RPS Exclusive: Soren Johnson on Everything PC

Soren Johnson has many things to be proud about. Most gamers would argue that being the designer of the magisterial Civilization 4 heads the list. We'd probably agree. But if you go way-way-way down the Good-Stuff-Soren-has-Done document, beneath even things like buying an expensive round in a bar a few times and opening a door for an old lady, you'll find that he was the first developer to ever mail RPS, saying that he'd like to do an interview. We'll probably love him forever for that.

After some of RPS' usual cheery incompetence, we finally got around to it after he'd returned from Brazil, visiting a games conference. The following hour long conversation wandered the entire expanse of PC Gaming – what's important, online sales, how smart Introversion are, how ace Desktop: Tower Defence is, how PC Gaming relates to Peru Ubu, the Sex Pistols and Television and pretty much everything else which wasn't to do with his current project, Spore.

How PC Gaming looks at the end of an exciting 2007, from one of the PC's most interesting designers. It all starts beneath the cut.

After this bit, anyway. We talked a lot about Civilization 4, but that'll be written into its own article into the near future. Also, some aspects – involving certain uses of AI, his influences and why games anyway are span off into a separate interview in a forthcoming issue of PC Gamer. Rest assured, we'll link to that whenever PCG stick it online.

But there's quite enough to be getting on with.

RPS: You're back from Brazil now – how did you find it? Was it interesting?

Soren: Yeah, quite interesting. It's a little like GDC, I guess, but a lot smaller as they're just starting their game development community down there. They have no real retail games market. I saw some of the pirate shops they have there, and they're stacked full of PS2 games for a couple of bucks.

RPS: I saw the Polystation on your blog...

Soren: I laughed out loud. I thought it was a rip-off Playstation, but it's much more like a megadrive for cartridges. Totally bizarre. But if you can't sell to your own market, how do you develop as a game developer?

RPS: You speak to people in Russia with people dealing with a similar problem. What are the developers like down there?

Soren: There's only a couple who are working on “standard” game developer stuff. The Deer Hunter series has actually been developed in Brazil for the last X number of years, which I didn't realise. I met the people who've been working on that. They [Brazilian developers generally] have a lot of the same frustrations that developers have with publishers, except maybe a little more extreme as they're so far more removed. Deer Hunter isn't necessarily their passion. I don't know if they hunt much in Brazil, but they were happy to have a contract. There were a lot of people hoping to make MMOs or online games or mobile stuff or whatever. Something will come out of it, eventually. We've seen a lot of the good stuff come out of Eastern Europe, and Brazil is one of these new frontiers.

RPS: The Eastern Europe stuff is interesting, especially for RPS, as it tends to be PC Stuff due to the lack of a licence or corporate approval of any kind. The basic democratic nature of the format. How important do you think that is?

Soren: That's the reason why PC is always going to be super important. It's always the democractic system or whatever. PC gaming is really starting to flower more now, when the last few years... well, the last four and five years we've seen the middle fall out of PC games development. There's a lot of cool stuff that's happening in the fringes, and there's still a few blockbuster Triple-A franchises that are still cranking out... but if you look back at 98 or 99 there was a huge range of games available. So if you wanted an off-kilter RTS like Majesty... well, that stuff can't be made anymore. I think this is changing with online and that people are getting better and better at making PC games – it's not as hard to make one with a small team anymore.

RPS: That's interesting. Ballooning team size is a trend which has been well discussed over the last decade, but do you think that's reversing slightly on the PC now? As long as you set your sights intelligently...

Soren: I'm in an odd position, going from Civ 4 which was a big project, to Spore which is a mammoth project... but it's just that I think most of the stuff which is going to benefit from smaller teams is going to be stuff comes across the web. We see that all the time now. It's finally a viable market. To me – and ironically, this isn't a web-based game – but something like Defcon. That game didn't have art. Which is brilliant.

RPS: Very artfully chosen unart, if you know what I mean.

Soren: It's a weird feeling – I play the game, and it looks great, because they chose a brilliant style. It doesn't need art. It just fits their game perfectly. But the interesting thing to me about Defcon is that the size of the game is right. It's a pretty good game – it's not quite a brilliant game, but it's a fun game to play. But you can't say that it's too simple or complex. It reminded me of a lot of RTS when I first discovered them a decade ago. Now it's really difficult with an RTS to...

RPS: Not submit to the endless feature bloat. You have to have all the bullet-points.

Soren: Yeah, you can't make it without a campaign and scenarios and an editor and cutscenes and all that extra junk. Really, that junk is preventing us from making more interesting games. It's kind of a paradox. Obviously, people want that junk, and it's a good thing for those people. And the editors which people make their own scenarios is great... but that stuff all comes
at a cost. I think maybe we're starting to realise that now. The answer is that the economics get turned entirely around when you don't have to deal with Best Buy and Wall Mart and what not.

RPS: People endlessly talk about the declining PC, but the figures never include those specific areas where the PC is expanding – the online sales, MMOs. When we started RPS, for me it was about trying to redefine what a PC Game is to include all that. I'm sure Peggle will be in everyone's top 10 games this year... but can you imagine a game like Peggle being included in mainstream PC talk a few years ago?

Soren: Or Desktop: Tower Defense. That's an awesome game too. I play that more than most strategy games I've played this year. Which is weird but... what does that mean? So yeah, absolutely. The PC Market is no one thing any more. There's no sales figures you can look at. The question is simply is “What is the variety coming through? What are the different options available we didn't have three or four years ago”. For me, PC Gaming should be like Punk Rock – being able to do whatever you want. And people are forgetting that the Punk period isn't just the Ramones and the Sex Pistols... it's Talking Heads, Televisions, Patti Smith, Pere Ubu, Gang of Four... this huge variety of stuff because people were making it up as they were going along. It was easy enough to make music that people did what they wanted to. And that'll always be the advantage of PCs.

RPS: I interviewed Doug Church about what it was like developing when the PC had started being a real gaming platform, circa 92 or whatever. And, basically, when doing Shock they just didn't know what they were doing. They were original by default. In the following 15 years, like the feature bloat, by learning what works, it also limits you a bit. The people in the mainstream need to work out what ELSE works. But the European teams try stuff which no-one else does, because they don't know any better. Like the Bohemia Armed-Assault guys, trying forever working on their butterflies...

Soren: That's games. And it's funny how much pleasure people can get from little things in games which you've never seen before. I know what you mean – think about RTS. What does that term mean? Now it means a very specific thing... but what else is Real time strategy? The first Sim City was real time strategy. Populous was an RTS. Rollercoaster Tycoon is RTS. You could say MULE was RTS. Obviously Defcon and Darwinia are. RTS should be the biggest category there is, but right now it's very, very specific. There are a few triple A titles which are trying to push it – World in Conflict was an interesting take on that. But you need people to come along who aren't intimidated by all the stuff that exists already in the genre.

RPS: Apart from the Online market, what's most exciting for you in PC? I mean, in terms of trends.

Soren: To me... Heh. I don't have answer other than online, because I think online is what makes it all work. I mean, obviously the open development and you're developing on the same platform for what you're developing for, which makes things easier. With Playstation you're always one step removed from the hardware. But I think online is really, really important. Those free-to-play models are finally taking hold over here – your Mythos or Travian or Puzzle Pirates and while I haven't got into it myself, I hear Maple Story is starting to take off.

RPS: I saw Rossignol's Girlfriend playing Maple Story, which says a lot.

Soren: Does she spend money on it, or is she a freeloader?

RPS: I think she's a freeloader.

Soren: When you talk to the developers of these games they say that freeloaders are very important. They provide the content for everyone else. If you didn't have the 90% of the people who don't have the super-special hats or whatever, then the super-special hats don't have any value. Its' not as if the free people aren't contributing.

RPS: I was reading your notes on the back of Civ 4, which were great. I got the feeling that sort of thing, which few people do anymore, was something you liked when you were younger. Walker always talks about he misses cloth maps...

Soren: Ah – another one of those lost art things.

RPS: What other things do you miss from the PC? That you wish we still had?

Soren: Well, I miss the fact that we can't count on people to have a joystick. That's a huge thing. I didn't actually start PC gaming until late – I mean, I did in the sense I was a C64 and Amiga and Mac Guy. But I didn't start PC gaming until I worked for Firaxis. But Amiga was in this great period where you knew everyone had mouse and joystick. There were even some games which let you use both, and assumed that's what you were going to have. It just frees up your designs.

RPS: One of my favourite arguments which I've heard writers use is kind of retroactively claiming the Amiga and C64 as PCs. I like that stance. What I find interesting about them is that they were a crossbreed, so you saw the games that you'd define as “console” or “PC” titles. But we just playeed everything. There's a bit more of that coming back. It's not there completely, but it's in interesting to think of the PC as an arcade machine again, occasionally. The good side of the convergence...

Soren: I'm glad its' starting to move back, but it's hard. As a mainstream developer, we can't rely on that yet. But you can't really look back – there's always a primordial soup thing with the C64. You can see the roots of a lot of PC games there, but if you think of it strictly as PC... you'll perhaps forget things like Sid's Pirates, back when it was on the Commodore and whatever else. It was really like a console game. It relied on the joystick. It wasn't one of the Microprose games which came with a big keyboard overlay... it was a joystick game, so essentially a console game. When Sid came to remake it 4-5 years ago, we were almost moving it from a console game into a PC game, because it was the first time it had ever used the mouse and adding some extra complexity it didn't have before.

RPS: Actually, spinning off, mind if we talk a bit about AI in games, since it's the area of design you're probably best known for? It's a topic much gamers don't know much about. It's something we talk about in the vaguest of terms.

Soren: Yeah. I know a lot of people who work on first and third person games, who talk about how when gamers talk about bad AI, they're actually talking about bad animations. But, please, go ahead.

RPS: Actually, it's kind of an open topic. I'm just interested in hearing what you say – it's kind of a set the topic then step back and listen.

Soren: Well, one thing that's kind of frustrating is that there's a sense of what academic AI is, which means there's all these tools they can give you – neural nets, genetic AI, machine learning or whatever. And then there's Game AI. And then there's this giant gap between the two. Even within Game AI, it means so many different things. In the shooter, AI is usually very scripted. It's not AI in the classic sense of the term – you go around the corner, and stuff happens. Guys either run to the right place, or they don't, and at some point something breaks and the guys just run into the wall. I've never worked on a Non-Strategy-AI title so it's difficult for me to know the problems they face, but a lot of times I've heard them talk about how AI happens inside the player's head. It's a bit imaginary. So when they notice something wrong with what the NPCs are saying or how they are walking, it breaks the illusion of a real world environment.

[At which point we go into a section on AI and what's it for which found a home in the Gamer interview.]

RPS: I remember seeing Gary Penn of Denki at a EGDC a few years ago arguing that AI's purpose is to die entertainingly – it may as well be an amusement. Of course, what you're doing is completely different. The AI has to play the game.

Soren: There's another level, which is the difference between the good and great strategy games, is that the AI is not only just trying to play the rules smartly – if you really want to take it to the next level, you need to think about whether the AI player is actually making the game more fun for the player. There are some strategies which you are going to want to simply take off the table for the AI, because while they may fit under the game rules, they're going to be annoying for the player. There's a fork in the road, where you can go down the optimal path – which for some things is appropriate, such as with a fixed problem like chess or Go, this is the road you want to go down – and then there's the other road, where you don't worry about that and concentrate on personality. For example, in Civ4, Tokugawa doesn't like to trade. Obviously, that's based of Japanese history, as he's really protective about his resources and his technologies. And that's NOT a good way to play Civ4. That's not going to do well for you. But it's interesting to have a character like that in the game who plays like that. I think it's important to set that distinction between good AI and fun AI.

[At which point we talk a lot about Civ4, which will get its own feature soon. Soren ends that by talking about how inspiring he finds the mod-scene, in terms of the ideas it germinates...]

RPS: This reminds me of some things Chris Taylor was saying in an interview to me recently, about Creators inspiring each other through their work. He was annoyed by the Zoom in a game, and inspired to make the infinite zoom in Supreme Commander. Musicians talk about stuff like that all the time – dialogue between creators through their art. Do you ever feel like that?

Soren: Definitely. It's tough. I have this document I keep around which is a list of all the ideas I'd love to make if I ever had time. A lot of them are variations of cool stuff I saw other people do – if I just changed this, or changed the focus onto this it'd be awesome. I'd love to make a competitive RTS that the focus was on something different from military... there's no end of stuff. There's not many games which come out which are shocking moments of innovation and brilliance. Those are few and far between. Games are tough. It's a huge culmination of all these different bars you have to hit. Not necessarily great graphics, but graphics which have a style and are appealing – a game like Travian is like an Asterix world, but it's an appealing graphical style. It's not a question of technology. And then you need to make the game understandable. If people ever get there, you want to make the game deep. It needs to hold together technically. It's a big challenge, and there's games which do some of those things right, and not the other ones. It's why progress in games development is slow.

But the ironic thing – I feel like a lot of times when I talk about some innovative concept or game I'd like to work on which I don't really see out there... a lot of those ideas are stuff which has been done a long time ago and kind of forgotten about. There are some great game ideas that lie buried in the 80s as they didn't spawn successors.

[Soren Johnson is currently working on Spore. His website can be found here.]