In this talk, Ian Tullis talks about puzzle design; in general and in the puzzlehunt style. He talks about what makes puzzles interesting in general; and the weird areas that the puzzlehunt community's been exploring for the past few years. If you read my 2-Tone Game write-up, you might remember where I said:
I had a halfway-brilliant idea: since this game wouldn't be a one-time event, since it would be this persistent thing, it would be a way to introduce new folks to puzzlehunts... I don't think I followed through well on this idea.
This talk gets into some ways that I could have followed up: remembering to include "fun" along with the "wow". Including what the Anonymice call a "crank" for the player to turn while they wait for inspiration to strike. Anyhow, some notes:
- We're talking about puzzles. Ian likes puzzles.
- Puzzles are little worlds in which everything's there for a reason.
- In real world, defining the problem's impossible, let alone solving it.
- You look at a puzzle, and when you understand it, you know you understand it and you know that your team needs to get back in the van and drive to "that Chevy's by the Bay Bridge."
- You get to exercise your <napoleondynamite>skills</napoleondynamite>. The German word for joy in exercising your skillz: Funktionslust
- Huntish puzzles are different: there's an answer to extract. You aren't done when you fill in that crossword. You're done when you extract the answer phrase out of that crossword. Whether you filled in the whole grid or not.
- There's a goal-oriented nature.
- [note-jotter's aside: Did the necessity of answer extraction nudge us towards multi-layered puzzles? After all, so many puzzle-huntish puzzles have a pair of steps: "fill in the grid" / "extract an answer phrase". If there's an answer-extraction step that's itself puzzly, then you've drifted into multi-layer territory.]
- There's a goal-oriented nature.
- Thomas Kuhn, the "Paradigm Shift" coiner uses "puzzle" to talk about useless scientific "problems". The stuff a scientist might work on because it's an obvious place to apply already-existing scientific knowledge; not shifting paradigms.
- And you can say that he's not using the same definition of "puzzle" that we are, but he does point out "the assured existence of a solution".
- Kuhn worries that these puzzle-y problems might insulate a scientific community from problems that can't be reduced to a puzzle-y form.
- [note-jotter's aside: Admittedly, I have made zero progress on eradicating Malaria while I worked on puzzlehunts]
- Ian says: "Our puzzles are different, dang it!"
- In our puzzles, you have to figure out the rules. Yar talks about "the pleasure of being stumped."
- Our puzzles point out connections between different things.
- E.g., Ian made a puzzle which gave you Latin names for critters ...
- which were of the form earth tiger or fire snake...
- which leads to Chinese sexagenary cycle (zodiac+five elements)...
- This leads to some design principles by which you can make puzzles that Ian will like:
- Principle Microcosms in which everything exists for a reason
Achieved by Everything points to the answer; no extraneous data streams
(With a shout out to Scott Blomquist for "data streams")
- Principle Unexpected connections between familiar things
Achieved by Novel subject matter or mechanisms; a “twist”
(This is to delight people. Not just to confuse the n00bs. Right? Right??)
- Principle Have clear answers (which may be part of a meta)
Achieved by Answers fit into metas, or at least fit their puzzles thematically
- Principle The pleasure of being stumped / figuring out rules
Achieved by Intriguing, but a “battle of wits that the writer expects to lose”
- Principle They put the Fun in Funktionslust!
Achieved by The solving process must be fun; should be a crank to turn
- Principle Microcosms in which everything exists for a reason
- Let's distill this into two metrics: "Wow" and "Fun"
- Wow: novelty, elegance, cool constraints, cool connections. Puzzle snobs write about them. They have a gimmick you can talk about, especially if you're a puzzle snob writing in your LJ.
- Fun: enjoyable process, humor, team-friendliness
- A Wow example: NYTimes crossword puzzle Dec 23, 2008. The crossword's answer is symmetric: e.g., if in one area you see EDISON, the radially-symmetric grid area will be NOSIDE.
- An amazing feat of construction!!
- But maybe it's not much fun to solve. There are some lame words in there.
- A Fun example: that Jumble acronym puzzle that's in your daily newspaper. They give you AGLND and you figure out GLAND.
- It's totally clear what to do. You've done a million of these already.
- But at the end, the most you get is a chuckle.
- So where are we now? We are in a golden age of puzzling. The field's wide open. Teh internetz have made the world's knowledge universally accessible and useful; this in turn makes people interested in concepts and cool connections—which our style of puzzle can provide.
- But bad news about the future:
- We might run out of forms. Our constraint is: somebody looks at this thingy and figures out how to get a phrase out of it.
- We're becoming less accessible.
- Let's talk more about the bad news. Hey, maybe we'll even figure out some workarounds.
- Our forms are constrained there are only so many ways to extract an answer
- Almost all puzzles use a mechanism from the short list on the slide.
- There are so many where you're getting stuff letter-by-letter such that a valuable team skill is looking at V__L__N_O_R_F__N__I_ and yelling "Vallaincourt Fountain! Get in the van now now now!"
- IWBNI more often it was implied via a connection, sequence, or omission
- Conventions "Old Standbys". Make puzzles less accessible to n00bs.
- Experienced solver is bored: aa aaa iai iii i Hey, it looks like Morse. Ho hum, I'm bored.
- N00b solver is flummoxed: aa aaa iai iii i ?!? Wha- no explanation. I am totally confused?!?
- Also, can be misleading. Example of a puzzle that looked semaphore-ish but wasn't semaphore. Semaphore is so common that if something looks too 45-degree-angle-ish, it could be a red herring.
- What's an answer extraction that n00bs can understand?
- _ _ _(_)_ _ Probably OK
- _ _ _ _ _ _ (4) Probably not OK
- neTWOrk (=2 =network(2) =E) "Oh I would never get that" (which means "I've lost interest")
- [note-jotter's aside: So when I talked about "too hard" puzzles, in hindsight I realize I was plagiarizing this talk. Though I didn't realize it. I guess I internalized this talk :-)]
- network 23 (= network(2) network(3) =ET) "Oh I would never get that" (example from Scott's recent BANG puzzle (?))
- In this list, an experienced solver will get the first two, but won't be especially impressed. Will get the third and be impressed. Probably won't get network23. Why not? It's not convention, though it makes about as much sense as _____(2).
- Hey, maybe ____(3) is a crappy convention. N00bs don't get it and l33ts aren't impressed. Favor "Jumble-style" over indexing.
- constraints inspire great art
- harder to write
- can result in a less-fun puzzle.
- [note-jotter's aside: Yeah. I think this is why "remote solvers" don't enjoy the "remote" version of the 2-tone Game. A puzzle that meets the constraint of using data from Coit Tower is impressive if you're sitting in Coit Tower thinking "whoa, they made a puzzle from this?". But if you're at home, looking at some photo of a Coit Tower mural, you think, wow this puzzle sure had to "stretch" to get some of these words. Wow, that's an awkward reach.]
- n00bs don't know enough to notice
- As we reach further and further for novelty, we leave the n00bs further behind. Are we painting ourselves into a corner? "Our puzzles are so cool, not like some word search" Hmm
- Ian's seen before:
Diverse! then Stagnant. In old
poems from Japan
- How can we do great stuff w/out locking ourselves in our own ivory tower?
- Keep doing what you love
- As you strive for novelty, maybe don't do it by combining existing conventions. The n00bs might recognize Morse, might recognize semaphore. But do you really think they'll spot the combination of them?
- "Where's the fun?" Don't forget the fun.
- Justin Graham asks a question into his hand and the mic doesn't pick up what he's saying. I'm nodding next to him and stroking my chin contemplatively, so it must have been pretty profound. Maybe we can reverse-solve the question from Ian's answer:
- Sometimes you get lucky and the constraints help point to an answer.
- But often you're so constrained that you don't have that choice.
- Eventually, you mature enough as a puzzle designer that you learn to let some ideas go.
- John Owens asks a question into the back of Deeann's head. We can't hear him, but he's making nicely expressive hand gestures.
- Yeah, there's the danger of an "arms race" between constructor and solver.
- Was looking at puzzles from an old puzzlehunt. By modern standards, those puzzles seem easy. But they're precise, clean, simple.
- Cultures go through this. In China, they look back to idealized ages when rulers were perfectly just.
- More recent puzzles can't be tackled by one person in an hour. You need a whole group.
- Who's she (24:48)? She must be saying something interesting; Sean Gugler is scrunching up his brows and thinking about it. Ian sez:
- We're lucky to be in the bay area where this scene has been happening.
- Thanks to Debbie et al, rest of USA has DASH
- Boston has the Mystery Hunt, but doesn't have something BANG-like. [self-destroying prophecy as BAPHL has since come along and is going strong as I write this in late 2010, a few months later]
- Brent Holman points out: maybe we shouldn't worry so much about the n00bs. There's this thing called teh internetz. There's dorks making spoilery game write-ups that give n00bs a baseline of knowledge. Ian sez:
- Yeah, I worried about using a puzzle concept that had one element similar to something from 1999 mystery hunt. A puzzle snob might have turned up his nose at that.
- Corey Anderson: view point of playing versus designing–
- I've forgotten what "playing" is like.
- You'd think you could give them a "codebook" and they could come to the game knowing about analog-clock-to-semaphore-to-letter code.
- For the Shinteki Disneyland hunt, we gave teams codebooks.
And there are Disney fans who do puzzly scavenger hunt games!
But they don't use our particular conventions.
And for every puzzle that relied on "aha, it's Braille" or whatever, those people hit a wall.