Previous Entry Share Next Entry
Puzzles I worked on (MIT Mystery Hunt 2013)

Quick summary of MIT Mystery Hunt 2013 — didn't work on as many puzzles this year, between (a) Not Trying To Win (after writing last year's hunt), and (b) Zachary. Did quite a bit of work on the hunt-solving software; discovered that the current limits to Meteor's scalability are "less than team Codex".

Puzzles I particularly liked:

  • Time Conundrum (except for final extraction)
  • Too Many Seacrests
  • Tuva or Bust (for which I successfully used Prolog in anger)

Other puzzles i worked on:

  • Square Routes
  • The Maze (helped guide final extraction)
  • Czar Cycle (which we never solved)
  • Road Trip (final extraction, with alexp)
  • Space Monkey Mafia (final extraction, with alexp)
  • Diagramless Crossmusic (knew how it worked, but meh)

There were a lot of very interesting puzzle ideas in this hunt. Several of them would have made excellent puzzles, given a bit more focused editing. In particular I want to single out 50/50 and Diagramless Crossmusic as great ideas.

Thanks to the Sages for all their work over the past year. The Mystery Hunt is a ton of work to write, and it's all done for the love of the thing.

I would definitely appreciate any feedback you have about how the software went for you. To my pleasure I did learn that it was "not less than team RoboPop", but I may have tailored what our app did to what I expected would work well (not to mention using it as inspiration for a bunch of optimizations over the last few months).

Post forthcoming! (either here or on meteor-talk)

Random is planning to try Meteor for our own internal app next year! I'm also interested in the scalability stuff (especially since Random's a pretty big team).

As the author of Tuva or Bust, I would like to state for the record, Prolog was not used in the construction or test solving of the puzzle, but I'm glad that it worked!