The making of GRID Autosport

Making a modern racing video game is incredibly difficult. Quite apart from the challenges presented by ceaselessly evolving hardware and the enormous task of coding a big-budget, multi-platform title, the audience you are working so hard to please is a disparate and ruthlessly discerning animal.

Make your product too simple and 'arcadey' and you switch off the purists; focus too much on accuracy and you dismiss all but a small section of talented players. Either way it's all too easy to dismiss the thousands of man hours that go into it, simply because the handling doesn't suit your taste.

A developer must weave a fine thread between accessibility for all levels and the true essence of the sport it is trying to depict - all the while maintaining enough of that 'just-another-go' gaming glue to ensure the product is a success. By these measurements games companies live or die.

The original TOCA games set a benchmark

Codemasters has a strong heritage in racing games dating back to the hugely popular TOCA series that dominated the freetime of Sony PlayStation players in the 1990s. Its current racing series, GRID - which evolved through TOCA and later Race Driver - is now about to be released in its third iteration.

The new title, GRID Autosport, is intended to bring the series back to those halcyon days of the TOCA Race Driver Series, and that's where we came in.

While the initial GRID title was a hit, comfortably sitting between game and sim, it delivered entertainment if not a true racing representation when it was released across PC, Xbox 360 and PS3 in 2007. Five years later, the sequel missed the mark. It went too mainstream and turned off the racing audience loyal since those TOCA days.

So when Codemasters came to AUTOSPORT saying they wanted to take the game back to its roots and asked for our journalists to give input on the handling model and access to our address book of racing drivers, we couldn't help but say yes.

After all, we all know the kind of racing game we'd want to make. We'd never done anything like this before and it was a chance to find out for ourselves just how tricky it is to get it right... Three and a half days of retina-searing intensive testing later, we'd all vowed to be a little less critical of games in the future.

THE PRICE OF DEVELOPMENT

Huge effort goes into getting tracks right

Since those heady days of TOCA Touring Cars - the vast sales of a game based solely on the British Touring Car Championship being hailed a massive success - the value placed on promotional licenses has amplified significantly, as has the technology.

The headcount required to create modern gaming software runs into the hundreds. Now the risks of getting it wrong are real and potentially fatal for an independent software house. Just to cover your bets, you need to sell more like a million units just to make a profit.

The statistics below give a sense of the scale of a modern game.

Development team headcount: 249*
Number of man hours: 261,307
Man hours to make a track: 5312
Lines of code: 3,708,453
Circuit development time: Two years per track

*excludes recording studios, localisation agencies, cast, etc.

EXPERT VIEW
Graham Bromley
GRID Autosport Lead Level Designer

We go to the locations and fully research, photograph and collect data from the track. We often collect over 10,000 images showing every detail in the track surface: the size, shape and appearance of every trackside object, rumble strip, kerbstone and drain, and any other features. From these the art team creates several hundred textures, and the hundreds of assets and buildings that make up that location.

The new game is designed with the motorsport fan in mind

In the meantime, the level design team will take the data and images collected, along with the first-hand knowledge, and start creating what we call a 'whitebox track' - this is a working, playable prototype of the location.

Once the whitebox is ready, the art team applies the photorealistic textures and assets that they've created to turn our functional prototype into fantastic looking environments.

It's also great to have been able to work with so many racing drivers who know the tracks in incredible detail, as they can help us make sure that we haven't lost anything important.

So, taking an existing venue, it would take one designer 30 days to prepare the track from data to whitebox standard. It then takes five artists 350 days to get the track looking photorealistic, followed by another three designers needing a further 50 days to fettle the details.

AUTOSPORT'S TESTING INPUT

Glenn Freeman
It seemed easy at first. We tried two types of BMW M3 handling model on a street circuit and had to give feedback on the pros and cons of each.

There was a stark difference, and one was really nice to drive, so naturally that got a lot of positive feedback.

AUTOSPORT's team of test drivers got a sense of how much goes in to a racing game

"Simple," I thought, assuming that it would always be this clear-cut which handling model felt best, and that there would always be a setting for each car that suited me.

But as we tried other cars through the day it made you realise just how gruelling this process can be, and how switched-on those doing the test driving have to be to make sure that the game developers are getting useful, accurate and consistent feedback.

Much like going testing in real motorsport, if the driver's head is not in the right place then his team is not going to get much out of the day.

Charles Bradley
Jarama is a classic old grand prix venue that I've been to loads - it was the first place I ever flew to on a plane!

Whether it was Formula Renault Finals, World Series by Nissan, European Super Touring, FIA GT or Superleague Formula, I knew the place like the back of my hand without ever having driven around it.

Now was my chance, as one of my tasks was to feedback on a Caterham SP/300.R, which was well-suited to this twisty track.

Of course, the focus was on giving feedback on the two very different handling models I was back-to-back testing, but the attention to detail of trackside aesthetics really put you there. It was also impossible to not notice the evil bump on the exit of the Portago right-hander that always tried to spit me into a spin...

I thought it was brilliant, and can't wait to try a new track I've never been to: Mont-Tremblant!