2019 MIT Mystery Hunt, Part 5: Acknowledgments

(This is a recap/review of the 2019 MIT Mystery Hunt, which happened in January. Puzzles and solutions can currently be found here. These posts will contain many spoilers. Maybe not so much this one.)

It’s an atypical Thursday night at Chez Dan, as Jackie is on an overnight trip to NYC, and I’ve been alternating between solving logic puzzles and catching up on wrestling. (Okay, my portion isn’t that atypical.) In any event, this seems like a good time to write up my penultimate Hunt post, which may be of less interest to puzzle enthusiasts… I’ve said what I want to say about Hunt content, and now I want to thank some people who made Hunt happen.

I should start by saying that everyone on Setec (listed here) contributed to Hunt and everyone was important; I’m not writing this post to try to identify everybody or try to single out who did the most work, but rather just to tip my hat to the people who contributed especially to my own Hunt-writing experience.

Setec picked up a whole bunch of new members (from Illegal Immoral & Fattening, Left as an Exercise for the Reader, and Om [some number of Noms that I can’t remember]) this year specifically for construction. Setec was not what I would consider a small team when we won, but we weren’t big either, and the Hunt has grown to a scale that I don’t think we could have written without recruiting. So thank you to Sami Casanova, Chris Cieslik, Jesse Gelles, Avram Gottschlich, Albert Lin (actually returning to Setec after a long time away), John McLaren (repeatedly called Australian John despite being the only John on our team), Matto Mildenberger, Julia Tenney, Julia Urquhart, Rebecca Vessenes, Mike and Sandy Walsh, and anybody else I may have missed. Without you, Hunt likely would not have happened, and I hope some of you stick around to solve with us in January.

We had our executive structure mostly figured out before the new folks joined us, which probably made it difficult for them to wiggle their way into big-picture roles. Chris Cieslik thankfully insisted on being involved in the nuts and bolts of releasing and ordering puzzles, and he built some simulations that provided a nice counterpoint to my approach of “my instincts from running Hunt before say this feels right.” These are the advantages of having an experienced game designer in your ranks… I now realize I’m too late to plug his company’s Kickstarter, but you can read about the successfully funded game here.

I’d also like to credit Rebecca Vessenes for taking control of the midgame (the birthday party interaction) late in the year along with Justin Werfel. I led the midgame construction in 2017, and a big priority was making the whole thing run efficiently, since I knew we’d be executing dozens of times with varying groups of people. My biggest worry in handing it off was that whoever took it over might not be as organized, but Rebecca did phenomenal work in gathering and constructing the props and equipment, and more importantly, idiot-proofing the organization and instructions so that anybody could portray the Fool and their handler. Given that none of us had done any event planning with Rebecca before, we realy lucked into having the best possible person doing this.

Speaking of Justin, I don’t know if he noticed it at the time (it might just have been my being pompous), but the first year we hunted together with Setec, our personalities didn’t gel well at all. But I’m glad we’ve now had the chance to construct together, because it turns out he’s a great collaborator, and in particular, as I said in my story post, I really appreciate the character development (and scriptwriting) he contributed to weave the April Fool into the plot. In my hands, the Fool would likely have been a generic chaotic evil villain (that’s what I do), but the story we told was a lot more nuanced, and a lot of that is Justin’s doing.

Scripts are only as good as the people who perform them, and so thank you to Marisa Debowsky, Philip Loh, Josh Oratz, Greg and Margot Pliska, Scott Purdy, Julia Urquhart, and particularly Jesse Gelles and Steve Peters for bringing the kickoff to life. I had never seen Jesse perform before, and I cast him based on meeting him for a couple of days at our construction retreat, and he surpassed my expectations as Jack. In contrast, I’ve seen Steve perform tons of times, as we did a lot of musical theater together at MIT. Given that Steve gave me my first significant role in college (as Adam/Noah in Children of Eden), it was nice to finally get to direct him in something.

I cannot even fathom Matt Gruskin’s computer skills. Certainly he contributed to the website and made Hunt happen from a tech perspective, but I lost count of the number of times we needed an interactive implementation of a puzzle or a simulation of a tester saying “what if I asked the coding guy on my team to do this?” and our default approach was, “Someone ask Matt.” Frequently I’d tell him something was no hurry and still get a response in less than an hour that was exactly what I wanted. I was sitting across from him in HQ on Sunday morning when the Hunt website went down–conveniently right after TK, our head of tech, had left to sleep–and I watched as he got on the phone with TK, and they calmly identified and fixed the problem. (To be fair, when I say calmly, Matt may have just hidden his panic well, and I only heard half the conversation, so for all I know TK was screaming.) Arthur C. Clarke said that any sufficiently advanced technology is indistinguishable from magic, and as far as I (with my pitifully amateur coding skills) am concerned, Matt and TK are sorcerers that put Mystereo Cantos to shame.

Speaking of TK (Focht), I believe that after 2017 I said he was one of a couple of people who kept me sane through the year. That was true this year as well, and since I was running metas and he was running story/structure, we had to interact a lot. TK and I often have matching perspectives about Hunt philosophical issues, and the nice thing about having a person on your wavelength like that is that when they do disagree, you tend to trust them. I know I’ve done it before, but at this point when I imagine writing a Mystery Hunt without TK on my team, I curl up in a corner and cry. Okay, to be fair, right now I do that when I imagine writing any Mystery Hunt. (I’m done for a while, for real this time.)

Almost last but by no means least, there is Tanis O’Connor, who was in charge of the whole dog and pony show. Unlike my creative rhythm with TK, Tanis and I frequently don’t approach things the same way by default, and in 2017, we had a significant number of arguments and personality clashes. So to Tanis’s credit, when we won the Hunt again, she immediately initiated a conversation with me about how we could work together without getting on each other’s nerves. We still didn’t always see eye to eye, and we still had a few disputes throughout the year, but we collaborated much more productively than last time, and I suspect that where we met in the middle was farther from her comfort zone than it was from mine. I am deeply grateful for her flexibility and her patience, and that’s just in dealing with me; as editor-in-chief, she had to keep watch over all the puzzles and deal with an entire team worth of constructors with idiosyncratic creative processes, and she made it happen. If it weren’t for Chris Morse, Setec Astronomy wouldn’t exist, and I’m glad he let me join the team back in 1998(!) when I was still in high school. But if it weren’t for Tanis O’Connor, neither of the last two Setec Hunts would exist, and I hope people appreciate all the time she’s volunteered.

Also, I don’t know if I talked about this in an earlier post, but when I unsuccessfully tested Funkin’ at the retreat (in its original form that involved mini-CDs rather than USB sticks) I thought the concept was a logistical nightmare. I insisted that we should have a website version ready to deploy if we didn’t get the donuts as intended, which I honestly didn’t think we would. I eventually leaned extra-hard into my role as a “donut truther,” repeatedly referring to “the donuts that won’t actually exist” and consistently generating rude gestures in response from Tanis, who was guaranteeing she could make the donuts happen. Hear ye, hear ye, internet. The donuts happened and were one of the most memorable elements of Hunt for many teams. Tanis was right, and I was totally wrong.

Finally, I want to thank Jackie Anderson, because if you’ve never lived with me in a year when I was writing Mystery Hunt, you have no idea what she’s been going through. (I heard at least four accounts of executive committee members’ spouses, some of whom are on Setec and some of whom aren’t, looking forward to late January when they’d finally get their husband/wife back.)

Thanks to everybody I mentioned above, and everybody else on the team; if you pick a name on the credits list, I could tell you something they did that made this a better event. But more than 1600 words in, I’m going to call this done, get some sleep before Galactic starts tomorrow evening, and hopefully moderate a backsolving discussion in a future post before the Hunt is three months old. Cheers.

4 thoughts on “2019 MIT Mystery Hunt, Part 5: Acknowledgments

  1. Can you give some more details on what Chris Cieslik’s simulations were like? Is it “assuming a team has size N, they’ll solve M puzzles per hour,” or is there something more complex?

    Like

    • The main purpose of the simulation was to get a view of meta-progress based on our unlock order. It took in the order puzzles would be unlocked, and had the simulated-team randomly solve puzzles every so often. The solve order and solve time were weighted in two ways:

      1) The simulation assumed that a team solving on average X puzzles per hour would begin hunt solving 2X per hour and end hunt solving X/2 per hour. I arrived at this figure from looking at past hunts, and keeping in mind that our early puzzles weren’t quite as easy as emotion/fish puzzles, and that backsolving (except early) was not easy at all in our hunt.

      2) The simulation assumed that a team was about three times as likely to solve the puzzle they’d had opened the longest as the puzzle they’d just opened.

      This is a big over-simplification of how teams solve, but I wasn’t looking to build an ultra-accurate AI-simulation of 40 people in a room, just a good estimate :)https://puzzlvaria.wordpress.com/2019/03/14/2019-mit-mystery-hunt-part-5-acknowledgments/#respond

      With this, I tossed some numbers into the simulator, seeing where hunt would progress for a team at an average 1, 1.5, 2, 2.5, 3, 4, 6(!), 10(!!) puzzles per hour. This was useful for a couple reasons. Most importantly, it gave a general idea of when hunt could end. Mid-Late Saturday if things went quite speedily, or sometime Sunday in all likelyhood (My guess was early Sunday evening, which turned out pretty good!) The other thing that it was very useful for, was getting a grasp on when the mid-game birthday interaction was likely to happen, and when teams might start requesting donuts, swanlumps, etc. This helped us plan staffing in an informed way, which was great.

      Once hunt started, every 3-4 hours I’d do some simple calculations of the top teams solves-per-hour and fit that to the simulation curve. That let us project where hunt *might* end, and it pointed at Sunday for most of the weekend. For some of Saturday we thought end-game might possibly happen around 5 or 6AM, but then the top teams slowed down (especially on metas).

      Like

  2. No idea why that URL got pasted into the middle of my post, oops!

    Anyway, it was lots of fun to watch teams progress over the weekend and guess when things might end, our big-board visual display was a great venue for that. John and Matt and the rest of the tech team did a great job putting it together!

    Like

Leave a comment