Greg (gfpuzz) wrote,

The Microsoft Non-Intern Puzzleday 2009

Puzzleday is one of the puzzle events Microsoft puts on for its interns, sponsored by the intern program. It's a light version of the Microsoft Puzzle Hunt (which is in turn an adaptation of the MIT Mystery Hunt for Microsoft); thirty-odd relatively easy puzzles in eight hours. This year, the staff tried an experiment of running the event again with no budget for non-interns, whoever might be interested in playing: teams of six rather than twelve, no free T-shirts or copies of physical puzzles for everyone, etc. It seemed like the goals were to appeal to a less hardcore set or people who wanted to try puzzling but were intimidated by the difficulty and length of the MS Puzzle Hunt or the running around of Puzzle Safari, as well as generally raise interest/awareness in potential future volunteers and such.

I solved with TBD, which was half a chunk of my intern team that won the 2007 Intern Puzzleday, and half assorted friends and friends-of-friends. Personally, I fondly remembered the '07 Puzzleday for being able to reuse the awesome wands from the Hogwarts Bay Area Game in 2006- I knew they weren't likely to top that, but it'd still be a fun experience regardless. And my expectations were fulfilled.

We arrived in the atrium of Studio A for the introduction, which was pretty low-key. Pavel Curtis had a table with a few of his nifty physical puzzles (minus the one that was part of the event, of course), and I spent some time wandering around seeing who all I recognized and counting puzzle shirts. That's the one way to objectively measure how experienced/hardcore the crowd was- only six distinct shirts, not counting this year's Puzzleday shirt that the staffers wore. (Three Intern Games, one past Puzzleday, one Mystery Hunt team shirt, and one Puzzle Safari polo) I expect the numbers to be higher at SNAP VI in a few weeks. Eventually, Puzzle Control gave the rules/intro shpiel, the skit the interns got, and passed out all the paper puzzles for the initial dash back to the conference rooms.

And then, um, we solved puzzles in a conference room for a while. Various subsets of us left the conference room from time to time to solve some live puzzles. (Because they didn't have the budget for this event to make a copy of the physical puzzles for every team, they set up the extra copies left over from the intern version in a room and had us send a couple people out to do them) It's hard to make much of a narrative out of that.

As we solved more puzzles, we began to start looking at the metapuzzles. Puzzleday has used a two-stage meta for the last couple years: a first meta that you need, say, a third to a half of the puzzle answers for, and a second meta that you need almost all of the puzzle answers for. Solving the first meta gets you the second meta; solving the second meta means you win. The first meta was pretty straightforward, taking the Nth letter of different puzzle answers, but we kind of choked on it by not getting until we had all the information we needed. I blame myself for that; I was the one looking hardest at the meta because my partial-solving engine is pretty well-tuned, but the two non-words in the answer threw it off. So once we solved that and received the second meta, a pretty interesting criss-cross where solving a puzzle would give you a transformation that would turn a word from a word list to one of the words that fit in the grid, I knew we had to focus on that. I don't think I solved any normal puzzles once the second meta unlocked, as we plugged in all the data we had from puzzles we'd solved before getting there.

Eventually, a few more answers trickled in and got us the breakthrough we needed to fill things in, and my partial-solve fu actually worked this time, so we got our answer. (Conveniently equidistant from all the rooms that teams could have been working in, too) A call to Puzzle Control let them know to expect us, and we sprinted out the door... only to wait for the slowest member of our team at the elevator. And then we got there a few minutes ahead of the staff, finding the stack of Kit-Kat bars (Marathon bars in the intern version, of course) and taking the first to label us the winning team. The staffers arrived, congratulated us and got pictures, and showed us back to their secret headquarters where we could watch the Big Board and chat for a bit.

This was about 2 PM, and the event ended at 6, so after cleaning up our conference room we hung out for a bit on campus. I went home and took a nap so I'd have some energy for the wrap-up. And so I got back to campus for the wrap-up pretty early, as the puzzles were set up with a bunch of green and red Post-It notes for us to label ones we liked or disliked, and returned to the "Who do I recognize?" game. (I didn't fully realize until they were going over puzzles from the solution site where the authors were credited that the one member of staff that looked kind of familiar but I couldn't place was from my D&D campaign last summer at Google) And so the wrap-up began, they showed us assorted videos from the intern version, went over all the puzzle answers, and finally formally announced us as the winners. We got a giant cookie as a prize, which we split later over our post-hunt dinner.

All in all, this was definitely a good time. There were a few logistical seams, some introduced by the format shift and some utterly random. Our people on the physical puzzles that turned into "Go X and do the puzzle" setups didn't know when they had enough data that they could just transcribe what they had and go back. This question wasn't even in scope for the intern version, since their copies of the physical puzzles were in the room with them. But on the sillier side, the first puzzle I looked at happened to be misprinted in such a way that you could do the interesting part of the solving, but when you tried to go through with the answer, you got garbage. But I'll forgive them those because it was, overall, a good fun:cost ratio for the community. And I hope it encourages new folks to either try the other events or look into helping out.

As with Puzzle Safari, it's hard to talk about specifically cool puzzles because the default mindset is to churn through paper puzzles as fast possible. But going through the puzzles on the intranet, here are a couple interesting ones:
•Zeus Juice, a Powerthirst spoof. I didn't solve it myself, but I couldn't help but laugh as I overheard my teammates playing it. Just watch the video.
•Sisyphus, a live event in the atrium of Studio A where the hunt began. We were given a five-foot beach ball and told to roll it up the stairs and around the building. As we went, we were stopped by challengers with pictures of animals to identify- a wrong answer would have the challenger throw the ball over the balcony and force us to start over. Fortunately, we only had to restart twice; the "Back so soon?" would have gotten annoying otherwise.
•There were a couple of well-done wordplay puzzles. I enjoyed my time with Mythic Mumblings (Mad Gabs clues of mythic-themed video games) and Hetero-Nymphs (heteronyms, words with the same spelling that are pronounced differently), even though they were pretty quick.
•I didn't get to see the Theseus and the Minotaur mod with two Theseuses and Minotaurs (who followed the closer Theseus, allowing for more ways to trick the AI), but after hearing about it in the wrapup, I want to forget the solution he showed us so I can try it on my own.
•Gateway to the Gods, a Rush Hour puzzle where you got the answer by looking at letters the moving blocks revealed. The message allowed for an interesting way around the Rush Hour component.

I will primarily put on my aesthetic criticism hat to say that, yeah, I know that constraining puzzle answers is a mess logistically, but Meta 2 would have been so much cooler (though also a lot harder to partial solve) if instead of transforming words from the arbitrary list provided, they'd been the other puzzle answers. And yeah, the theming in 2007 with Harry Potter and the wands and stuff was a lot better for this, but I would have preferred a little bit more than "Hey, you're solving puzzles for some reason".

They posted some statistics at the wrap-up showing that basically team #N to finish in the non-intern version was about an hour ahead of team #N from the intern version. Doing a little math from that and the smaller team sizes means that, in terms of raw puzzles/person-hour, one person on TBD is worth 2.5 interns. I will set aside the differences in experience and added difficulty of mobilizing a 12-man team as opposed to a 6-man one and merely point to this as evidence that I should be paid as much as 2.5 interns.
Tags: puzzleday, writeups

  • MSIG 2013: Seattle Experimental Telemetry Initiative

    Another year, another Intern Game. This year followed the Seattle Experimental Telemetry Initiative- on the pre-game website, the interns helped…

  • WarTron

    I have gotten kind of lax about writeups, haven't I? I could do a very quick rundown of past events, and post that WHO writeup I've got lying around…

  • Microsoft Puzzlehunt 14

    Trying to get through my backlog of events to write up- I'm tackling the Puzzlehunt now, since all the puzzles are available online to refresh my…

  • Post a new comment


    default userpic

    Your IP address will be recorded 

    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.