Previous Entry Add to Memories Share Next Entry
2012 Mystery Hunt
[info]cananian

This year I was a member of Codex, the writing team for the 2012 Mystery Hunt. I'm going to describe some of the puzzles I wrote for "The Producers" hunt, in release order. BEWARE SPOILERS!

  1. One of the early theme proposals for our hunt was "Alice in Wonderland." Casting about for novel meta ideas, I hit upon the idea of a round with purely numeric answers, 1 through 29,394, which would resolve to words via "looking glass numbers"—that is, numbering all the words in "Through the Looking Glass". It occurred to me that you could make your numbering system self-descriptive if you used certain words; for example, if you wanted to make clear that hyphenated words should be counted as one (instead of two), you could include "great" and "half" on either side of "arm-chair". The numbering of "great" (164) and "half" (166) would make it clear that "arm-chair" should be treated as a single number (165).

    This didn't survive as a meta, but it eventually became a puzzle, called 1207 1370 (which translates to "Looking-Glass Words" using its enumeration system). It also served to ensure that teams had a good wordlist by the time they got to the Charles Dodgson meta...

  2. Blinkenlights. A recursive-structured puzzle inspired by (but not reaching the greatness of) Derek Kisman's Maze from Setec's '05 Hunt. If anyone is mourning the lack of Jonathan Coulton-related puzzles from this year's hunt, blame me: I stole the answer PROTECTORS which Andrew Lin had earmarked for a JoCo puzzle. ("Did I say overlords?")

  3. Caterpillars. I like giving physical objects to teams. This was another failed meta—you would have assembled the pieces out of words, then would have to assemble the jigsaw from the word-pieces. The location of the caterpillars' heads in the final assembly would spell out the final meta answer using an overlay. But the puzzle is more fun with tangible pieces, I think.

  4. B.J. Blazkowicz in ‘Wintertime for Hitler’. I was writing the meta for this round and trying to find non-dictionary words. I needed "CAR..." as a prefix to make the chess game work, which suggested CARMACK as an answer, and the puzzle just wrote itself from there. Scott Handelman contributed the title. This puzzle was going to be distributed on 3.5" disks (remember how I said I like giving teams physical objects?), but the last 3.5" floppy disk puzzle was Blue Steel in '06. (Redundant Obsolescence doesn't count, since the 5 1/4" disk was redundant.) The past six years have not been kind to the 3.5" floppy; ultimately we decided we didn't want to deny teams the pleasure of playing the game because they couldn't locate a floppy drive. It's more important that puzzles be fun than hard!

  5. Charles Lutwidge Dodgson meta. I began writing this puzzle immediately after the 2011 hunt, dissatisfied with the mechanism and final clue phrase of that year's Racking Your Brains. I thought I could write a better puzzle using Scrabble Solitaire as a mechanism.

    Slightly later it became part of the "Alice in Wonderland" theme proposal, with Jabberwocky words. Then I spent a couple of months away from the hunt, getting married.

    Upon returning we badly needed critic metas so I dusted off the puzzle, adding an Alice chess frontend yielding the tile string in order to make it a shell meta. The puzzle can still be solved as pure Scrabble Solitaire (ie, without the given "scores after each play") but it's easier for humans to solve with the frequent checkpoints given. For what it's worth, I constructed the chess game with a reasonably-deep alpha-beta search, so all the moves "make sense" as much as is possible given the constraints of the puzzle. And it ends in a clean checkmate, obviously... I have no idea how BENOISY snuck in there.

  6. Ben Bitdiddle meta. The idea of making an electronic circuit which was impossible to assemble incorrectly had been in my "Mystery Hunt ideas" folder for years. A coworker at OLPC mentioned the odd power-pin configuration of the PIC chips one day, which gave me the "flip" mechanism. Brainstorming with Andrew Lin brought it the rest of the way.

    I promise never to abuse an optoisolator in this way again.

    (Of course it turned out when constructing this puzzle that Ben Bitdiddle really needed to use the show answers CARPAL and THESOUTH because of their length in morse code, so I ended up having to rewrite parts of Dodgson to make Bitdiddle work. In the rewrite CARPAL became CARMACK... and B.J. Blazkowitcz was born.)

  7. JFK SHAGS A SAD SLIM LASS. One of my earliest puzzle submissions was, "A puzzle contained only in its title." Again, the fabulous Codex editor team turned this into a real puzzle.

Some puzzles I enjoyed editing:

  1. Revisiting History — I commissioned a Doctor Who-themed puzzle for the answer TORCHWOOD (see the final clue phrase for the reason why) and contributed the "location of the word 'who'" mechanism.

  2. Gibberish and More Gibberish. I liked the idea for this puzzle enough that I shoehorned a suitable answer into the Charles Dodgson meta... and then had to do some heavy lifting to get the puzzle finished and into the hunt.

  3. Sounds Good to Me. It was immediately obvious this was a brilliant idea from Seth Schoen. But the twin barriers of toki pona and hiragana threatened to make it unsolvable. I'd like to think I played a role in making this an accessible and solvable puzzle.

  4. Itinerant People of America. Same deal. Squiggles had bequeathed the world the facial expression described as, "That's my brain leaving out the back door while my face distracts you." My contribution here was solely instilling the fear of God into the authors. Scott Handleman describes how he and Emily Morgan took that advice and constructed a kick-ass puzzle.

And that's it for my puzzles! I also did a heck of a lot of other stuff for the hunt; I hope y'all enjoyed it. (My own favorite part was the wrap-up, since all my responsibilities had been discharged by then. I could just watch Patrick rock my hat and accordion, play along on ukulele, and sing tenor with Francis at the end.)


some of these are just lovely. thank you for posting.

i suspect the other ones are just lovely too. but require more energy and attnetion to appreciate that fact. but just idly flipping through JFK SHAGS A SAD SLIM LASS is very elegant. and the ben bitdiddle meta is just epic and very very clever.

Two quick questions:

1. Were there any paper progress sheets this year? Not hearing about them, I figured they went the way of the dodo, but I've also thought they were going to for years so I wanted to check!

2. Are there any spare Ben Bitdiddle components anywhere?

1. No paper progress sheets. Just the "big board" view from Novalis' hunt-running software

2. There are extra Bitdiddle and extra caterpillars. We'll be selling them to recoup expenses; stay tuned to puzzle-hunters.

Not that I mind a small contribution, but what if we just never picked them up?

We discussed this, and decided that teams had all weekend to pick them up, so it wasn't worth our bookkeeping to try to give out freebies only to teams which hadn't bothered to grab them during the hunt.

How did you make the caterpillars? 3D printer?

Laser cutter. We had on-team resources, but due to some unfortunate events danger!awesome came through at the last minute and made most of them for us. They are awesome.

Thanks for all your work-- this hunt seemed to go more smoothly than ever, and having answers posted *before* wrap-up was truly awesome.

Will there be more solving stats released? Specifically I'd like to see charts of chutzpah over time, and/or open puzzles. I'm curious how far ahead of the timed releases we (Central Services) stayed, and just when on Saturday did we get overwhelmed...

If people are making requests, let me suggest solutions for the meta-metas (including whatever happened with the critics and tickets, which seems to not be described), as well as puzzles (such as they are)/solutions for the events. I don't care if I read this stuff now or in three months, just that it is posted for a complete Hunt record before someone forgets. (You guys have generally done amazing on the posting front so far!)

Yes, those are all being worked on. I should be able to do another big update this weekend.

Fantastic! Maybe the best (fastest, most complete) solutions, etc. work anyone has ever done!

Timed release this year worked a bit differently than it has in the past. All releases were a function of time and solves. Every batch had a base release time, and a minutes early per chutzpah, where chutzpah=nsolved*(nsolved+1)*(2*nsolved+1), and nsolved=number of solved puzzles + number of shows produced. So, solving even a single puzzle put you ahead of time release.

Ah, I saw the formula but didn't quite grok its implications. So there was a single fixed release schedule, and it just got moved forward whenever you solved a puzzle? If we solved zero puzzles, would we still have had all puzzles open by Sunday at 3pm?

Yes, but the schedule was moved up by a variable amount (more for each solve, but less for later batches).

The final batch's base time was 44 hours after the hunt started, so Sunday morning.

Yes, actually you would have opened the last batch 44 hours after the start of the hunt (so by 9am Sunday, since we started at about 1 Saturday).

Each batch moved by a different amount for each solve; so solving one puzzle only would have had nearly no effect for the later batches, but by the time you were on the pace we had primed the releases for, the later solves moved the last batch by around a half hour.

This actually isn't too far different than the way it worked the last two years, as in the history hunt the clock kept ticking backwards (as in a time release), but every puzzle solved jumped it back by even more (and when the clock got to a new year that had an associated round, it got released), and last year the points had a time accumulation component as well as a puzzle solving component. There are differences, but for the most part they're more similar than different.

Wow, thanks for the "Maze" name drop. That made my day. :) I worked on (and very much enjoyed) Blinkenlights this Hunt, and it definitely reminded me of Maze, but I just chalked this up to serendipity. It's cool to find out that it wasn't a coincidence ... and that somebody actually remembers Maze fondly. ;) Of course, the big difference with Blinkenlights is that it's actually hand-solvable, rather than reducing to a grueling math puzzle. Oh, and you fixed the reverse-engineering issue...

I was hoping more people would treat Blinkenlights as a math puzzle instead of a clicking exercise. That was one of the successes of Maze; but Blinkenlights wasn't *quite* big enough to keep people from clicking at it for hours.

Was there ever a time during development when each critic was to be matched puzzliciously to a different show than the one that critic reviewed? There were some noticed correspondences: Watson 2.0 + Okla-Holmes-a, Sheila Sunshine + Into the Woodstock, Ben Bitdiddle + Phantom of the Operator, William S. Bergman (Burroughs) + Mayan Fair Lady...

Great hunt, btw (I was a remote solver on Luck).

No, that was a total coincidence... which we decided to take advantage of in the critic-interaction part of the endgame. Like we meant to do it all along! ;)