They started with five cars, going as far as to take them apart to measure the several hundred points of research data, from simple things like weight distribution to measuring inertia on the driveline. Over the years more cars have been quantified, some to a greater degree than others. This has built up a database that provides the starting ground for each of the cars in the game.
Turn 10 doesn't use any data submitted by the manufacturers, who might not be above sending them a ringer. But their system is surprisingly accurate. "Even though we don't have the real numbers, we've actually been able to find errors in cars through this simulation," Greenawalt claims. All this work results in around 9,000 parameters for each car that are used by the game's physics engine to simulate everything from a tiny European hatchback to a Le Mans winning prototype racing car.
For Forza Motorsport 4, the team has thrown away all its previous tire data and started from scratch. Forza Motorsport 3 had data from Toyo, Michelin, Pirelli, and FAE, and it had to be manipulated so that it would all work together, despite being gathered at different testing facilities and with different testing methods. This time the game is built with new data: custom tests from Pirelli across a huge range of variables (tire height, width, compound, temperature, pressure, camber angle, etc) that was imported directly into the game. "That was the only way we say to be as up-to-date as possible. We weren't copying the textbooks of racecar engineers from a couple of years ago but putting in the data that would write the textbooks of the coming years," Greenawalt explains.
