(no subject)
Jan. 24th, 2019 09:50 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
seekingferretMystery Hunt was amazing, but it took a lot out of me. I took Monday and Tuesday off work as recovery days. The combination of low sleep and inconsistent eating (I ate a reasonable number of real meals, but not always at normal times) and adrenaline and a six hour drive home on icy roads post snowstorm and probably a bit of a cold beat the hell out of me, and I barely left my bed the past two days, and never left the house. Back at work Wednesday, still a little off my game.
The theme was a combination of the Nightmare Before Christmas's holiday metaphysics and the Great Molasses Flood Centennial. Something was wrong in the Holiday Forest involving molasses and puzzles and it was causing unfortunate conflicts between the different holidays. This was a less theatrically themed Hunt than last year's Inside Out hunt, which had lots of interactions with costumed characters from GC. But it was deeply, deeply themed in terms of the puzzles, which for me did a lot of the work of conveying the story. Answers often connected to puzzles, and metapuzzles generally connected well to their themes (and were often funny).
Also there was a Bloomsday Round and it was great. Although it means that if Palindrome ever wins, I won't be able to pitch the Bloomsday Round I always fantasized about, which would have looked similar to SETEC's in some ways, but would have had some significant differences.
Because of the notion that conflict between holidays was causing the problems, the Hunt's metapuzzles had a novel mechanism where a solver didn't know which meta a puzzle's answer belong to- there were several metapuzzles connected to each round. A Christmas-Halloween meta and a Halloween-Thanksgiving meta, for exmaple, both used some of the answers from the Halloween round. The result of this was very difficult metapuzzles that often seemed to have been designed to be difficult to solve without a high majority of the answers, if not all the answers. Backsolving, also, was harder than usual because the universe of possible puzzles an answer in a solved meta could belong to was higher, and spamming twenty puzzles with a couple of potential backsolve answers would be frowned upon. (As the Hunt dragged on, we became less reluctant.)
These mechanisms seem to have stemmed from a combination of just a general desire to show off a novel metapuzzle structure, and a desire to not repeat the result of SETEC's last hunt, where a number of teams finished the Hunt in less than 24 hours, partially because the metas were straightforward and backsolve-friendly. It may have been a slight overcorrection, but the Hunt was fun and none of the metas felt unfair. I'm curious how the Hunt seemed to less sophisticated teams than ours... sometimes the really long, challenging Hunts are less fun for weaker teams, but my guess is that the puzzles were clean and satisfying enough in general that even weaker teams could get enough satisfaction.
For most of the Hunt, Palindrome rocked these puzzles. There were dual puzzle release mechanisms- teams that were solving fast enough got one puzzle released every time they solved a puzzle, and teams that were below a preset time curve got extra puzzles released on a schedule so that all puzzles were released by noon on Sunday. Palindrome was one of only two teams that stayed ahead of the time curve for the whole Hunt. And, to judge from the graph of puzzles solved released at wrapup, we were neck and neck with the other top team for the first twenty four hours before pulling ahead Saturday night.
But then we got logjammed on the last handful of puzzles. It was a combination of some tough late appearing puzzles and some puzzles that we'd just gotten stuck on for 24 hours or more and turned out to be meta-critical. We spent a good portion of Sunday stuck on about ten puzzles unsolved and after starting Sunday with four or five metas left, knocked away the last few metas about one every three or four hours, agonizingly slowly.
We solved our penultimate meta at about 5:15 on Sunday afternoon and a few minutes later, someone from SETEC came to visit us and warn us that even if we solved our final meta quickly, we had been beaten to it by another team which was at a sufficiently advanced place in the final runaround that they could not be caught. Fifteen minutes later, the coin was found; We solved the final meta about fifteen minutes after that.
It was frustrating to come so close to winning and fall short, but hunting with Palindrome was incredible fun throughout and I'm looking forward to next year.
Standout puzzles I worked on include Loaded, a pattern recognition puzzle involving polyhedral dice that was the one puzzle all Hunt where I had all the key insights to solve.... American Icons, a puzzle that started with us cutting out and dressing paper dolls of American feminist icons, before I had the insight that the dolls were analogs of American Girl dolls (thanks to my sister for this osmosed knowledge)... The Turducken Konundrum, an amazing and hilarious set of nested Duck Konundra... Clued Connections, an Only Connect extravaganza... and Bloom Filter, a lovely combination of a programming puzzle, and a word connections puzzle.
I think the personal highlight of the hunt for me was figuring out the final answer to American Icons on the walk from our secondary room to our primary room for the team meeting where we had been about to report that we were stuck and needed help figuring out the final answer to the puzzle. I immediately turned from a walk to a sprint to get into the room and ask someone to call in the answer, and when our team captain asked us how we were doing on the puzzle I was able to say "In about a minute they're going to confirm we got the correct answer."
The complete list of puzzles is here: http://web.mit.edu/puzzle/www/2019/puzzle.html
The theme was a combination of the Nightmare Before Christmas's holiday metaphysics and the Great Molasses Flood Centennial. Something was wrong in the Holiday Forest involving molasses and puzzles and it was causing unfortunate conflicts between the different holidays. This was a less theatrically themed Hunt than last year's Inside Out hunt, which had lots of interactions with costumed characters from GC. But it was deeply, deeply themed in terms of the puzzles, which for me did a lot of the work of conveying the story. Answers often connected to puzzles, and metapuzzles generally connected well to their themes (and were often funny).
Also there was a Bloomsday Round and it was great. Although it means that if Palindrome ever wins, I won't be able to pitch the Bloomsday Round I always fantasized about, which would have looked similar to SETEC's in some ways, but would have had some significant differences.
Because of the notion that conflict between holidays was causing the problems, the Hunt's metapuzzles had a novel mechanism where a solver didn't know which meta a puzzle's answer belong to- there were several metapuzzles connected to each round. A Christmas-Halloween meta and a Halloween-Thanksgiving meta, for exmaple, both used some of the answers from the Halloween round. The result of this was very difficult metapuzzles that often seemed to have been designed to be difficult to solve without a high majority of the answers, if not all the answers. Backsolving, also, was harder than usual because the universe of possible puzzles an answer in a solved meta could belong to was higher, and spamming twenty puzzles with a couple of potential backsolve answers would be frowned upon. (As the Hunt dragged on, we became less reluctant.)
These mechanisms seem to have stemmed from a combination of just a general desire to show off a novel metapuzzle structure, and a desire to not repeat the result of SETEC's last hunt, where a number of teams finished the Hunt in less than 24 hours, partially because the metas were straightforward and backsolve-friendly. It may have been a slight overcorrection, but the Hunt was fun and none of the metas felt unfair. I'm curious how the Hunt seemed to less sophisticated teams than ours... sometimes the really long, challenging Hunts are less fun for weaker teams, but my guess is that the puzzles were clean and satisfying enough in general that even weaker teams could get enough satisfaction.
For most of the Hunt, Palindrome rocked these puzzles. There were dual puzzle release mechanisms- teams that were solving fast enough got one puzzle released every time they solved a puzzle, and teams that were below a preset time curve got extra puzzles released on a schedule so that all puzzles were released by noon on Sunday. Palindrome was one of only two teams that stayed ahead of the time curve for the whole Hunt. And, to judge from the graph of puzzles solved released at wrapup, we were neck and neck with the other top team for the first twenty four hours before pulling ahead Saturday night.
But then we got logjammed on the last handful of puzzles. It was a combination of some tough late appearing puzzles and some puzzles that we'd just gotten stuck on for 24 hours or more and turned out to be meta-critical. We spent a good portion of Sunday stuck on about ten puzzles unsolved and after starting Sunday with four or five metas left, knocked away the last few metas about one every three or four hours, agonizingly slowly.
We solved our penultimate meta at about 5:15 on Sunday afternoon and a few minutes later, someone from SETEC came to visit us and warn us that even if we solved our final meta quickly, we had been beaten to it by another team which was at a sufficiently advanced place in the final runaround that they could not be caught. Fifteen minutes later, the coin was found; We solved the final meta about fifteen minutes after that.
It was frustrating to come so close to winning and fall short, but hunting with Palindrome was incredible fun throughout and I'm looking forward to next year.
Standout puzzles I worked on include Loaded, a pattern recognition puzzle involving polyhedral dice that was the one puzzle all Hunt where I had all the key insights to solve.... American Icons, a puzzle that started with us cutting out and dressing paper dolls of American feminist icons, before I had the insight that the dolls were analogs of American Girl dolls (thanks to my sister for this osmosed knowledge)... The Turducken Konundrum, an amazing and hilarious set of nested Duck Konundra... Clued Connections, an Only Connect extravaganza... and Bloom Filter, a lovely combination of a programming puzzle, and a word connections puzzle.
I think the personal highlight of the hunt for me was figuring out the final answer to American Icons on the walk from our secondary room to our primary room for the team meeting where we had been about to report that we were stuck and needed help figuring out the final answer to the puzzle. I immediately turned from a walk to a sprint to get into the room and ask someone to call in the answer, and when our team captain asked us how we were doing on the puzzle I was able to say "In about a minute they're going to confirm we got the correct answer."
The complete list of puzzles is here: http://web.mit.edu/puzzle/www/2019/puzzle.html