Via Robert Buckingham, a graph of the length of the MIT Mystery Hunt:
I always quote 60 hours to shock people about how long it is, but that seems a little extreme.
January 5, 2011
History of Mystery Hunt Lengths
Posted by Anand Sarwate under Uncategorized | Tags: MIT, puzzles |[8] Comments
January 5, 2011 at 8:02 pm
Chris Morse is sitting next to me and questioning some of those data points — particularly the 2000 hunt, which he ran (Wizard of Oz). He claims that it was the shortest of all the hunts he ran and that it ended on 11:30 on Saturday. He is threatening to go downstairs and dig the proof out of the basement. I suppose I should tell Robbie this too…
January 6, 2011 at 4:38 am
Yeah, 2000 was a short one; I was on the team that won (Palindrome), and can confirm that we finished Saturday night.
January 6, 2011 at 7:12 am
As a follow-up, Chris did go down to the basement and get the logs of that hunt. PAINTNIAP (aka palindrome) found the coin at 2:17 AM Sunday, putting it at 38 hours. The Tech article Robbie sourced his data from was off by 24 hours.
January 6, 2011 at 12:29 pm
Doh, there goes the nice –..–..– etc. pattern.
Also, not to be too pedantic, but you mean “when the winning team found the coin” which is not necessarily synonymous with “the length of hunt” (see ’07 and ’10 at least, where many teams continued hunting after the coin was found).
January 6, 2011 at 12:32 pm
The html ate two of my dashes.
January 6, 2011 at 1:00 pm
Correction posted in a separate post.
January 6, 2011 at 1:12 pm
That two long, two short trend wasn’t quite there to begin with — Normalville (2005) is hiding itself in an almost perfect linear trend from 2004 to 2006.
January 6, 2011 at 12:55 pm
[...] made some corrections to his earlier plot: LikeBe the first to like this [...]