Log in

No account? Create an account

6th place! - Chronicles of Harry

2014 Apr 6

17:12:00 - 6th place!

Previous Entry Share Next Entry

Stopping by here briefly to record this. (I think I may actually begin posting some more once my defense and everything is done. For now, though...)

SO! Yesterday was the Microsoft College Puzzle Challenge. (Apparently "college" includes grad students. And people who finished their PhDs a year ago.) It's a 10-hour puzzlehunt for teams of 4 people. (You have to be in one of several particular locations to do it.) They do it every year but I'd not done it before. Dave Putnins was running a team, Good Fibrations. (A team name with a history of doing well at CPC, winning UMichigan every year since 2010 and winning the whole thing in 2012.) The team was him, me, Jordan (hence the "and people who finished their PhDs a year ago" above), and Leda, who is not from the math department but instead from the solar car team (and the only actual undergrad on the team). There was another math department team, consisting of Ari, Chris, Alex, and Dave Prigge; I'm not sure what their team name was.

Long story short, we came in 6th overall and won UMichigan. (Second place UMichigan was 20th place overall.) We didn't solve the meta; if we had we'd have been 3rd. We were also damn close to 5th... more on that in a bit. I was kind of hoping for that Xbox One (which, yes, I would probably just use as an expensive Geometry Wars machine), but, a good showing nonetheless.

A note on the format: The format was apparently a little different than in previous years. Instead of the winner being the first to solve the meta, there was a scoring system. Solving a puzzle was worth 100,000 points ("dollars"; there was a Las Vegas theme this year, ugh), while solving the meta was worth 1,000,000; you could keep going after the meta for more points until time ran out at 10:00. Not sure how much the pre-event puzzles were worth.

Additionally, on the ordinary (non-meta) puzzles, there was a bonus for solving the puzzle early -- 10,000 to the first team to solve it, around 7,000 for the next team, and decreasing pretty quickly; you had to be very fast to get more than 4,000 on bonus. Much more important than the bonus, though, was the betting. On the ordinary (non-meta) puzzles, you could bet up to 10,000 points that your first answer was right (much larger than you were likely to get from the bonus). You could not bet on answers beyond the first. Also, if you bet on your first answer, but you got it wrong with one of the "we'll give you a hint" possibilities, the bet would just be cancelled (though you still couldn't bet on later answers).

Pre-event puzzles: Yeah, I didn't do any of these; I have enough on my plate. Fortunately someone on our team (presumably Dave) did.

All Flights Lead to Vegas: We got the airlines, got the terminals, and then got stuck. This was one of only like 5 we didn't solve.

Bad With Names: I didn't look at this one until pretty late; somehow nobody on our team noticed before me that the length differences were all distinct and hence likely a sort order.

Beat the House: The metapuzzle. Ugh. We spent a while thinking this was not the metapuzzle but rather something weird like Don't Trust the House. We also didn't start on the meta until Wave 3, when we probably could have started on it earlier. Ultimately we didn't even get as far as finding the first message, because we apparently messed up the dealing. We were also misled by a mistake on this one -- instead of there being two instances of 7♥(BURN), there was one instance of 7♥(BURN) and one of 7♥(START) (the latter was a mistake they informed us of pretty late). So until pretty late we thought that meant you had to do the whole thing twice, once with 7♥ being burned and once with it not.

Don't Trust the House: Jordan got this one. I dunno how he came up with it.

Game of Numbers: This one had a typo in it originally -- the one with the factorials; originally those + signs were all - signs. This yields a number far larger than 26! (And yes, that's still true if you interpret it was "far larger than 26 factorial".)

It's In Your DNA: This was the last one we solved; it was in like the last 20 minutes or so. For most of the hunt we made a point of betting 10,000 on everything, which generally worked out for us. However once we had like 20 minutes left, we started spamming answers and betting 0. Jordan came up with the correct solution to this one after we'd already spammed a few DNA/RNA sequences. (Why would they give a clue saying to transcribe it if they wanted thymine and not uracil? Gaah.)

Keno: I'm pretty sure of all the ones we didn't solve, this is the one wasthe easiest, or at least, the one that the most other teams solved. (I think some teams may have solved it all the way back in Wave 1.) While I didn't get what was going on, I did while playing around notice the invariant that the +'s always summed to 186 (!). But that's not very helpful by itself. Jordan finally figured out the +'s near the very end. We never got the "..."s, which given the team you'd think we would have. But part of the problem is correlating them to one another.

Mating Games: This was another one we solved late; of course, this was also one of the hardest puzzles. They kept giving out such useless hints! Jordan finally put it together near the end.

Mobius Strip: This one frustrated us for a while, because the people doing the original maze hadn't crossed out the letters used. Finally Jordan suggested doing this, but the strip had already been cut, making retracing the maze difficult; we had to retrace the maze on an uncut PDF copy while matching the path to our cut-up copy. Once we finally got the answer, there was the question -- do we go with "Monte Carlo Resort and Casino", or just "Monte Carlo"? After the fiasco with "The ABCs of Nevada" (see below), we decided to go with the former, bet on it -- and lost. After "Monte Carlo" turned out to be the correct answer, though, we complained to them, and they honored it.

Natural Odds: Dave spent a while thinking that the winner actually becomes the loser here (effectively making it loser wins), until I suggested that perhaps we should just pay attention to the winners for the purposes of filling out the bracket, and worry about "You are what you eat" later.

Ocean's Eleven: The betting system largely worked pretty well, but in some cases it was a little bullshit. It ultimately ended up not applying here -- our first answer was "The Stardust", which was a hint answer, so we didn't win the bet but didn't lose anything either. What's a little bullshit is that our next answer, "William John Brennan", was counted as wrong, when the right answer was "Bill Brennan" -- good thing we didn't get it "right" the first time! (That said, if we had, and complained, they would have honored it, I think. See Mobius Strip above.)

Please Don't: The team was kind of annoyed at this puzzle. We spent so long on it and then so much of it turned out to be irrelevant. (People also found the answer a bit weird, because it's not any sort of common phrase or anything.)

Revolving Roulette Wheel: This was a silly puzzle. We did nothing with it for quite a bit, then I bothered to actually translate the text on the bullet ("Russian alphabet"), prompting Leda to ask, how many letters are there in the Russian alphabet? And the rest was straightforward. When we got the puzzle we were quite confused about the backwards numbers, but once I started actually decoding, it made sense. (Also, I did not realize that "Russian alphabet" and "Cyrillic alphabet" are not synonymous.)

Seattle Seattle: The T-shirt puzzle. Since of course you get the T-shirt before the contest really starts, and it wasn't a hard puzzle, we solved it before the contest really started, for a nice bonus once it actually did start. (Well, we didn't find it a hard puzzle, anyway.) (I'm a little disappointed with the shirt as a shirt, actually, since the puzzle doesn't obviously look like a puzzle.) Anyway, we noticed the variable lengths early, but were focused on the spots, not the gaps between them; there are 3 lengths of spots. Jordan started tracing one of the chips because, why not, and I pointed out that this could be used to make a ruler allowing us to sort them into discrete lengths. Giving us some ternary numbers, until someone (I don't know who) suggested measuring the gaps instead.

Show Me the Money: Ugh. I spent a while trying to find an online multicolor nonogram solver; all the ones I could find were single-color. Finally I found a downloadable one (for Windows, so had to run it under Wine) that would do multicolor. But on downloading it, turned out it was shareware, and multicolor was a paid feature! So, I started doing it by hand. It was pretty easy... except for the greens. I just got totally stuck on the greens. Eventually Jordan got the greens but said it was one of the worst nonograms he'd ever done. Wondering if we might have been able to do extraction without all the greens.

Soccer Facts... and Fiction!: Another one we didn't solve. We were kind of on the right track. I thought that the numbers, because they were a sort order, wouldn't also be an offset. I was wrong.

Table Stakes: We didn't get this one until the hint, after which it became easy.

The ABCs of Nevada: Ugh. Leda and I solved the clues, then got stuck. Fortunately Dave and Jordan unstuck us and we got the answer (with some mixups; I miscounted the number of letters in SECTION). (It does seem a bit odd to me to have all those middle boxes if they're just going to be filled in with the alphabet, in order, but whatever). But then there was the question of just what to submit as the answer. The whole book title? Well, Dave tried that first, but it wasn't counted because apparently colons aren't a valid answer character. I made the suggestion of just submitting "Silver", which we did, and bet on, and lost. Ugh. After that we tried the whole book title but without the colon, which worked; apparently omitting the subtitle would have worked, too, but not just the word "Silver". If not for that, we'd have had 20,000 points more, enough to put us in 5th place rather than 6th. A bad idea.

The Line: Aside from "Over is dash", we got nowhere on this.

What Happens in Vegas, Stays in Vegas: Ugh; not a good puzzle. Leda was handling this one, but got it wrong the first time. So what did we do? Well, there were only finitely many possible answers, and we'd already used our bet, so we brute-forced it!

When the Chips are Down, Look to the Stars: Another puzzle with lots of unnecessary shit going on!

OK. That's all I've got to say. OK maybe not all but I'm not writing any more.



[User Picture]
Date:2014 Apr 7 10:01:49 (UTC)
Hay! Good luck with your defense!
(Reply) (Thread)
[User Picture]
Date:2014 Apr 7 18:33:43 (UTC)
Oh, thanks!
(Reply) (Parent) (Thread)