It's "The Famine Game Postmortem", a talk by Todd Etter and Phil Dasler, the Lead Gamemakers of the Famine Game This here is my notes. [My rambling asides are in italics] and I take some pretty egregious summarize|rephrase|totally-change-meaning liberties with other folks' words, too. Original videos and slides at this here link.
- Backstory Todd and Phil flew out to West Coast to play in
Ghost Patrol
game back in 2008 (see
Todd's writeup).
- Emerged thinking "oh it would be awesome to run something like this"
- mind you, Todd and Phil hadn't met yet. Todd flew out from DC area, Phil from Texas; played on different teams.
- Debbie introduced them to each other.
- Our prezo looks better than most because Nat Parisi designed 'em.
- Basic idea: theme a The Game on The Hunger Games
- 24 teams, each assigned to a Hunger Games district (lumber, electronics, fishing, etc).
- Teams "kill" each other by solving puzzles
- Teams start with no solving gear, they earn it
- How to make the theme fit while keeping the game fun? Killing is mean.
Do we really want to eliminate teams? [no]
- Instead, themed the puzzles on Hunger Games districts, and in the story of the game, your team was eliminating other teams. But in practice, you weren't really eliminating other teams; they weren't eliminating your team.
- Solving a puzzle gave you a message that you'd "killed" one or more teams associated with that puzzle.
- But if your team was associated with that puzzle, for that puzzle you'd instead get the message that your team had "survived" due to completing that puzzle.
- Inobvious detail: the last puzzle had to "kill" more than one team: if it only "killed" one team, when that team solved it, they'd get the message "you survived this... and thus won the game". But if they were already the only surviving team, that wouldn't make sense.
- We like metapuzzles.
- Hunt had three parts, like three books: Arena: 23 "kills" 3 metapuzzles; Clock (12 puzzles and a meta); Revolution: head into DC proper for puzzles and metas to "take over the capital"
- Wanted to have many many shorter puzzles instead of smaller number of "grindy" puzzles. 75 puzzles, most of which were in the 20-40 minute range. A few longer ones.
- "Clock" part was a set of 12 puzzle rooms; teams could go into a room to
try to solve for 15 minutes. But if they didn't finish in time, they had
to go on to another room.
- If all 24 teams had showed up at the clock at the same time, would have been too crowded.
- So, rather than working to reduce the spread in the first part of the game, kinda let it slide.
- But it'd be good to manage the spread after the clock.
- In the clock: First few teams at the Clock didn't get to see its metapuzzle until they'd solved all 12 mini-puzzles. But later teams got to see the meta after solving fewer minis.
- In Revolution, the last part: To get all the teams back together again, of the 27 puzzles in this section, a relatively slow team might get skipped over 23 of them.
- Set up a big spreadsheet to simulate the experience for fast, slow teams.
- Wanted to come up with automated closing times for puzzles.
- So sketch out a set of times, simulate it—see who would get skipped over what.
- if teams would be skipped past part of the "core experience", you know you need to re-jigger the timing.
- Wrote an app to check answers, dispense hints, dispense bits of story
- oh, nicely designed
- needed an alternate version of content for each team-elimination puzzle: what to show to that team so it became their "survival" puzzle.
- for story content, needed
videos
of teams being eliminated
- asked each team to make a video acting out some scene.
- then applied FX to change the meaning, lest that team get an "edge" solving their own puzzle based on their video direction.
- Hint system very much inspired by Shinteki—understand that few teams will ask for hints. But they will take time-release hints.
- Worked on game from March 2012–October 2013.
- Really glad had an earlier-than-theoretically-necessary playtest. Forced folks to be ready on time. Gave more time to fix things.
- Using an existing story is great; you don't need to do so much storytelling
with the game itself.
- The Hunger Games districts are already themed: the lumber district, the electronics district… That's pretty nice for inspiring puzzle ideas.
- Trying to figure out which ideas would be fun and which would backfire. Taking away teams' gaming equipment at beginning: fun or cruel? (did that) Having them ride public transit into DC: fun or annoying? (didn't do that)
- (Part 2 video)
- More goals:
- puzzles should fit locations
- which worked out really well for the Pet Resort setting of the Muttations puzzle—the Pet Resort manager was into it. "We have a dog costume." (Tell folks about the event, they might pleasantly surprised me)
- It's OK if not all puzzles are super-memorable. Especially since you're going to have to skip some teams over some puzzles. You don't want them to feel like they missed part of the core experience.
- Still, don't make lead teams suffer too much. It'd be bad if the story
got out "We were way out in front, so we were sent to Langley to
solve a puzzle called
Kryptos…"
- Team Golden Smokingjay did break the game, got way out in front of the pack. Got to one site a few minutes before it opened, had to wait. But maybe that's not a disaster.
- Working around "The Stupid Hours"
- Things that are part of the core experience: put those during the day, at times when teams will be high-energy.
- Late at night: shorter, simpler puzzles. Physical activities.
- Late at night: this was the Clock arena; comfy facilities inside (a school). Breaking up the puzzles into 15-minute bits probably also kept teams from slumping into sleepyheadedness.
- If you need many many little puzzles, you want Ian Tullis on game control because he has more puzzle ideas than is theoretically possible.
- Getting teams to mingle
- encouraging same-district teams to collaborate ahead of time
- Friday night party-ish events
- getting team logos, bios, etc into the answer-checker app so that teams could look through it easily
- DC has some cool locations that don't necessarily fit the game theme; but some were cool enough to use anyhow.
- "Polarizing" puzzle types: some folks will complain about any of these.
[as I look at their list, I realize I'm pretty complain-y]
- cryptic crosswords [yay]
- Smith Jones Robinson [boo]
- Nikoli grid puzzles [boo]
- string of abstract encoded data [hmm]
- Music-based puzzles [boo]
- In hindsight, we could have…
- …done better at herding volunteers
- could have communicated better. could have done things to make them comfier.
- …warned teams: "It's not just the application. If you get in, we're going to ask you for a kill video, a team logo…"
- Didn't have app data or puzzle docs until a couple of days before the event, and then only because of Phil putting in herculean effort.
- Q&A
- (apparently, question didn't make it onto video, d'oh)
- video actors were kinda improv'ing based on the story
- Bob Schaffer asks: I had a great time at this game.
Picking teams to play from many many applications
is hard. What did you do?
- We stuck to our policy. But we agonized over it.
- And frankly, we weren't 100% pleased with how it went.
- But it was a relief when players from not-accepted teams got "adopted" by other teams.
- Allen Cohn asks: Did this game have a scoring system?
- Yep, based on the Hunger Games theme, we presented that score as number of fans.
- But didn't emphasize it during the game. Players who cared about scores could look. But many players don't care. And newer players could have found it discouraging.
- Corey Anderson asks: In your game development timeline,
there were several months between first meeting and first puzzle
ready to playtest. That's a long time. What all happened in there?
- Making sure that we actually want to do this thing.
- Game structure: if you write a bunch of puzzles before you figure out game structure, you're just gonna have to rewrite them. So figure out structure early.
- Figured out scheduling: when do we want the hunt to occur? To make that work, when do other things need to be ready?
- Curtis Chen asks: Team Snout had a great time. Two questions: how
often did GC meet? When are y'all running your next game?
- At first, a little more than once a month. Then fortnightly. Then weekly. Leading up to the game: might meet daily or more often.
- GC was geographically scattered, especially when Natalie moved to Seattle. Yay for Google Hangouts.
- Corby Anderson asks: What was up with the next of kin puzzle?
- We had team members' loved ones record a puzzle for them. The Hunger Games had Jabberjays, so this was a thematic way to deliver a puzzle.
- In the applcation, we asked for "next of kin"
- Debbie did a lot of hard work getting recordings from folks
- But most puzzlers "in puzzle mode"
didn't recognize the voices of their loved ones. Hmm.
- In Team Bloodshot, these families are friends and hang out together: but still didn't recognize voices.
- Wei-Hwa points out: besides, it's not like my sister and I usually talk with each other in English
