Corin Anderson (magellanic) wrote,
Corin Anderson
magellanic

Mystery Hunt 2013

Ah, the MIT Mystery Hunt. Another year, another event we'll be talking about for years to come.

I had the pleasure of playing with Leftout's West Coast contingent this year. We numbered about 16 at the peak on Friday and Saturday but ended with maybe 8 - 10 by Sunday evening and Monday morning. Rich's house is nicely arranged for these events: table in large kitchen is one solving zone; table in TV area is a second solving zone; some loose chairs make up a few more zones; and the living room in the front of the house is yet another zone. With exception of the living room these spaces are adjacent and without walls between them so it's easy to command everyone's attention or to eavesdrop on another conversation. We had a Hangout with Leftout East Coast on the wall-mounted TV so we had a view inside Boston. I made a nest in a corner between two zones and joined others here and there.

The event starts at noon in Boston with an in-person act layout out the story. Apparently, Enigma Valley Investment and Loan lived up to their evil name and "invested" the MIT Coin (ie, stole it). We want it back, so teams were to plan for its heist. How, exactly, we'll find out soon.

Ordinarily, the puzzling starts about 15 minutes after the end of the acting event, but GC told us they had some technical difficulties, please hang tight, first for an hour, then for another 45 minutes. We're champing to solve some puzzles, and we even ponder whether this is a puzzle -- after all, we're suppose to break in to this company's vault, right? But no, it's just real world technical glitches. The event does start, just two hours late (11a Pacific time). Great.

The event starts with a set of six puzzles which each solve to a name, the names of the people we'll recruit for our heist. Whom do you recruit for your heist? How about Danny Ocean, Maxwell Smart, Marty Bishop, and the like? Yes - these names feed into a meta puzzle that unlocks our full event: we're going to recruit each of six people (those three plus Indiana Jones, Richard Feynman, and Erno Rubik) for our heist. Each person represents one phase of the event with associated puzzles (with more or less theme) and meta puzzles. Also, after completing each phase (ie, after recruiting each person) the local team in Boston completes some additional activity for a particular obstacle in the way of our heist: dealing with the guards, cracking the safe, getting past the lasers, etc. The structure's pretty clear, actually very clear, and I really like it (in part because I like heist movies).

So we now begin the main part of the event. We have six phases but we've opened only some of the puzzles in the first phase. We expect (correctly) that we'll open further puzzles as we solve these, and so forth. I let other people pick up puzzles first, because I don't know exactly what I want to dive in to, but after some dust settles I start in on a neglected puzzle. There's a few MP3 files that play overlapping songs. I go about identifying the music and make a nice spreadsheet of everything. I'm kind of stuck. After a while I see someone adds a column for "album" and there's the insight -- each song comes from an album with a color: the White album, the Black album, etc. The puzzle flavor suggests we should "add another electronic band to the mix" which leads us to map the colors to numbers like resistors (there's four songs per file, too). But then, what? We have some very nice data -- we harvested data from the puzzle, did research to get more data, and had the insight about colors and numbers, but we're stuck. We try adding the resistor values and spelling words with black = B, red = R, etc., and we find a good word (ROBBY) but it's wrong. We try ever worse ideas for a while but we're stumped. We've invested three or four person-hours into this puzzle (working in parallel), we've done much of what we think we need, but no answer. Urgh.

And, that's a refrain that'll play out for much of the weekend. We expect puzzles in the Mystery hunt to be more challenging than those we face in most games, because we have the luxury of having hours to solve them. But this is a different challenge: we've done a lot of work, we have lots of derived data, but we're still stuck for what to do next. This pattern hits us for dozens of puzzles, and apparently not just for our team but for other teams as well, we learn later.

It's Friday evening now and I work on some other puzzles. I'll skip details of each of them but let me name a few from the weekend overall in just a little bit. Dinner arrives magically and we continue to puzzle. Melinda shows up around 9p, after class, and stays for a few hours before heading home; she teaches Saturday morning, too. I leave around 11p because I want to be refreshed for Saturday.

Typically, Mystery Hunts run until Sunday about noon Boston time, so until mid-morning for me this year. But heading into Saturday evening the team sees this event is going to be epic: we've completed none of the six phases and we've seen puzzles from just two of the six. And, we're apparently one of the teams doing better than most. Yikes. I decide to head home Saturday night, around 2:45, for some sleeping, too. Melinda stays up puzzling, having come by after teaching. Driving I-280 home is nice at 3a. Plus, the sound of the road is a wonderful break from music identification and Gangnam style (see below). I'm back Sunday morning to find we have puzzles from most of the phases available but we've completed only one meta. Sunday morning GC starts sending their e-mail messages: first, that their prediction for the end-of-hunt is Monday 9a. Second, that they're going to increase our rate of options accrual (the currency we're using to get a free answer on puzzles). We run with these changes through Sunday, but even by Sunday evening it's grim. GC drops more things: we need to complete only five phases, not six, to claim to be done, and, we get a free puzzle answer every hour. GC says they want the event to end soon, too, and they're doing all they can. Word from other teams is they're having not much better progress than we are.

We spend Sunday night and the wee hours on Monday sorting through the puzzles in order to identify which we'll take the free passes on. It's a little disheartening to take a bye on a puzzle you've spent hours on earlier (eg, the resistor code puzzle) but the answers feed into other puzzles that need progress.

I sleep at Rich's house Sunday night, just for three hours. I force myself awake around 5:45 because I don't want to catnap through the end of the event. It takes me 30 minutes to come back online -- I'm bleary-eyed and cold. We have a kitchen full of snacks, and a visit to Starbucks soon after helps. Monday morning GC lets teams start asking yes/no questions for hints in order to speed things up. We manage to eke out a few more answers but finally, just three minutes before noon (in Boston) Monday, GC announces that another team has completed all the puzzles. Woosh - just shy of 72 hours, 70 if you discount the late start, but 72 again if you include the time for the runaround at the end (even after solving everything there's typically a physical trek around campus to find the coin itself).

Melinda and I didn't stay much past the end at Rich's house -- we were ready for some sleep. A few people did stay and solved a few more puzzles but I'd exhausted everything I wanted to see.

Now, about those puzzles.

Megamix. The colors in the album names was a neat observation. It's a good example of what I like about puzzles: many disparate things (in this case, those songs) have something in common (an attribute of the album names) unexpectedly.

Diagramless crossmusic. This puzzle was renamed after I'd worked on it for hours - I wish I'd gotten this hint earlier. It's 263 MP3 files, each about three seconds long, and playing two overlapping songs. I thought about making a grid but the task seemed daunting. We identified maybe one quarter of the files. Did we need all of them? Dunno. We got this puzzle Saturday night but already our numbers were decreasing (many people at Rich's house could puzzle only until Saturday evening). I regret not trying to make a grid for this puzzle earlier. OTOH this could have been like the many other puzzles: invest a huge sum of time to make a spreadsheet of data and then get stuck at the last step. 8/

Sages Style. A YouTube video remix of Gangnam style parodies. Too bad Melinda was teaching when this puzzle came out. Justin Graham, Trisha, and I IDs a bunch of the videos, and the puzzle had a list of questions such as "number of ducks * 4". Lots of branches in the solving path: do we need to identify the original videos? Do we need to count the things ("ducks") in the complete, original parody or just in the clip as shown? Some parodies have none of the items listed, and some of the listed items don't seem to be anywhere. Some of the parodies aren't complete videos, just one or two scenes. We collected a bunch of data here and gave up because of these uncertainties. This puzzle is an example of us having ideas to try but without confidence things are correct we're unwilling to explore.

Space Monkey Mafia. Given is score after score of musical notes and lyrics, sorted by note and syllable. From the name one might find a stanza from "We didn't start the fire" among the syllables, and one can make that tune. Great. Now, identify the other 9 songs from this jumble. We identify three more of them but then stall. Hard to identify a song given 6x the possible notes and pieces of lyrics, and if those lyrics are "a" and "n't". I liked the core piece of this puzzle -- identify collections of things (syllables into songs) from a bigger soup of them. I wish it had been more tractable.

So good they named it Hull. I worked this puzzle with Doug. The cute insight here was using that cities in New England have the same names as those in England.

De-coins. Mystery Hunts often will have a scavenger hunt puzzle and this one was about coins. Currency coins, non-currency coins, alternative meanings of coins ("quartering of soldiers"), and Mystery Hunt coins. I contributed photos of several coins from my collection for this and was pleased with myself for having a few of the more rare items.

Golden Images. A bunch of photos of things, seemingly unrelated. What they do all have in common, it turns out, is that they're depicted on the Golden Record on Voyager. Some are images, some are sounds. Not sure how the puzzle part worked but neat to research about the Golden Record.

Phone tag. A phone system, call it and extract a bunch of data. It's sets of cities and area codes and lots of references to traveling salesman. I think we needed to find some traveling salesman problem solutions with some of the sets of the cities, but this puzzle has high branching factor -- lots of things one could do, all of them somewhat expensive to try so I'm reluctant to try any of them. I do try one or two and they don't work out. This marks the second time I've taken on a phone puzzle in a Hunt and given up. 8/

Snow Day. Themed around Blizzard games. Other people did the hard part but Jessen handed it to me when he had an inspiration (which proved correct) but couldn't try it at home. The puzzle ended with collecting data from high scores of a game one downloads for this puzzle. I was unreasonably pleased with myself on completing this puzzle, perhaps because it was a Sunday solve, at a time when we were running low on real solves (many of our "solves" were exchanging options for answers at this point).



So, the hunt was a much harder slog than in years past, but I still enjoyed the experience. Rich's house is a comfortable location for solving, I got to interact with people I don't usually play with (them being on other teams), and the weather was pleasant. I'd enjoy playing in Los Altos again and I'd enjoy going back to Boston.


(If you'd like to comment please do so on my G+ post.)
Tags: puzzling
Comments for this post were disabled by the author