Inside Petri Purho's Philosophy — creativity can't be scored, so hand the judging to the world

Intro — the man who attacked his own award-winner

Petri Purho is a game developer based in Helsinki, Finland. He won the 2008 IGF Seumas McNally Grand Prize with Crayon Physics Deluxe, a puzzle game in which crayon lines you draw become physical objects, and he now works as one of three people at Nolla Games on Noita, a rogue-lite in which every pixel is simulated. Outside the English-speaking world he is probably better known for the latter.

My reason for writing about him now is simple. He set out to make a game that rewarded the player's creativity, and he failed. And then, the year after winning the Grand Prize, he stood on a GDC stage and listed those failures himself. The reporter present described the talk as surprising the audience "with a cutting critique of his game" (The A.V. Club, 2009 ↗). Designers always know their own game's faults best; not many perform that fact quite so dutifully.

This is a study of the person, not a write-up of the games. I quote only sources whose text I read, and I keep his own words, the reporter's framing, and my own reading strictly apart.

Background — one game a month, as self-punishment

He entered the industry in 2005 as an intern at the Finnish studio Frozenbyte, working on Shadowgrounds and its sequel. In 2006, inspired by the Experimental Gameplay Project, he began making one game a month on his blog Kloonigames (Game Developer, 2019 ↗), spending no more than a week on each.

Crayon Physics was one of those — a five-day prototype. Asked what motivated it, he answered: "It's hard to say what motivated me to do it. Probably the fear of public humiliation for not coming up with a game on time" (Gamasutra, 2007 ↗). Of the full version he said the motive was the prototype's popularity, "And also by the fear of public humiliation."

After the 2008 Grand Prize he went full-time indie, and his explanation of why is very much in character: "I figured I'd work on my own games until I ran out of money and then I'd cut my hair and get a proper job. Luckily I haven't run out of money yet" (Game Developer, 2019 ↗). In 2016 he founded Nolla Games with Olli Harjola (The Swapper) and Arvi Teikari (Baba Is You). Noita grew out of experiments he had been poking at alone since 2011.

Philosophy — not the right solution, a creative one

Lay his statements out chronologically and one line barely moves. In 2007, as an IGF nominee, he explained: "While the game is a puzzle game, it's very different from other puzzle games in that sense that it's not about figuring out the right solution to the puzzle. It's more about coming up with a creative solution to the puzzle" (Gamasutra, 2007 ↗). In the same interview: "No two players play the game exactly alike."

Two years later, at GDC's Independent Games Summit, he said almost the same sentence: "The game is not about finding the right solution to the puzzle, it's about finding a creative one" (Gamasutra, 2009 ↗). He also explained the design's origin there: physics games like Armadillo Run are engineering-like with one or two solutions; at the other end sit sandboxes like Line Rider. He aimed to split the difference.

Twelve years on, with Noita, only the vocabulary changes. "Having a highly-simulated world feels much richer and much more interesting to explore. Not just explore in the spatial way, but to explore as a set of systems." He calls it an "emergence-generating machinery" and notes "there's always something new to be discovered, even for us" (Game Developer, 2019 ↗). The constant, as I read it, is that the designer refuses to close the solution space.

Obsessions — build to discard, and build the tool first

The first obsession is volume. On prototypes: "doing prototypes really worked out well for me," partly because you can "get the bad game ideas out of your system" (Gamasutra, 2009 ↗). He also used download counts as a crude signal of which experiment would break out: his prototypes averaged 2,000–3,000, the Crayon Physics prototype did around 25,000 in a month, and 250,000 once his YouTube demonstration spread.

The second is refusing to hold a finished picture in his head. "I've made a lot of prototypes where I had a good idea of what I wanted the game to be like, down to the graphics and sound effects, and what the game mechanics would be. Usually those games turn out to be the worst ones" (The A.V. Club, 2009 ↗). Which is why, he says, he can afford the stance of "This is not going to work at all, but I'll make it anyway, just to see if anything can be done with it later."

The third is tools. On Crayon Physics Deluxe he built the level editor before the levels (The A.V. Club, 2009 ↗). A good editor, he says, needs two things: to be easy to pick up, and to be powerful enough that small components combine into big ones. His good example is the Wii's Mii editor; his bad example is Photoshop — "You can do a huge amount of stuff, but it takes hours to learn, so such a small percentage of people actually end up using it." He also offered, as a joke, that you can tell a user-generated-content system is working by how many crude doodles players draw with it.

One detail of continuity is worth noting. Crayon Physics Deluxe in 2007 and Noita in 2019 use the same library for rigid bodies: Erin Catto's Box2D (Gamasutra, 2007 ↗ / Game Developer, 2019 ↗). For twelve years the skeleton of his worlds has been made of the same part.

Failures, and how he talks about them

Start small. Asked in 2007 what he would do differently if he could rewind: "Not much. I'd skip the part where I tried to code my own concave polygon collision algorithm. After much struggling I got it running, but in the end I didn't even use it. That was two weeks wasted" (Gamasutra, 2007 ↗).

The large failure sits in the concept itself. By the reporter's account, he had originally envisioned a game that somehow quantified creativity and rewarded right-brain thinking, and was disappointed when practical reality forced him to scale that ambition back (The A.V. Club, 2009 ↗). The question he threw at his GDC audience survives verbatim: how do you detect when players are being creative? He added that a lot of people are lazy and will go for the easiest solution possible. The upshot, he said, was that for many core players the game was somewhat too easy — and that his idea of the target market had not been clear (Gamasutra, 2009 ↗).

How he frames it matters. "I didn't want to talk about it at my talk like, 'This is how awesome my game is. These are the best things about my game.' I didn't want to do that marketing stuff, like these talks usually do. I wanted to point out, 'These are the problems I had with the game, and maybe you guys can figure out solutions'" (The A.V. Club, 2009 ↗). He does not keep his failures; he leaves them behind as open problems for the community.

The same habit runs through Noita. He volunteers the ideas that sounded great and died in practice. Buildings that were fully simulated and could collapse around a fight "sounded really cool in theory. In practice, you very often ended up with just a bunch of rubble at the bottom of the screen." Enemies whose limbs you could shoot off would be "fun to watch for a while, but do they really provide a formidable and interesting enemy for a game?" (Game Developer, 2019 ↗). The way through is always the same: implement, test, discard.

Dilemmas — playground or goals, and four feelings about clones

The sharpest dilemma is playground versus goals. Asked whether he considered removing the goal-oriented layer entirely: "It crossed my mind. That's how it started out. I made the level editor before I made the levels, and I played around with the game in sandbox mode a lot, and found that to be the most interesting thing about the game." His reason for keeping goals anyway is in the same breath: "But games have these great structures for motivating people to do something, and inspiring people to do something. Like, the number of hours people put into World Of Warcraft is insane. So I thought I could do that, but do it with something that's more useful for people, so they learn to be more creative" (The A.V. Club, 2009 ↗).

Which meshes badly with the previous section. He wanted to reward creativity; creativity cannot be detected; and once you reduce it to something rewardable, players take the cheapest path. He never solved this. He said so from a stage.

On clones, four feelings coexist. Anger first: after the first few appeared, "it took the fire away from doing the game a bit." Then acceptance: "I was kind of happy that there were so many of the clones because Crayon Physics got a brand name from it — people were referring to them as Crayon Physics clones." Then discomfort: an iPhone game called Touch Physics offered him money as a tribute, and he felt uneasy about accepting it (Gamasutra, 2009 ↗). Then regret: "I wanted to take more time to do the game, but because there were these clones coming out left and right, I had to push the game and be done with it. I felt I had to push it out a little early." Few developers admit all four about the same event.

The Noita-era dilemma is more technical and more blunt. "It's easy to think that having a complex or highly realistic simulation will auto-magically result in interesting gameplay. Often, that's not the case. The simulation can, in fact, be in the way of more interesting gameplay" (Game Developer, 2019 ↗). In the same interview he names the asymmetry between engineering and design: "Technical questions are much easier to answer. Is this running 60fps? Yes, then it's working." Against that: "Is this fun to play? Well, it depends who is playing."

One Kizuki-ish note. In 2007 he credited Braid with the idea that players should not be punished for failure. Noita, in 2019, has permadeath. It is too early to call that a contradiction: in the same interview he explains that he chose permadeath and procedural generation after Spelunky so that players would learn "not just through rote memorization, but in a way that allows them to apply that knowledge to a variety of situations" (Game Developer, 2019 ↗). The purpose of punishment moved from punishing to reshaping how learning happens — that, at least, is how it can be organised.

Influences — the ones he acknowledges

The biggest conscious inspiration for Crayon Physics, by his account, was Crockett Johnson's children's book Harold and the Purple Crayon — with an honest footnote: "Actually, when I worked on the game I hadn't read the book, but I knew the plot and the magic crayon" (Gamasutra, 2007 ↗). Two games follow: "The whole physics puzzle thing came from Armadillo Run. Players being allowed to create whatever they wanted and not being punished for failures came from Braid. Also the absence of Pavlovian reward system is something I picked up from Braid."

The crayon itself was an accident. He was making a title screen for another game about George W. Bush, Daydreaming In The Oval Office, reasoning that Bush "would definitely use crayons." Realising he could produce that look algorithmically, he connected it to the drawing-physics game he had wanted to make for years. He also did not want "a cheery kids game, where you'd have bright colors and cheerful music," but rather "the kind of way I really remember my childhood, not the cheery, fake stuff that tends to be put out there" (The A.V. Club, 2009 ↗).

He names non-game influences too: an obsession with John Cage's 4'33" ("as a piece of music, it's so interesting") and, with the caveat that he could not recall the artist's name, "these big blank canvases — white paintings." Those two produced 4 Minutes And 33 Seconds Of Uniqueness (The A.V. Club, 2009 ↗). In the same interview he praises Jason Rohrer's art games as work with "deeper meaning" that is neither funny nor about shooting people. His board games of choice: Settlers Of Catan and Puerto Rico.

For Noita he names two. Derek Yu's Spelunky, which "forced you, as a player, to learn how the world worked in a more systemic and deeper level than traditional linear games" (Game Developer, 2019 ↗). And the wand system, which "was inspired by the deck building games like Donald X. Vaccarino's Dominion" — modifying wands means moving spells between them "much like the cards in a deck building game" (80.lv, 2019 ↗).

Kizuki's reading — the designer who stepped down as judge

What follows is my reading, not his statement. I read Purho's twenty years as a slow abdication: he tried to score creativity, failed, and stepped down from the judge's chair by degrees. The Purho of Crayon Physics Deluxe wanted to be the referee who detects creativity and pays it out. He could not be. So in the next work he resigned. In Noita, whether a player's solution counts is not a matter of his taste; whether oil ignites, whether ice shatters onto an enemy, whether acid dissolves a wall, is settled by per-pixel rules. He is free of the duty to grade creative answers — which is why he can still be surprised by his own game. In 2019 he reported, plainly delighted, that he had killed an enemy above him, the body fell and broke an oil lantern, the oil spilled on him and set him ablaze, and that this had never happened to him before. Only a designer who has stopped grading gets to sit in his own board as a guest. That is how I read him.

Closing — where to start

If you want to understand him, start with Crayon Physics Deluxe and refuse to progress. Solve one level ten different ways. You will feel the point he conceded from the stage — that if you are playing just to reach the next level, you are missing it. Then play Noita for an hour. You will probably die. If you can accept that you died because of your own judgement, you have received what he received from Spelunky.

Some onward paths. His Noita collaborator Arvi Teikari ↗ designs by letting you play with the rules themselves. Jason Rohrer ↗, whom Purho singled out for praise, plays with life and death through constraint. Line the three up and you can see late-2000s indie developers in Finland and North America working the same question — what are games for — from three different directions.

Sources

Primary sources used in this article (text verified for all):

・Gamasutra, "Road To The IGF: Crayon Physics Deluxe Sketches Self-Expression," 8 November 2007 (interview) ↗

・Gamasutra, "GDC: Crayon Physics Creator Purho Prototypes Hard," 23 March 2009 (report on his GDC Independent Games Summit 2009 talk) ↗

・The A.V. Club, "AVC at GDC '09: An Interview with Crayon Physicist Petri Purho," 28 March 2009 (interview) ↗

・Game Developer (formerly Gamasutra), "Road to the IGF: Nolla Games' Noita," 13 March 2019 (interview) ↗

・80.lv, "Noita: a Game Based on Falling Sand Simulation," 5 April 2019 (interview) ↗