In a crowded sea of games this year, developer Ben Esposito launched Donut County, a quirky, personal game filled with hand-crafted puzzles where players control a hole in the ground that sucks up a fictional Southern California county.
Esposito spent six years of his life making this wonderful game, and he learned a lot along the way. We sat down with Esposito on the Gamasutra Twitch channel and picked his brain -- here's what he had to say.
Edited for length and clarity.
I knew I wanted the game to be about erasure because it's a game where you play as a hole in the ground, and you erase all the things that are in a place.
Because I knew early on -- you play as a hole in the ground but the hole's not a character, the hole is just a concept right? So the real character of the game is the stuff you put into it, so the stuff has to be really important. Then literally what's happening in every single level is you're erasing someone's home, essentially.
"I can tell this story from the perspective of an idiot tech gentrifier person."
I wanted to keep the game pretty human scale because I think your connection to those objects and the way you expect them to behave... I knew it was a game about erasure, but i didn't know what the real story was.
So coming back around to "ok, well, how do I do a story about erasure that I can build off of my personal experience?" I know my personal experience is moving to LA to work on video games and seeing neighborhoods change as a result of a bunch of tech workers moving into a specific area. So an area like Venice Beach, which was like totally taken over by Facebook, Google, and etc., boxing the existing residents out and raising the rent and, you know, stuff like that.
I was kind of aware I was part of this process, and seeing this neighborhood change I was like, "Oh cool, I love LA and this is something that is just happening around me, and it's something that I'm kind of complicit in. I can tell this story from the perspective of an idiot tech gentrifier person." And that's where the raccoon BK came from; he's like a cartoon version of my character type, so to speak.
I ended up with more adventure game-style puzzles than strictly systemic puzzles.
There is a version of this game where it's a puzzle game and it's got way more systemic rules and you treat it more like a Sokoban or something where it's like, "Okay, I have these types of objects and I have to get this over here, and these are my constraints around it."
But I think in terms of getting you to be immersed in the logic, the silly logic of Donut County, the puzzles need to have their own consistency that you learn over time. The way the game is silly and the way the game has really stupid rules, I thought that was the most fun way to explore the puzzles -- to introduce you to new concepts that told you a little bit about the feel of the world.
The style was informed by the type of illustration stuff that I'm familiar with and in terms of the silhouettes and broad shapes, informed all the characters and things like that.
But it's also efficient, because I knew that if I was going to model almost everything in the entire game, I needed a process that scales. Although stuff is modeled in Maya, because that's what I was familiar with, people are doing really similar work in Blender, and you could do it in anything.
My rule was no textures, so everything has to be built with flat colors. The meshes are cut up, just so. So if you see the rocks with those dark red streaks along them, that's actually just cut into the mesh, that's not a texture. And that was my approach to everything: if I want that vector quality of this game to scale to any resolution, I have to build it into every single mesh, and so I had a rule of like no more than four colors per object, unless it's like a really large object.
So I built this in Unity, and if you're familiar with Unity, i mean a lot of game engines work with the same type of material system, but the way i did this was I modeled most of them in black and white just using four different materials. I would kind of have a light one, a dark one, a middle one, and an accent. Then I would bring them into the engine, and since they were made with separate materials which are each their own asset, I would color the scenes in Unity. I'd have the thing open and then I'd just be tweaking the individual colors one by one until I had the right feel for the scene. So that was my process.
Unfortunately, each material is a draw call so it's not that efficient. I could have gone in and like baked them all into textures or vertex colors after the fact, but the shaders are simple, it's just showing a color, so I think it was fine. It could be optimized further, but I just liked the convenience of being able to tune in the actual engine.
The Unfinished Swan [on which Esposito was a level designer] is a short game that is very experimental and has novel mechanics that we tried to use in interesting ways, up until the point where we felt like we'd explored them and then we move on to the next concept.
So I think my process for working on this game [Donut County] was probably informed by the type of work we were doing on that game because I was coming from a place of "ok, I want to explore a novel mechanic and I want to tell a story." That puts a lot of constraints on what you can do. That's usually the reason why you end up with something that's like pretty short and tight.
A donut without a hole is called a NUT.
Also, The Unfinished Swan was a really interesting game we tried to make really accessible to non-gamers, but there was a limit for how much we could do that because it was a first-person control scheme...We were thinking we were making something very, very accessible to gamers that maybe don't even play FPSes, but it still had a limit to how accessible it really could be because a lot of people aren't able to both move and look independently on each stick, because it's a lot to ask for.
So the controls [for Donut County] were kind of a reaction to that and [the question], "ok how can I take it a step further and build a game where it doesn't rely on any gaming knowledge in order to get started playing?" So that's why there's only a tap or a click or one-button and a movement, and that's it -- and you almost never need the tap anyways except to advance the dialogue.
So I built it with that constraint because I wanted to push the accessibility even further -- so my grandpa can play the game! He's 91 and he was like, "wow I was really impressed that I got 30 minutes into it." You know his approval isn't always easy to get, so that was a big deal.
Probably the biggest reason I went with a publisher [Annapurna Interactive], the most important reason to me, was that I knew that with my limited amount of time and energy remaining to work on the game, I wouldn't be able to make a splash in the way that I think the game needed to succeed.
I could have put it on Steam by myself, and then hope that it goes viral and people tweet about it and stuff. I had a feeling that people would dig it, but I didn't think that that [earlier] version of the game -- [the one] that was a bit smaller and that was only on one platform and that was not marketed to any significant extent -- I think the game had more potential than that.
"If me from the past showed me this game I'd be like, 'dude, you don't even know how much time this is going to take for you to finish. You think you have it figured out but you haven't even started.'"
And so working with a publisher was really, really nice for that reason. They brought it to a bunch of conventions, they also helped negotiate deals with console, they had a relationship with the App Store. So those kinds of soft connections and them taking care of that stuff for me helped really expand the impact of the game.
I think that was really, really worth it, and also just having someone else to put some external pressure on the process in terms of like, "hey, why don't we launch on this date, stop noodling on things." That stuff was extremely, extremely valuable. Otherwise, I wouldn't have have known when it was ok to stop working on the game. It helped for a lot of reasons like that.
So much time passes when you make a game. In my case, the 2013 me who started working on this game full-time is a different person than I am now. I don't think that I'm going to stop growing and changing, so this was a really good idea for me to make when I was however old I was in 2013. It was exactly what I should've been doing then, and it was exactly what I wanted to do then, and it's different now.
I can look back on it and say "I can't recommend anyone else do this," but I wouldn't have done it any differently -- I had to learn that stuff for myself. The reason I just can't suggest it is because it might not have turned out good. There's a chance that things might not have worked out the way they did. I don't even think I would give myself the advice to keep going.
If me from the past showed me this game I'd be like, "dude, you don't even know how much time this is going to take for you to finish. You think you have it figured out but you haven't even started." So I think [the game's development] is a representation of me in the sense of this is me growing up, this is me over six years working on a project, improving my skills, but also changing my sensibilities and my approach to design.
You can kind of see the game changes near the end. Some of that is a reaction to the work I was doing at the beginning and the way the industry has been changing, stuff like that. I feel like it's me but it's not like I wrote it all down on paper and I stuck to that and I adhered to that. It's more that this is what I was able to produce and it represents me at all those different points of my life.
For more game developer interviews, be sure to follow the Gamasutra Twitch channel.