How Creative Assembly's Process Breeds Quality

Tim Heaton is the studio director of Creative Assembly, the development studio behind the popular Total War PC strategy franchise and an upcoming Alien action game. How does the developer, which has been part of Sega since it was acquired in 2005, manage to maintain its quality standards for the franchise?

In this interview, Heaton describes how the team is structured to deliver 90-plus Metacritic games, how it maintains its autonomy from Sega while still availing itself of the support a publisher-owned studio expects, and explains the push-and-pull between two teams with two different goals and cultures, but one style of management.

So you oversee all aspects of Creative Assembly?

Tim Heaton: Yeah, absolutely. And we're a kind of a split studio; we're a schizophrenic studio. There are only two things going on in the studio -- one half is Total War and one half is this big console game that we are in production with now. It's interesting. When I came in, those two components existed, and they possibly have even grown farther apart since I've been there, deliberately, as a part of the strategy.

Total War is a massive, heavyweight PC game. We're doing rather interesting stuff with it. We're a very established, 11-year franchise. Some of the key players on that team have obviously been there for 11 years, so they are really sure about the game that they make.

And then the console team is a little bit fresher and newer. And we run the teams in vastly different ways -- the console team is more rock n' roll, and the PC team is a little bit more staid. And we project manage them differently, etcetera. So they're very separate teams, and I'm one of the only people who kind of actually bring those two strengths together.

Obviously, Total War is a very established franchise. So that's as smooth sailing as you could hope for, I guess, in the sense that you know it's going to get the resources and the audience, right? How do you work with Sega as your publisher, as an internal studio?

TH: I think we're really independent. Before Sega bought CA, it was a very independent-minded studio, and since we got bought, we've maintained that kind of attitude -- to the point that we run the marketing for Total War from inside the studio. It's something that normally a publisher will completely control. Recently, we've said we want to be really specialist on our marketing, and have really focused people, and so Sega has allowed us to build a team around that.

So it's good; they're a good dad to have, because they provide support where we need it. They tend to be watching the market for us, to steer us into the right direction that they think we should be going. And then infrastructure stuff, like Q&A and etcetera is great, just to have them as a big structure there. But certainly for Total War, I feel like CA totally guided that franchise; we're the kings of that franchise.

With the big console game -- which is based on the Alien IP, with the Fox IP for the films, and which Sega working with already, with Aliens: Colonial Marines from Gearbox -- that's been more interesting, because that is where we need more of the publisher guidance, due to the relationship with Fox, and so on. Sega certainly wants to keep a view on that.

But still, it's run as an independent team. We kind of decide our own milestones and we communicate with all the parties, whether it be the licensor or first parties. We go direct to those guys.

Total War: Shogun 2

When it comes to running a studio with two teams working on two different projects, with different characters about them --

TH: Oh, totally different.

Where are the things that come together, and where are the things that stay apart, and how do you identify those?

TH: So I think if you're a mega-studio -- like your Ubi studios and your EA studios -- I think there's a huge opportunity for cross-fertilization, and becoming more efficient through that. A medium-sized studio like us, I think you've actually got to build a wall between them; you've got to create a gap. Well, you've not got to allow yourself to be distracted by the problems of one team pulling away from another.

And I think that's an issue that CA has had in the past with Total War and Viking -- Viking was the last console game that CA produced, and that was a flawed game. To some degree it was flawed because we were also trying to make Empire: Total War at the same time. So yeah, I actually think with a small studio, it's useful to have dialogue between the two teams, and not useful to share resources.

Now, with your console team, obviously the PC team is strategy, strategy, and more strategy, right? Does the console team do other genres?

TH: The console team is creating an action-genre game on consoles, so, yeah, they are. And we recruited in. We started with a really small team on console two years ago with about 20 core staff, and we built out to 70 now. So we've been looking for a very specific kind of person to make action games. We've brought loads of experience in.

Luckily, bittersweet-wise, we've been able to pull staff from other studios that have closed down in the UK, and there have been a very few of those around us. But we're also now pulling from the States, and we're looking at all those UK dev staff who went to go work in Canada recently and over the past few years, some who haven't found it quite as interesting as they were hoping and want to come home, or have had a family, or whatever, want to come back to the UK. So I'm sure there are expats out there who want to come back. We've already taken a few on.

Given the state of the UK, exactly as you say, there are a lot of studio closures and a lot of people relocating. Has recruitment been a challenge?

TH: Yeah, it has been a challenge. It's been helped by the fact that we picked people up, along the way, from there. And I don't necessarily see any particular big studios in difficulty at the moment, so I think that's kind of dried up, which I'm pleased about. We want a really strong UK scene. UK started really strong but it has got a touch weaker, but there's still a depth to that experience there. There are some great developers for sure; it's not dried up in any way.

You think of the UK as a strong base for game development, and it has been since the '80s, right?

TH: Absolutely. It was the place.

It's just more market forces than I think anything to do with the UK intrinsically as a talent pool, or anything like that.

TH: Absolutely, and we look at our burn costs and stuff, and we're absolutely competitive with the U.S. And you know, we would like tax breaks, which we don't have over there. [Ed. note: A UK government tax break plan was revealed soon after this interview was originally conducted.]

I actually don't think tax breaks are the be all and end all; I don't think they're the definition of victory. If we get tax breaks, you end up just recruiting a few more staff, and maybe that's not profitable. You don't necessarily make great games with just loads of staff who are not quite as good as you need.

What do you think makes great games?

TH: I think it's building quality into the team. And you don't add quality in the later stages of the game; it comes from each individual. And I see it on Total War. Interestingly, in my previous life, I worked at Electronic Arts, and worked really closely with Crytek. I saw it at Crytek, too -- that they strive for quality at every single point along the way. And when the core staff on a team take that as the number one priority, then it feeds through really great. The average Metacritic on Total War is 90 percent; that kind of proves a point, and that's the strength of CA.

Napoleon: Total War

When you're talking about striving for quality on every step of the way, can you enumerate on that in the sense of process? Do you have reviews, for example?

TH: Yeah, you bet. We go through the design process, which is pretty wide and all-encompassing, and then scope it out and then start cutting features, because hopefully we've got such a huge bag of features that we want, that it makes it physically impossible to ever make a game with all those in. So we'll cut some straight out, and we'll then start prototyping and judging the quality all the way through.

And our creative director Mike [Simpson] has been on Total War, as an example, since the first one, and he owns the scope of the game, and he will pick and choose his feature set as he sees it bubble up through prototyping. But equally, all the key leads on the game are working to certain quality thresholds as well.

While we're prototyping we're also figuring out how long the work that we'll need to do will take. It's a piecemeal work. And sometimes we'll need to pull stuff out, because although we can reach quality threshold that we want, it will simply take too long. So we will cut features, rather than delivering an 80 percent quality feature. And that all builds on top of each other, and all the features need to be of high quality; we don't really want to let any substandard features go in.

And then, through production, actually, we do what we call "Metacritic analysis." So we will break those features down into subsets, and we both look at it from a player's point of view, and a reviewer's point of view, and we'll weigh certain features as to how we see players and reviewers look at them, and they'll build up to a 100 percent score, and then we'll judge where we feel we are on those individual feature sets, and see the momentum on those and the velocity on those, too.

And so if we see one flat line and it's not where we want it to be, we then will cut it. Well, we'll cut it really late in the day. I think teams are really scared about doing 90 percent of the work and then cutting it. It's kind of like, "Well, it's nearly finished; I... I've done all the work! Please don't cut it! I'm sure I can make it better." And we're fairly brutal on that.

I'd much rather not see a feature in the game but still pay for it than risk... You know, every step of the way -- from the beginning to the end -- we're talking about a 90 percent Metacritic. That's our goal. That's what we tell Sega. And we communicate that through graphs, basically, of where we think we are.

We build into that also, on that Metacritic analysis, external events. So if we think we've done a really great PR job, if there's an individual event that we've done really good, we might add, you know, a .5 percent Metacritic. If we think it's fucked up or somebody's not done their job right, or miscommunicated something, or whatever, we'll see that in our Metacritic analysis. And we share that with Sega on a weekly basis, so that they can figure out how we're doing, too.

Do you have anything like a safeguard to not cut too much? It seems like cutting too much could be a danger.

TH: Oh, for sure. Well, we start with a bigger scope than we need, until we know what we can cull.

Intentionally?

TH: Yeah, absolutely. And we do work to a date, and we absolutely want to work to a date. We almost have to be constrained; if we weren't constrained, we'd never finish a game. We like the constraints, and we like making money, and we know we need to be ready for certain dates and events; we know how marketing works.

So it is a fine balance, but the beauty of Total War is we've got 11 years of experience of that balance, and we pretty much prove we can get it right. And we're doing exactly the same thing on the console game. I'll be honest. We don't have 11 years of pure experience on 360 and PS3, but we go through exactly the same methodology.

You talked about getting a piece of the game 90 percent there and then cutting it anyway. Does that have a hit on morale, or does everyone understand?

TH: Yeah, it does. So there's two hits on morale while we make the game, and we constantly debate this. At the beginning we don't tie down a lot of stuff. We let a load of stuff bubble around, we don't define it too hard, we take some risks...

And certainly some elements of a team -- and this always happens with every team I've ever worked with -- just go, "Come on, just give me a game design document. Just tell me what I need to do and then I will do it to the best of my ability". And we slowly, hopefully, educate people that's just not the best way.

And we will enter a fog of ideas for quite a long time, and some of those things will have risks against them right up until the end, and then we might pull them. And really yeah, it does. It pisses people off, absolutely, but it's for the best. But nothing makes the team prouder than delivering a 90 percent game and selling two million copies. So that's the bottom line, and people do come to understand that.

Do you just wait for the end and that payoff for people to see, or do you have anything that you do to boost morale?

TH: Well, we do regular updates, obviously, and we do monthly or bimonthly sessions, but there's so many of us now that we can't do it in the studio. We go to the local cinema and we do a half day presentation of progress, and share some of the thinking that's going on, and we'll do a report from GDC, and so on.

So we do that, and equally we have the Metacritic analysis, which we keep showing to people. And as long as people can see that, actually, by the time we get there, we'll be where we want to be -- then people will understand what the decisions that are getting made late, or people's work that doesn't get into the final game, is still a valid way to think about the big project.

It sounds like also you try to avoid documentation as much as possible. How do you use documentation in the studio? When you're talking about this "bubbling up", is that a prototyping process?

TH: It is. The team is about 100 development staff on Total War. And so we break the team into functional groups -- so there's a campaign group, and a battle group, and a UI group, etcetera. So we devolve responsibility down to those guys. And to be fair, on the amount of documentation that we do, we almost leave it up to them.

At top of those functional groups, there's communication between them, so the leads on those groups need to have regular dialogue, and we'd much rather they'd just talk than do anything too systematic about that.

And below that, some of the more buttoned-down teams and do a huge amount of documentation. And in general, the overview of the whole project management, the team is reasonably waterfally; we're a touch old school on it. But actually, when you get down to the lower level of the functional team that, if they want to, they will go Agile. We've done a fair amount of support on teaching people about Agile and supporting them.

So it's great that we can go, "Hey, you just figure it out." We went through a phase when I was early in the studio. That's what I wanted to do, and everybody went, "Oh, that's great! We really want responsibility! Give us responsibility to make decisions!" And then we'd sit there and they'd go, "Right... So what do you want us to do?", and we'd go, "Okay. You're not getting it yet. That was your decision to make."

You talked a little bit about the fact that you take into account how reviewers will perceive the game, versus actual players. Can you talk a bit about why you approach it that way?

TH: Yeah, we do. Reviewing is important; it's still important. And there are some hygiene factors that perhaps you need for some reviews that maybe aren't as strong for players. At the moment, Total War is primarily a single player game, and we have talked about how valid multiplayer is in the way that we've sometimes done it in the past. We actually feel we'd be really stung by reviewers if we just took multiplayer out. It would maybe be a tick box feature that would disappear.

And for the players, if potentially 80 percent of players are playing single player only, then the vast majority aren't going to be affected by that. Given that, just to be clear, some of the multiplayer ideas that we've got in development at the moment are clearly going to define the future of Total War, I think. So we're not going to cut multiplayer.

But things like that, just the perception of the first five or 10 minutes of the game is super important for some reviewers -- not hardcore reviewers, but more mainstream reviewers. Whereas our average play time on Total War is 40 hours; 50 percent of our players play for 40 hours, and 25 percent of our players play for 100 hours or more.

So I've heard some stats about Skyrim and other games, and I think we've beat Skyrim on playability, and we see that day to day, direct through Steam. So that's not research; that is real data from our players. It's quite amazing.

When you hear the statistics about the number of people who actually complete the campaign in even a relatively short game -- like a 10, 15 hour console game -- they can be quite low. I'm surprised to hear that.

TH: I know; it's really cool. We have a niche audience, and they just love it, and they come back for more. And downloadable content -- we do loads of downloadable content now for Total War, which elongates the game and adds to the game, and that's been hugely successful.

Do you think it's your audience? Or do you think that it's something that you intrinsically do with the design of the games? Or is it a blend that keeps the engagement like that?

TH: Yeah, I think it's that people make their own game within Total War. We're not a story-driven game; we're not about a directed campaign. And we'll occasionally do some elements of that -- but yeah, absolutely, you make what you want of it. So when you're allowed to do that, you will follow it through; you attach much better to it, I think.

And it's such a huge, deep game, and that's one of its selling points. I've had conversations with people at Sega who almost feel like we're doing too much -- that there's too much content there. Sometimes a publisher will go, "Ehh, there's 100 hours of gameplay for 40 bucks; maybe that's not the best way to do it." It absolutely is; it's just one of the things that we can do.

How do you make that case? Because I understand exactly what you're saying...

TH: Because we're unique. I mean, we're not unique, but we are one of those players that can offer a huge amount of content with the way the game's made up, and it's quite a unique kind of game style. I think if we cut it, a lot of our hardcore fans would be very, very disappointed. Day one, that massive, deep experience.

And my guess is that a lot of Total War players don't buy many other games; they want a strategy game, and Total War fulfills all their needs, to some extent, or certainly for long periods of time. So it's one of our pillars, to give huge amounts of content.

It's the kind of thing that you can't really illustrate with a spreadsheet, which is how publishers make decisions. So I'm curious about that.

TH: Well, that's the beauty of us still feeling quite independent. And Total War is perceived throughout Sega as highly successful. So luckily we are in a strong position to defend our case. And if we don't sell games, we will certainly quite happily rethink that policy, but until then, if we hit a sweet spot...

Plus, I think that heavyweight gaming experience, that tentpole experience, actually allows us to do really interesting things with extra content, with some of the ideas we've got around multiplayer at the moment. And it's actually going to allow us to do some clever things that we couldn't do if we were really a constrained, small game. And so it's a great strategy for the future, I think.

Total War: Shogun 2 - Fall of the Samurai

Do you have any more detailed analytics around what your players engage with and how?

TH: Well, we do. We have an in-game metrics system, which has collected one and a half terabytes of data so far from Shogun 2. And so we absolutely can see how people play to a huge, deep level, and that is feeding back into the design of our next game hugely.

And then we also see how people play through Steam analytics, and some other data that we collect, and we can see how people purchase, we can see at what price point they like to buy, how long they wait till they buy downloadable content, whether they buy one piece of downloadable content or all of it, etcetera, etcetera. How many hours they play.

So bringing all that together, it is the future. Metrics is the future. And we hear that from social all the time, and we truly believe that, and we have a telemetry set of people within the studio who do that analysis on a daily basis.

Like you say, you have 11 years of experience, so there are people at the studio who are positively sure they know exactly what the audience likes, wants, and how it functions. Has it yielded any surprises?

TH: Yeah. I find it tough to pick one particular insight that would be pertinent, actually. I'm not sure.

But have you had that experience?

TH: Oh yeah, we have. I mean, I won't be able to say what areas of the game worked or whatever, but, certainly, absolutely, there are elements that helped our balance hugely. But we absolutely can see areas of the game that don't get played, and don't get played a surprisingly large amount, and we see how people want to auto-resolve battles, and things like that. So it definitely steers us in a particular direction.