Back in PlayStation 1 days, Driver was the height of gaming, a pioneer in the open world mission-based driving genre. And Newcastle-based developer Reflections was a power-house. But Driver 2 stalled and Driver 3 crashed. Driver: Parallel Lines was better.

Atari flogged the IP and developer to Ubisoft, and PSP game Driver '76 emerged. Then, this week, along came Driver: San Francisco - a towering return to form and the best Driver game yet.

The Driver: San Francisco development team peaked at 220 staff. The project took four years. Then five, following a year-long delay. A bespoke engine was created from scratch. The game's promise? Driving in a world created by the comatose brain of hero Tanner following a car accident. And a Shift mechanic that allows Tanner's spirit to leap between hosts. In other words, Driver: San Francisco was one hell of risk.

Martin Edmonson, founder of Reflections (now Ubisoft Reflections), must deliver one hell of a sales pitch. After all, it was this game he returned to his studio to make, after having walked away in 2004 following the rushed release of Driver 3. But Driver: San Francisco wasn't flawless. And so Eurogamer sat down with Edmonson for a post-mortem, not only of Driver: San Francisco, but of the series as a whole.

Eurogamer: How significant is the release of Driver: San Francisco?

Martin Edmonson: Driver was a massive franchise back in the late '90s with Driver 1 and Driver 2, and then suffered its various issues in the last two iterations. This was a big chance to come back and do something that was different and meaningful and quality. From that point of view it has been really important for the studio. I wasn't there the whole time towards the end, but you can imagine it's not great to be working on a project where your publisher has no money and they want the game out within nine months.

Eurogamer: What happened with Driver 3?

Martin Edmonson: A number of things. We had come up with something that was totally original. The first Driver game was the first open city-ish driving game. And then Driver 2 was the first one to feature getting out of the car and running around. Some people think Grand Theft Auto was the first to do that, but it was actually Driver 2.

We had this rolling ball of innovation, and then Driver 3, we were guilty of falling into a rut of continuing in that vein when GTA had already gone very strongly in that direction: out of the car and with guns and so on. The basic problem with it was that we had not properly finished the out of car sections.

Despite the media-savaging it got, not everybody thought it was poor. That reflected badly on the other bits in car, and actually we were really happy with the car sections of the game - they looked beautiful, the lighting was nice, the cut-scenes were great, the music was great, the cars, the crashes, the physics and so on. But the out of car stuff was just not finished and had been bettered by other games.

Then the publisher situation and release dates: you get to the end of financial [year] and you have absolutely no choice but to put a game out, whether it's finished or not, and that's a difficult situation to be in as well. That's not to say it was entirely the publisher's fault, because we were also running behind on certain aspects of the game.

Eurogamer: Was putting out an unfinished game why you chose to leave Reflections?

Martin Edmonson: That would also be a bit unfair. My contract had also run out. It wasn't a case that I got so upset with them that I decided to leave. But had I been extremely happy then, you know, we would have presumably carried on.

Video: Eurogamer video talkthroughs Driver: San Francisco.

Eurogamer: Why did you return?

Martin Edmonson: There are several aspects. Atari had got itself into a financial situation where it needed to or chose to sell the brand and Reflections. It had been sold to Ubisoft, which was very committed to the idea of bringing back the game in a big way. They weren't going to just take the name and bang something out in a year and take advantage of what we had. [Ubisoft] wanted to form a big team, long development schedule, technologically be on the cutting edge. That, to me, was very exciting, the idea of really brining it back, and that is part of what drove me to come back in.

From a personal point of view you have a studio that I set up, co-founded, and my brother had been promoted to the managing director and I knew lots of people in the team anyway. It's my hometown where I was raised. All of these things collided.

Eurogamer: How do you feel about Driver: San Francisco's reception?

Martin Edmonson: What we are happy about is that we're, so far, very consistently getting bang, eight; bang, nine; bang, eight; bang eight. Now this is not something we're used to! What we're used to is polarised, love it or hate it. And so far the reviews are coming in pretty positively, so we're obviously happy about that.

What I'm pleased about is that the reviews are recognising what's different and interesting and innovative about the game. Remember this comes from a position of having a feature [Shift] which is very difficult to sell on paper and describe, unless you play the game. And when you play the game - I'm not talking about the demo - it's nice to see how many more people get it and enjoy it.

Eurogamer: The demo seemed to put people off. Did that surprise you?

Martin Edmonson: It's difficult. This game is not the sort where you can easily present one mission to someone. This game is about what happens to Tanner and how he gets himself into that situation. Without giving away half of the game you just can't [convey] that in a demo.

And Shift itself is something that, again, doesn't come from nowhere. You have to play through the story to understand where it came from, why he has it, what its power is, how it becomes stronger, why it becomes stronger, what happens in real-life that is affecting his coma, what he's thinking about in his coma that influences what happens when he comes out of his coma. We just can't think of any way of doing that easily - here's a bad guy, here's a good guy, have a chase, have a crash and that's it. It's got more to it than that. We always knew it was a very tricky one. We tried to choose missions that were giving a little bit of a taste of it.

The multiplayer demo was a lot easier for us as a proposition, because you know what the rules are; there's Shift, it's Tag mode, get on with it. Story doesn't come into it.

Eurogamer: Is there anything that has been misunderstood in Driver: San Francisco?

Martin Edmonson: Most people that play the game think the actual function of Shift is good fun and bringing something different. There are reviews that get the story, they understand it, they see how Shift came and how it develops. They also recognise that we're showing Tanner's schizophrenia deliberately. One minute he's doing a police investigation and the next minute doing something completely crazy with a bunch of Japanese street racers or a learner driver.

And this is by design - it's an intentional thing. One or two of the reviews just go, "Oh this story is completely ludicrous." You need to stop for a minute and think why have we done it this way. This is why: because he thinks he's having a mental breakdown. I don't think they've thought about it carefully enough. What we have done is, very deliberately, treat it with a light touch. Jones, for example, his partner, is constantly taking the piss out of him. And this is because it's such a ludicrous situation. All the time we're being very, very deliberate about this. Tanner has to basically convince Jones, and he starts to think very quickly that things are not quite as they seem.

There are matters of taste: handling, for example. The way I view handling is, first of all, handling is very much back to the roots of the first game, that physics-based handling model. The disadvantage is that it can be trickier to do simple things like get around a bend by just steering. You have to do things like use the handbrake, you have to get the back-end out, you have to be drifting, you have to temper your throttle.

You can't be on the throttle 100 per cent of the time like you can in some video games. That can be tricky for the first few minutes until you get the hang of it. The advantages are that you eventually get used to the subtleties of how it behaves and it allows you to nail this perfect drift around a bend. That, for me, is far more rewarding than the handling model that has a button for drift sort of thing. It also means you can exploit the handling as you get used to it. There aren't too many video games that go into that sort of depth.

Video:

Eurogamer: Did the mechanics Boost and Ram need to be there?

Martin Edmonson: The reason why we put that in is that Tanner is in a strange state of mind. If you imagine you're driving a car and you have this urge and want it to go faster, be stronger... This is the thing that has been lost in some reviews, that it is not nitrous, not a nitrous boost. This is Tanner concentrating. It's like a kid in a sweetie shop kind of thing: if I could do anything with a car, what could I do? And that's one of the things people would love to do with a car. And you'll notice that he doesn't have boost before he has the big crash. You'll also notice that any time he's not in a coma he doesn't have boost.

Eurogamer: Was San Francisco under-used as a setting?

Martin Edmonson: I'm not sure I agree with that. We've got Downtown, we've got Russian Hills, we've got Marin County, Golden Gate Park, Sutro - we've picked the juiciest areas. If you look at a map of hot-spots, heat-spots, of where all the missions are, they are spread across the entire city in a variety of environments. There are not too many driving games that have big, off-road dirt tracks, which we do, with big hills for the jumps; massive, sweeping, almost F1 circuit courses.

That's the one thing that when I read in a review I can't understand where they're coming from. But that is probably the only one where I have nothing to say other than I just don't understand what they mean.

Eurogamer: Why did you use the sepia-like haze effect to the degree you did?

Martin Edmonson: You'll see that that's something that only happens when he's in the coma. When he goes into a coma there's a different treatment, and that treatment is to slightly affect the lighting. Some of the more intense missions have that effect that is more obvious.

I did read somewhere that it was some way of hiding the scenery, which I thought was terribly unfair, because if you're at the top of Russian Hill or Twin Peaks, you can see the whole city in its entirety - there's no fogging. We draw the whole world, and that has been one of the things we're most proud of, that we have this enormous city running in 60 frames-per-second and we do not occlude - we don't hide stuff.

That was one of the mandates at the beginning that we put to ourselves: it's going to be San Francisco, it's going to be at the tops of hills looking right across the Bay to the other side of the City, right over to Marin, many, many miles away. And we're not going to fog it, we're going to draw the world to its absolute extremity. That's the kind of thing that fills technical guys with utter horror. That's what we said to them and that's what they delivered.

The other thing to say is that it was a stylistic choice. The game is heavily inspired by '70s movie car chases. You'll see an element of film grain, you'll see that sepia, slightly aged film look, and it's entirely intentional. It's about the look of movies at that time. If you go and watch a car chase from the original Gone in 60 Seconds or French Connection or The Driver or Vanishing Point, they have that genuinely old, aged, celluloid feel to them, and that's part of what we wanted to do.

Eurogamer: Is there anything you would change in Driver: San Francisco - is there anything you left out you would have rather put in?

Martin Edmonson: Not many things. We were lucky enough to be granted an extra serious lump of time on the project, because originally it was supposed to come out in September last year, and we were coming up with some really cool ideas like split-screen, the Movie Challenges - ideas that came in late. We were so adamant that we wanted to include these that they were either going to be some kind of DLC, or we extend the length of the project and include that stuff. We got a lot of the stuff that would have dropped by the wayside into the final game.

In terms of things we still didn't manage to get in: obviously there are always going to be things. A few more functions in the Film Director would have been nice; there were a couple of movies we wanted to include in the Movie Challenges but we ran out of time; and we wanted to be able to do things like changing the car's colour in the garage when you take the car out. It sounds like a trivial thing but it's still going to take a week to do it.

Eurogamer: Is there post-release DLC planned for Driver: San Francisco?

Martin Edmonson: That's a question for Ubisoft. It absolutely would have been, but because we got the extra year of development time, everything that was planned for DLC - the serious, meaty stuff - we put into the full game package. DLC was getting bad press at the time for stuff actually being on the disc and then being unlocked - that's the sort of thing that really irritates people. We went the whole hog and put all of this in the game.

Eurogamer: How much is riding on the success of Driver: San Francisco - if it fails to hit is sales target will we see Driver again?

Martin Edmonson: For a sequel, for other games to come out, then a game upon which it's going to launch a sequel has to be successful. Yes of course the game has to sell well. But there are other ways of brands continuing. If you think about the end of Driver 3, Driver 4, a lot of people said that's the end of Driver, you'll never see another Driver again. And yet here we are with not just another quick version of Driver, but a huge project and, in terms of time and investment, the biggest Driver yet.

Eurogamer: Would you stick with the Shift mechanic, or is its appearance in Driver: San Francisco a one-off?

Martin Edmonson: It's too early to say. It's a feature, a function, that has an enormous amount of development time, effort, energy, tech, behind it. It would be a shame not to use it again. What we found is when we play some previous Driver games you find yourself wanting Shift. That's always a good sign.

Eurogamer: What you clearly couldn't do is have Tanner have another accident and go back into a coma again.

Martin Edmonson: You can imagine how difficult this was as a proposition early on, and it's all credit to Ubisoft that they are willing to experiment with things like this. I can imagine showing that, as a concept, to some other publishers, that they would be like, "What the hell?! This is far too risky." One of the great things about Ubisoft is that they are happy to try this different stuff. [The coma] did allow us to have great creative freedom with the types of missions, with Tanner's schizophrenia, and to tie it all up at the end and make sense of it.

Eurogamer: And are you done with on-foot sections after Driver 3?

Martin Edmonson: No, it definitely doesn't mean that. But it also doesn't mean that we're now planning on having Tanner getting out of the car. I had no objection to the concept of Tanner getting out of the car or firing guns or any of that on-foot stuff. It's not to say we couldn't do it, it's not a lack of talent within the team or anything. It's just time and energy and focus. And if you focus on two things, you're not going to be as focused on one of the two things.

Eurogamer: Does having the new engine mean a new Driver game would take less time to make?

Martin Edmonson: Using the same engine, yeah. This is an engine that we built from the ground up - proprietary tech. We had to do that because we were doing some crazy stuff with the tech. An enormous amount of development is taken with building proprietary tech that you could have otherwise just taken off the shelf in no time. Now that tech is built and we own that tech, it is entirely possible to use it to make other games. But it was very much tailored to that game.