Crash Bang Wallop

Pseudo's was always going to be tricky task. Entering the arcade driving genre nowadays is a daunting prospect; trying to introduce a potential genre-bender like unwreck, its time-rewind technique, didn't exactly simplify matters. What's more, the idea was to do all this in the traditionally difficult 'driving shooter' area. Bit of an ambitious mix. "We've done much to ensure that they all work in harmony, although we'll be the first to admit that perfect balance is a difficult thing to achieve in a brand-new IP," says producer Smith candidly.

The XNA thing was a bit of a break though. "The Crash demos allowed the studio to prototype both realistic car modelling as well as realistic procedural destruction," he says. It served as a springboard. A year after the second Crash demo aired at E3 2004, Pseudo was back with a fully working two-level 360 playable demo - one of the only publicly playable 360 games at the show that marked the console's actual unveiling. In it, players raced around a city smashing through shopfronts, street furniture, other cars, even walls, firing rockets and machineguns and the like at rivals and using unwreck to undo mistakes. We were among those impressed.

"We had unwreck working in prototype form way back in August of 2004. Even back then - with no visual effects and sparsely populated game worlds - it showed great promise.

"We entered full production in late October of 2004, but initially we were building tracks from a very layered approach. We were also busy prototyping our gameplay mechanics and attempting to understand just how far we'd be able to take our vision given the new hardware," says Smith of the development process.

Indeed, whether they've taken it too far is actually a point of contention following previews like ours that pointed to a bit of slowdown, and performance issues with the demo that appeared on Live Marketplace on January 25th. Smith says not to worry. "During the final phases of development, we made every effort to utilise all three cores through threaded rendering and other techniques.

"I believe you'll notice a very large improvement in overall performance, although we plan to push the limits much further now that the hardware is finalised and we have time to re-engineer our technology to maximise the benefits within each area of our codebase." Which is the developer for "it'll be quicker out of the box, guv".

He agrees that games like Burnout and Twisted Metal formed part of the inspiration for the game, but suggests it went wider. "I suppose Full Auto is still a simmering stew of ideas and inspirations," he says. "We planned to hybridise what a number of racing games had established in the past."

And hybridise, apologies to the Queen's English, it certainly did. As well as arcade handling, a range of weapons and an impressive time-rewind feature, Full Auto also presents an enormously destructible environment.

"For all of our modes we had to be very thoughtful about retaining a proper raceline and not trading the intensity of attaining first place into boring free-roam exploration," Smith says of the carnage. "With multiplayer, our biggest constraint was in perfectly duplicating and transferring each and every piece of debris to all clients. We ensure the most relevant objects and events sync directly, but ignore the inconsequential for the sake of performance." Which is presumably why we couldn't get the 360 debug to "see" Kristan in a competitive context last week.

But yes, multiplayer. How come there's no unwreck there? "It was more of a question of game design than just tech. As we played more and more multiplayer, we decided that [unwreck] was not core to Full Auto multiplayer. Focus was then turned to balancing the core elements and getting a fair and fun experience." That said, "you shouldn't count out the possibility of it in the future". Hrm.

Ultimately, it does all come back to that core playability. As good as it all sounds on paper, people are only going to spend their money if it's as good as or better than Burnout, Ridge Racer, Gotham, Need For Speed, etc. There's lots of competition. Settling on a handling model in a field of such diversity and refinement sounds like it posed real difficulties.

"Our real concern with handling came down to being mindful that we place players in a dynamic world where they not only have to drive fast but also aim and shoot," says Smith.

"Although we've experimented with a much more complex and 'touchy' driving model to draw out the personality and challenge of each of our vehicles, we needed the control to be more responsive and accessible to stay in line with our 'arcade' sensibilities.

"As for other arcade racing games, without need to concern themselves with environmental destruction or weapon-based combat, they can obviously offer up a greater sensation of speed along linear tracks. We don't begrudge them that, and we're all huge fans of the top contenders, but it should be publicised that our game has different tendencies and therefore different needs." Consider it publicised.

As for those who had a bit of a pop at the demo last week, Smith insists that the game's deceptively deep. "Witnessing people play it for the first time is riveting because almost everyone has fun from the very start. That said, there's no lack of depth to how players combine their racing, unwreck, and combat skills. It takes mastery of each and an understanding of the game's subtleties to truly become a pro.

"Speaking of subtleties, we're also extremely proud of just how much we've packed into the world. Pay close attention within the one-touch replay feature and you'll continuously notice small elements and interactions that you've never realised before. As late in development as the day before final submission, I was still witnessing unique moments that I didn't know were possible."

We'll let you know whether they're magic moments closer to the game's European release on February 10th. And if we're wrong, you can shoot us in the face with a missile launcher, presumably.