SnoutCast #175: The Famine Game Reaping

This week, we talk to Todd Etter–previously profiled in SnoutCast #101, now head Gamemaker of the upcoming Famine Game–about that event’s recently-concluded application process!

[ Download mp3 – 32 MB ]

00:00 – teaser: “several moments of [swearing]”
01:18 – “hungry”
04:45 – a GC perspective on the application process
07:42 – were final application numbers consistent with the initial survey?
10:40 – the value of GC’s Selection Criteria
15:41 – assigning teams to Districts (cf. Hogwarts houses)
17:31 – bribery is bad, m’kay
22:00 – application videos (including Team Snout‘s alleged swear-o-rama)
26:51 – application puzzles: hidden angst
30:32 – “Dad, is this another puzzle?”
34:12 – The End

Tell us we’re wrong on the Internet! E-mail podcast@snout.org or post a comment at www.snout.org/podcast.

Music: instrumentals from “Code Monkey” and “Ikea” by Jonathan Coulton

[ Subscribe to SnoutCast / iTunes link ]

Curtis DeeAnn Todd

0 Comments

  1. It sounds like Famine GC toyed with the idea of having some application puzzles tough enough such that not all applicants could solve them. The idea being: this gives the judges another axis on which to rank teams. But a big disadvantage: really, you want application puzzles to be a filter: Will teams enjoy your Game's puzzles?

    Because my brain latches onto impractical questions, I keep wondering: what if you weren't interested in the will-teams-have-fun filter and you did want to come up with a game that gave some kind of rank? Could you design a set of puzzles (or a puzzly game) that ranked teams? What if you couldn't grade on time?

    I came up with some ideas that wouldn't work in practice. E.g, "Your team must come up with a pangrammatic sentence, but if it shares a word with another team's entry, both teams lose." So far, I've failed to come up with an idea that would work.

  2. Everything is "graded on time", it's just a question of how much time (seconds, hours, weeks). Puzzle hunt puzzles are supposed to take hours at most. You can write puzzles that are hard even if you have a month — ARGs are full of them — but they are by definition a different genre of puzzle.

    Personally I wouldn't worry about collusion, or cheating on a "start when you're ready" clock, any more than most of us worry about cheating in-game. If you do want to run a bulletproof cheat-proof system, like you're giving away a million dollars, that's also necessarily a different type of puzzle presentation.

    So… what's your question? You want a month long puzzle that is hard (has discriminative value) but also resembles as closely as possible the type of puzzle and requires the types of skills used in a conventional puzzle hunt? And you don't trust teams not to collude?

  3. > "So… what's your question?"

    Suppose you relaxed the feels-like-a-hunt-puzzle constraint, but still went for something puzzle-y. But suppose you also wanted to be able to use the results to rank teams. What activity would you design?

  4. IIRC, the Amnesia Game (1999, Stanford) did something like this for their application; a few of the challenges were optimization problems, e.g., using given shapes to fill in a grid with certain parameters. Teams were scored on how successfully they optimized their solution to each problem. I don't recall if there were other, more subjective parts to the application.

    Jessen Yu might be able to tell you more; I believe he was on GC for Amnesia.

Leave a Reply to Curtis Chen Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.