Optimal decisions: 100 Rogues architect Keith Burgun on game design03 Jun 2013 0
Keith Burgun is the game designer behind 100 Rogues, a widely admired roguelike game from 2010 that I think is one of the best examples of the genre on iOS. We talked a couple of months back about Empire, Burgun's upcoming minimalist riff on Civilization-style 4X games -- he's simultaneously working on AURO, a tactical game with a superficial resemblance to last year's divisive Outwitters, though unlike Outwitters it's a single-player game. Mostly. We'll get to that.
I wanted to talk to Burgun about these upcoming games, but once we got on Skype we spent a lot of time talking about game design in general. He has some decidedly idiosyncratic views on the subject -- if Burgun were an economist, I suspect he'd be one of those that believed that people are always rational actors and make optimal decisions with their money. Having spent a few years living in Southern California, I saw too many racing exhausts on Hyundai Elantras for that sort of thinking to jibe with me.
But agreeing or disagreeing is beside the point -- there's no question that Bergun is a thoughtful, intelligent game designer with a lot of interesting things to say. Our conversation ran for so long that even after making some very unsentimental edits I still ended up with a monster piece, so I've cut it in twain. Tomorrow, we'll hear all about Empire and AURO, two iOS games that we can expect to see this year (AURO later this summer, Empire hopefully this fall) -- but for today, Keith Burgun on game design. Why the designer of 100 Rogues doesn't play roguelikes anymore, his disdain for the term "permadeath", and why he thinks that naming your XCOM squaddies after people you know doesn't matter one whit.
Owen Faraday: I know you've got a background in game design and you've even written a book on the subject. How did you start down that road?
Keith Burgun: I've been designing games with pen and paper since I was a kid. When I was like 14 I got a computer and I started programming games in QBasic.
The first time that I really said, 'OK this is going to be my career now' was in 2007 when the iPhone came out. I was working with a guy making free games and he got a job with a web development company called Fusion Reactions and they wanted to make an iPhone game. He brought me
in as the game designer and an artist named Blake [Reynolds] who does the art for our games and that's how we got that gig doing 100 Rogues.
I had been writing my own blog about game design around that time. I was reading everything I could find about roguelikes and I found a writer named John Harris who is like, the guru of rogue likes. He wrote for GameSetWatch, he was basically a walking dictionary of roguelikes. He got me a gig writing for Gamasutra and after a couple of years I decided I wanted to write a book about game design.
I studied music in college. One of the things that attracted me to game design was the idea that I could help more. Music has so much history, so much established knowledge and theory of how music works. In game design we're still just shooting in the dark. The whole discipline of being a specialist game designer is less than a hundred years old. That's very new.
OF: So Roguelikes are part of your origin story then. And you're making 100 Rogues in the middle of this roguelike renaissance, with games like FTL and Desktop Dungeons making such a big splash.
KB: I only learned about roguelikes at all in 2007. I'm not a big roguelike guy, I don't even really play them anymore. But John Harris is still really an important part of what I do. He had a game Kickstarted a couple of years ago that I'm still working with him on, a cave exploration game called In Profundis. I'm helping him with music and art for that one.
But rogue likes get one or two basic prerequisites of good game design right. Those would be random generation for single-player, and what people call "permadeath". I don't like that term, I just prefer to say that you can lose. There's a win and loss condition. Those two properties are essential for a single player game. Tetris has those, for example.
Choices really need to matter, they need to create tension. Having a fail state in a game gives you real consequences and avoids a positive feedback loop where the player is perpetually getting more powerful. Games should get harder as they go because the player is ostensibly getting better at it. In a game where the character is also getting better. In a lot of RPGs, the game gets easier because the player levels and the player gets better. The original X-Com even, is a great example of this, because the game gets easier as you go.
OF: Once you get mind control it's a walkover.
KB: Exactly. Mind control, flying suits, OF: But you can also make the argument that making the player feel like a badass is a reward for persisting to that point. There's a certain pleasure from wiping the floor with these aliens that were mercilessly kicking your ass when you started playing the game a couple of weeks before. It's the conclusion of your Hero's Journey.
KB: I agree mostly. I usually use the term 'fantasy simulation' for that. I think that's a different value, that's about wish fulfillment. That's different from what a strategy game should provide. Those two values conflict with each other. Historically, game developers haven't drawn that line very clearly, and there's a lot of bleed-through between real strategy games and fantasy simulations.
OF: So how do you create a feeling of progression if you're consistently ramping up the difficulty to meet a player's increasing skill? How do you avoid the feeling that you're on a treadmill and not getting anywhere.
KB: Think of it like playing a sport. Games are a discipline, like playing a sport or the piano. The progression is the progression of your skill. Your chess pieces don't level up, but your chess 'army' gets stronger because you're learning to position it better. I reject the notion that games need an external progression because it's already there in the skill that you're gaining.
OF: To me that reveals some insight into why you're making AURO [a tactical combat game] and Empire [a Civilization-style 4X game]. With Empire in particular, there's a real divide between the people who play Civilization at increasingly higher difficulties to get better scores and faster victories and there's people who play Civ for the experience of role-playing the immortal ruler of a nation-state.
KB: That's true, but Civilization has those elements I look for: random generation and a loss condition. That's why it's such an evergreen single-player game.
OF: If those are the elements you value so highly, why don't you play roguelikes anymore? They're the epitome of those two game design values, it seems to me.
KB: There's a lot of things about roguelikes that repel me. There's things that are central to many roguelikes that are just bad game design. Hidden information, for example. In the roguelike community there's things they term 'spoilers' that are really just basic mechanics. Like 'how does this potion work'? That's pretty silly. And there's wasted mechanisms like exploration. In rogue likes you often have an 'auto explore' button because it's such a no-brainer decision, why is it even in the game?
OF: So in your opinion, what are the classic game designs that embody your key principles?
KB: It's difficult if we stick to just video games -- I'd feel more comfortable answering that question if we can also talk about board games.
OF: Yeah, certainly.
KB: Specifically European-designed board games are awesome examples of what I'm talking about. And they're often not single-player, but to me it doesn't even matter. I don't even draw the line between single-player and multiplayer. If we're talking about strategy games, there's almost no such thing as a single-player game.
Stuff like Reiner Knizia's games like Through The Desert. Puerto Rico. A lot of these European games are very efficient and well-designed systems that are about making interesting decisions.
OF: The knock on Knizia games in particular and Euro games more generally is that you're essentially playing a spreadsheet. The themes are sort of pasted on and couple be interchangeable with any other theme. 100 Rogues didn't feel that way to me.
KB: I think theme primarily is there to communicate game mechanisms to the player. A theme gives you free information: if you give a player a sword, he immediately expects that he can swing it. If you give him a hammer, he thinks 'okay, I can nail something down, or break something maybe.'
OF: Theme as an implicit tutorial.
KB: Right. The other aspect is the 'fantasy simulation' side of it. 100 Rogues definitely has some of that -- we use some of those vestigial Dungeons & Dragons tropes like strength and to-hit chance and little things that aren't necessarily the most efficient and elegant things to have in a strategy game but they come from D&D which is really a simulation, generally.
OF: And it's a common frame of reference. D&D concepts are a lingua franca.
KB: Yeah, for people who are in that.. milieu. [laughs]
The other aspect of theme is the commercial side. Making things that look attractive and sexy and that people want to buy. Euro games I agree are bad on the fantasy simulation and bad on the sexy, but I think they're pretty good at the aspect of theme that communicates how mechanisms work.
While you're making a really difficult decision, and when you're really immersed in the game and thinking hard, you don't really think about theme. In that moment you're not thinking, 'I'm the hero and I must save the princess' -- you're thinking like a spreadsheet. That's my point. Your brain doesn't have the resources to think about thematic details while you're trying to make a difficult decision. You brain melts that stuff away when you're immersed in that decision.
That said, a lot of video games aren't very hard and in those we are revelling in the theme a little more, and it can be more of a fantasy simulation, theme-driven game. It's not a hardcore strategy game like Go.
OF: My immediate response to that is that some of those games that present a good fantasy simulation, as you call it -- they do take you out of thinking like a spreadsheet. When you've got one move left in XCOM and you're backed into a corner, and the guy that you've named after your uncle or your best friend is almost certainly going to die when you hit 'end turn' -- there's an emotional element to that. That's when a game becomes a story-telling medium.
Of course you're trying to find the optimal decision but you want that optimal decision because you're invested in the unit. Even in a Gary Grigsby wargame I can be worried about the fate of an infantry battalion represented by a NATO symbol because they've been with me since the start of the campaign.
KB: I'm not convinced that's the case in the situation you describe there. In XCOM, those characters that you've named and had around for a while also tend to be your best characters, they're tactically the most valuable to you. Now, if all the characters were exactly the same and you found yourself making a sub-optimal decision because you wanted to protect the guy named after your uncle, then that would be good evidence for that case. But I don't think someone would do that.
OF: So you think that I'm telling myself I'm making an emotional decision, but in fact I'm making the same decision I would have made anyway if they my characters were faceless drones with identical stats.
KB: I think that's probably the case. We're assuming a few different things here: I have to assume that you're playing on a difficulty that you find difficult and you're not just steamrolling the game. If you're playing on a setting that you find difficult then you really have to believe that you're making optimal choices.
OF: Because I'm sabotaging myself if I don't.
KB: Right. If you're playing XCOM on ironman, you can't afford to lose one of your best guys. That's why you save your uncle -- you're making the smart call, not the emotional one.
Part two of this interview, with more details about AURO and Empire tomorrow.