||[Jan. 21st, 2009|02:35 am]
This was a strange mystery hunt for me. My team (Metaphysical Plant, which wrote the S.P.I.E.S. hunt) got our asses kicked, solving close to all the regular puzzles but only one second-half meta. Yet unlike a lot of recent years which defeated us, we were steadily productive and having fun throughout. The puzzles I worked on were my largely wonderful; I have few complaints about them. And the theme, web design, kickoff, and so forth were great. This added up to my favorite hunt in several years (excluding, of course, the S.P.I.E.S. hunt, about which I cannot be objective). It ran long, but it didn't feel long until the very end.|
This all made it a successful hunt in my book, and yet it was not unflawed. The metas and structure were overly complicated and too hard, and the hunt as a whole rewarded large teams to a degree that makes me uncomfortable, even if I can't come up with a good solution to it. (Maybe this is not a real problem; if the hunt can support a dozen competitive teams, then it's unclear that it's a problem that a lot of them are 50, 75, or 100 or more people. But it means no individual can see more than a small fraction of the hunt. I had my most successful solving year in a long time, and yet there are great puzzles out there I never glanced at.) The second-half metas in particular required too many aha moments. I am firmly of the view that one should be able to describe how a meta works (really, how a puzzle works) in two sentences: one for the aha and one for the execution. (So for our two hardest metas in 2006: "The answers contain overlapping four-letter strings. Form a chain with these strings and collapse the chain to form the answer"; "The answers contain three-letter abbreviations for amino acids. Translate the codes into the one-letter amino acid abbreviations and read in map order to form the answer.") This favors what we on Plant call "pure" metas (which require only the answers, and maybe an external ordering) over "shell" metas (which require external information to solve the puzzle), but similar rules could probably be written for shell metas. (In particular, "avoid having meaningless extra information; it will almost certainly lead teams down the wrong path.") Yet as described at wrap-up (and I haven't seen the solutions; perhaps they will seem more reasonable in writing), it seemed a lot of the second-half metas had about three steps too many. Again, I don't want any of this to take away from what a great job the Bombers did — this was the most fun I've had solving a Mystery Hunt, throughout, possibly ever.
I'm not sure what I think of the different round structures in the second half of the hunt. I've thought about doing something similar should we write again, and I generally liked the variability and innovations that the different structures let them come up with. But it might have been a little too much; it was hard at times to keep track of all the different rounds. Seven might be too many; maybe three or four is more reasonable. The Dr. Who round, in particular, was the sort of multi-level structural innovation that really should have been appreciated over multiple rounds. (For instance, we never really got to the point where we treated the two halves as parts of the same puzzle; they were in our database separately and solved separately, which caused some issues.) I did love the different ways the Bombers incorporated interactions with other teams, including the online game and the scavenger hunt. I thought a lot in 2006 about how to do a new and different scavenger hunt, and didn't come up with any great ideas; this one was brilliant.
I loved almost all the puzzles I worked on this year, especially Jenn Braun's dancing puzzle, the airport codes half of the cricket puzzle, and Astro Traffic Control (which we did not solve, for our own dumb fault, because we tried every conceivable ordering other than the entirely reasonable one which was used). There are probably others I'm not thinking of; I will undoubtedly have more comments when the individual puzzles and their solutions are up on the web.