the constructors had some downtime after finishing the Mystery Hunt early… So as a member of Setec, I take partial credit for miscalibrating the difficulty of our Hunt! In any case, this was a great addition to the online huntscape. Given that it was free, the constructors have no obligation to give us another one next year. But I hope they will, and I hope they start writing now just in case they don’t have quite as much free time next MLK weekend.
(Note to self: If I’m going to keep writing these recaps regularly, I’m going to have to start making them shorter.)
