JOIN BREITBART. Takes 2 seconds.

‘RimWorld’ Developer Tynan Sylvester: ‘Don’t Listen to What People Say. Watch What They Do’ to Find Developers to Trust


In the last of this three-part interview with Tynan Sylvester of RimWorld fame, we explore what inspired the game, the way it’s built, and how Tynan confronts the challenges of Early Access and ongoing development.

Nate: Right. Now we’ll dig into RimWorld. Let’s start with influences. Of course, Dwarf Fortress is an obvious one. Then aside from the character interaction influences from The Sims, I would imagine maybe simulations like Theme Hospital also contributed?

Tynan: Well, stepping outside of — you covered the games pretty well. The biggest influence is probably Dwarf Fortress. And then there is sort of a genre of games that are similar to that, like The Sims, Prison Architect. Like you mentioned, Theme Hospital is more of a distant cousin. There’s basically a genre of games where you look down on people, and you kind of can get to know individuals to some degree. I think that would be the primary influence.

But stepping outside of the games themselves, for Dwarf Fortress there’s this whole subculture of people who write stories about what happened in the game. It’s been around for years. So there’s these giant long stories where people will pass around their saved games and each one play a year or something in the game. And then each one will write or illustrate things that happened.

Nate: Like Boatmurdered.

Tynan: Yeah, Boatmurdered, Gemclawed. So, before I made RimWorld I read these stories, and I would say that that is more of the inspiration than the game itself. I played Dwarf Fortress for like eight or ten hours, but I read these stories for many more hours than that. And I really just wanted to make a game that could create a story like this, because it’s fascinating how people could take something so abstract, and so simple, and imbue it with such meaning — which is the same mechanism that people do when they read words on a page in a novel.

So, that’s a huge inspiration for me, but it’s not an inspiration like “We’re going to do it like this,” but more “this is possible, this is something that works,” you know? It’s the same way that Pong could be an inspiration for a shooter, where you’re like “people are really compelled by shooting things with things, you know?” Or like Galaga or Space Invaders or something could be an inspiration for an FPS. It’s like, “this works, this is something that works.” Now, that’s well mined out I think.

Nate: Yeah, I think as an industry, we’re pretty much covered as far as “shooting things at things.” There’s nothing wrong with it, but it’s not exactly a new frontier.

Tynan: The way people would see a story in a game, and the way the game does things to make that happen — to me, that was a huge inspiration. It’s really those stories that inspired RimWorld more than anything.

But beyond that, in terms of the fictional backing and theme, there’s these great novels by Allistair Reynolds, there’s Dune from Frank Herbert, and Warhammer 40k, like the slurry of ideas that’ve embedded themselves in my brain from a lifetime of absorbing these types of things, definitely. But, I mean, it could have been a different fiction.

I would say, really, what’s unique about the game is that it definitely came from those stories — stories of Dwarf Fortress, old Starcraft battle reports that people used to write before YouTube was around — they would report their Starcraft game with all of these emotional turns, and little stories of like this space marine who was trying to stop the zerglings and he couldn’t. Creating stories from simple game mechanics I think is just something fascinating to me. Whenever I’ve seen that, it’s always been interesting to me, and those are the things that inspired RimWorld.

Nate: I’ve noticed right up front in RimWorld that you have a lot of tools to give the players a chance to choose how they tell their story. You can choose an AI storyteller, you can generate a map and choose where in the world you want to be. With the scenario editor, you can add twists to the mixture — I tried a cannibal tribe, for instance. It’s a lot more robust than “easy, medium, or hard.”

Tynan: Well, the thing with easy, medium, and hard is that it ultimately revolves around the idea that you’re going to get your rocks off in this game by triumph, you know? [laughs]

Nate: Right.

Tynan: You know, so it comes back to the same thing. All those things, those old patterns, they come after the same thing — which is okay, it’s a challenge, the game is about challenge and you want the challenge to be appropriate for you. That’s part of RimWorld, because there is definitely a challenge, and “triumph” is definitely a part of it, but the reason its set up that way is — it accomplishes a few things.

There are different types of players who want to have different types of experiences. Or maybe the same person, who just wants to do something different. Some people just like building stuff in the game, so I wanted to create an AI storyteller that would support that. Some people want more of a steady, kind of classic challenge curve, so I created one for that. And some people really just want to try something random, so I created that, too.

Nate: Random is by far my favorite. That’s the Dwarf Fortress way to play. Lots of “FUN!”

Tynan: It just does random stuff like 4 raids at one time, just crazy things. I would say there’s another aspect of it, which is if you let people choose these things, they will be more accepting of them when they happen. Especially random — I wanted random to be part of the game, but I don’t feel like I just force that on everybody because it’s unfair. It’s literally unfair, and kind of ridiculous at times.

Nate: But awesome.

Tynan: Yeah, like there’s such value in it, because it creates these absurd extreme situations, which are great for the drama. It’s great for the intensity. But it also, the way it does that is just by saying, you know, “fuck your rules, fuck your pacing, fuck your difficulty level. I don’t give a shit, we’re just doing something, right? We’re just doing something totally random.”

And to just give that to everyone, I think I’d get a lot of complaints, because a lot of people would just get killed — they’d build some colony, and fall in love with it, and then get murdered. So I want it to be something that you’ll choose, and if you choose it, you’ll accept it. If you know what you’re getting into, you’ll accept it and enjoy it. It’s basically trying to make it acceptable in that sense.

And the other thing is, it’s just basically an easy sort of a cheap way for me to make the game appeal to a broader variety of players. You know, I’m always looking for something that’s easy to implement, and essentially a requirement of how I design the game was that everything has to be able to happen at any time. There can’t be any restrictions. Which means that every event, every incident, has to be made in this really robust way, so that it can occur in context.

So your guys can try to get married while there is a pack of man-hunting wolves at the door, or something, right? And the systems will account for this, and they will work together. And once you’ve done that, it becomes very easy to just sort of throw all those events together in different configurations. And so it’s just basically a cheap way of trying to push the game system in different ways. Once the game system is robust, why not make it work in different ways?

That goes for the biomes too — why not play in the steamy jungle, or this frozen wasteland, or a nice forest, or this desert? The game supports all these, because it supports every range of temperature from like -100 degrees Celsius, to +100, where everyone’s dying of heat stroke, or getting burned and dying. So if it’s going to support that for seasons, you might as well let people play in different types of environments. Once you built the game properly, it’s easy to do things that let people create an experience which is different from their last one.

Finally, it’s also about replayability, you know? Once you played your normal storyteller in the forest, maybe you want to try a random one in desert, or an easy one in the ice sheet to get a different type of experience, or go in the jungle and deal with tons of disease. All the systems are there, so let people create an experience that they can own. Basically, it’s easy to do, so do it, and put the variation in because it’s already there, so make it work.

Nate: So, speaking of storytelling, because that’s obviously a cornerstone of the game, of this sort of emergent experience —

Tynan: Story generation.

Nate: Generation. Yeah, sorry. [laughs] Will we be seeing any kind of combat log or event log that will kind of help us to sort of narrate the stories, so that people can more easily share them?

Tynan: Well, I mean it’s certainly a classic idea, and I think it’s a strong one. But, as a general rule, I don’t discuss future features, just because I don’t know — I don’t have enough knowledge to predict what’s going to happen. There have been lots of times in the past where I was like, okay… Just thinking to myself, “next month we’re going to do XYZ,” and by the time next month rolls around, like a million things have come up. You have better ideas, someone has suggested something better, and there’s some tax paperwork to do, and it’s not going to happen, right? So, I don’t like to speculate.

We’re working on some really exciting stuff. Alpha — the next alpha, I think it’s Alpha 16 — is going to be huge. But I can’t say what that is. 16 has kind of a mega feature that isn’t talked about. But no, I can’t say specifically about the future because I don’t know. I can tell you something but then, you know, it might not happen and people would get pissed off. I’ll let people be happy when we can guarantee that their happiness is coming to fruition.

Nate: Do you have any idea in your mind of what RimWorld 1.0 looks like?

Tynan: Not really. I mean, what I have is this giant design queue of every idea I’ve thought of, some half-written designs. Basically just a giant Google Doc — it’s like 40 pages now — of all this stuff, and I just sort it. And like the idea you mentioned, the combat log, is on there, as well as a million other things. I’m not going to do all of them, there’s no way I’m going to do all of them. I mean, this document grows faster than I can consume it, because it’s easier to have ideas than to implement them. But, there are certainly some I would like to do.

But no, in terms of specific target, I’ve never really worked that way. It’s always been — to me, it’s essentially sort of a self-soothing, misleading approach to game design, when you’re in my business position, to sort of imagine exactly what the final game is going to be. The reality is, as a designer, I don’t know. If I’m being honest with myself, I know stuff is going to come up that I don’t anticipate, and I know that I’m going to have to change the design around. So I just don’t bother looking that far into the future. Rather, I just keep all the ideas, and when it’s time to do something, I do the analysis at that time when I really know what is going on.

Nate: With something as open-ended as RimWorld, I can imagine that’s probably the only way you can do it without making yourself crazy.

Tynan: Yeah, it’s a design approach that I came up with while I was writing the — well, I sort of developed over time, but for me it works really well. I’m not sure it would work well for a big studio project, where you need to have a marketing plan like two years ahead of time, or something like that. But staying nimble like that, and letting the “future you” make the decision instead of forcing the “present you” to decide a year ahead of time, leads to much higher quality decisions, in my experience.

Nate: So I’ve observed that as features get added to RimWorld, it seems to be that the game gets essentially “darker.” The more things that the colonists can do, the more ways they can horribly kill themselves.

Tynan: [laughing] Well, I guess. I mean, yeah.

Nate: Is the increase in complexity just inevitably destructive? Or is there a way to engineer these experiences with a more positive bent? I guess I’m asking whether that’s a deliberate design decision, or a by-product?

Tynan: Yeah, I know what you mean. I’ve been reevaluating this recently. What you’re describing is essentially the result of sort of my internal process, where I’ve been targeting the things that seem to have the most story drama potential —

Nate: I love that sort of grimdark comedy it creates. I’m definitely not saying it’s a negative.

Tynan: Oh no, I didn’t think you were. I think targeting the drama has been effective, because people are telling their stories of like,” oh my god people have gotten addicted to this, and it was really difficult to get them off, but their wife helped them through it,” et cetera.

We did add relationships and stuff, so colonists do get married, and they also get divorced and stuff. There are definitely happy moments in there. But in terms of the negativity, I would say it’s a side effect of the pursuit of drama on my part. And now I’m sort of noticing — and this has been brought up a few times — a lot of the game mechanics sort of point towards these sorts of grim situations, which is a side effect of the fact that grim situations are really dramatic.

Nate: Understandable. A lot of the fun comes from harrowing situations. That sense of peril, hovering on the edge of total catastrophe.

Tynan: Yeah, it’s a lot of fun. There’s danger and risk and things are at stake, you know? It also ties into game systems, where we physically track if the people are hungry, they’re tired, they’re wounded. It’s harder to track positive aspirational things about a character; those tend to be more, “I feel self-actualized,” or something. It’s very difficult to simulate that.

Nate: Yeah, I follow.

Tynan: It’s like Maslow’s Hierarchy, right? It’s easy to calculate things at the bottom of the hierarchy. It’s very easy to write a system to simulate that. But to write a system to simulate the top of Maslow’s Hierarchy is like, that’s a strong AI problem. I can’t do it. So, you have to sort of fudge it.

In any case, yeah, so I’m seeking drama, but it has ended up with these extreme situations — most of which won’t happen in any given game, but yes you can have drug-addicted cannibal psychopath starving murderers, if you want. But this is something I was thinking about, and I was thinking it would be cool if I added some more positive-type emotional valances to the game, just to balance it out, to have a back and forth, sometime in an upcoming update.

In terms of emotional valence, the game is really hitting the notes with — think of the most grim survival movie you can, like The Grey or whatever — it’s definitely got those feelings in it, and they’re really dramatic, and it’s great. But, it would also be really cool if it had enough variation to have, like, “rom-com” emotions in it too, you know? If you could go back and forth.

And right now you can, to a degree, because there is a relationship system, and people will get in quirky relationships where one guy shacks up with the mother, and then they break up, and then he marries the daughter. You know, you can imagine some goofy story about this if you want. But, yeah, no. That’s definitely a good point. It’s been tipping into dark territory, but I hope to balance that over time.

Nate: So we’ve talked about different aspects you’re considering, things you’re progressively tackling, and the forty pages or so of potential features. Do you plan any of this more than one update in advance?

Tynan: No. [laughs] Well I have a lifetime worth of content — it’s all there swirling around, right? Uh, no. Planning? The word “plan” implies that you’re going to do it, like you really think you’re going to do it; there’s a very high percentage chance you’re going to do it.

Nate: Right, that’s where I’m going with this. [laughs]

Tynan: Which I don’t do. In fact, I’ll keep that to under two weeks if I can. We might add more features to the next alpha, or we might just release it as soon as the current workload is done. There’s just no point; why plan, why try to guess at the future and then lock yourself into some specific course of action? Just decide later, because you’re going to know more, right? And you’re going to have a better decision if you just do it later. So, let “future you” decide, you know? Because “present you” is an idiot.

It would be like if I let my life decisions be made by the 5-years-ago version of me; he was an idiot, he didn’t have all of my experiences. So, I try to decide as close as I can — essentially decide things with as much knowledge as possible, which implies deciding things as late as possible. Which, I can do because it’s an indie company with just a few people, and we have that kind of nimbleness. It’s an advantage of being indie, and it’s one that I do play to the hilt, every time.

Nate: How’s that been working out for you so far? The Steam release, obviously, has been pretty successful. Steam Spy’s latest estimate is over 320,000 on that platform alone, but you’d been selling it well before that. Was there a big jump once you released to Steam?

Tynan: Oh, yeah, definitely. I would say sales probably went up over tenfold, I think, when it went on Steam. Obviously it did fall down; it was crazy in the first few days, and in the first week, it’s gone down quite a bit since release. But the game still is sort of clinging to in-and-out of the Top 10 sellers on Steam. I think it was in the Top Ten today and yesterday. And this is over two months after the release, and it’s not even a final game; it’s early access.

So, I’m pretty damn happy with it. I mean, what can I say? It’s been doing really well, it’s been doing great. So, it’s a fantastic validation that somebody likes what we’re doing here. And, thank you to everyone for that. What else can I say? It’s been doing really well.

Nate: Does that level of validation make you want to try new projects, or are you sticking with RimWorld for the foreseeable future?

Tynan: Well, again, I don’t know, right?

Nate: That’s a question for “future you.”

Tynan: Yeah, seriously. It’s funny. I mean, every time I do an AMA on Reddit, or I did one on Kotaku last week — Literally half the questions are, “Is this going to happen? Is this going to happen?” I’m like “I don’t know! I don’t know! I don’t know the future, guys.” I’m much better at talking about the present and the past, because otherwise I’m just guessing at it all.

Nate: I think people get really jumpy because a huge percentage of early access games just sort of fade into nothingness. It makes us cagey. We want to know what’s going to happen.

Tynan: And I think it’s totally pointless, too. Because you can ask for giant promises, and a developer can give you big promises. It doesn’t fucking mean it’s going to happen, you know?

Nate: It’s true.

Tynan: In fact, I would say, you should look for the developers who don’t make those giant promises. Because the ones who are like, ‘Yeah and it’s going to have sharks, and it’s going to have nuclear bombs!’

Nate: More sharks and bombs than you can ever potentially see. Eighteen quintillion —

Tynan: [laughs] I’m not naming anyone, okay?

Nate: [laughs]

Tynan: But the bigger the promises, the more likely they are to be disconnected from reality. So, I think an intelligent gamer who wants to find a developer they can trust, should look for someone who’s making measured statements, and is not saying things that are too distant from reality. Because they really shouldn’t be making those claims, because they really don’t know. And that’s what I try to uphold all this time.

Obviously, there’s a lot of new players coming into the RimWorld sphere — especially recently, with the Steam release — but it still surprises me sometimes when I get people who are like, “Oh, are you going to abandon the game? Are you just going to drift away?” I’m like, guys I’ve been doing this for three years! Every single month, somebody’s like, “Oh, you’re going to abandon the game now,” and they’re like, “Oh, tell us what’s going to happen. Otherwise, it’s obvious you’re just leaving.”

And every time I release something and it goes great and there’s a new update, all those criticisms are essentially silenced and proven to be wrong. But, then there’s always a new generation of people who are like, “Oh he’s going to abandon it this time.” I’m like guys, please just look at the record.

Nate: It’s a very paranoid market.

Tynan: Go look at the record; I’ve done exactly everything I ever said I was going to do, and more on the game. I’m pretty much one for one on promises in RimWorld. So, I’m saying I’m going to keep developing it, and it’s going to be finished in a satisfactory way. So, that’s going to happen. That’s all I can say. I really do value credibility, and I think it’s really important to be serious about saying things that you know are actually true, or will become true.

I take that very seriously, which is part of why I don’t give a lot of speculation about future events, because I don’t really know. Yeah, I try to build that up, and I think that the players who have been around for a while will know that. But, yeah, all I can say to anyone else is, “Don’t ask me about the future; I’d rather you look at my past.”

Nate: I do agree that the past is perhaps the strongest predictor of the future.

Tynan: Well, I mean, there’s the old adage, right? Don’t listen to what people say. Watch what they do. And that’s how you really get to know who people are.



I want to once again thank Tynan Sylvester for taking the time for such an extensive discussion, and for his candor. If you haven’t already, do check out part 1 on Tynan himself, and part 2 on the state of the gaming industry as a whole. Finally, don’t forget to read my take on the ups and downs of running a colony of cannibals, drug addicts, nobles, and cave men.

Follow Nate Church @Get2Church on Twitter for the latest news in gaming and technology, and snarky opinions on both.


Please let us know if you're having issues with commenting.