The first season of something always has some problems as the quirks are worked out. The 2004 X-Factor football season was no exception.
Though our virtual Jayhawks finished with a record (5-6) just one game better than the real-life football team (4-7), the way our video-game players reached that mark was often far from similar to the manner the flesh-and-bone players finished three games under .500.
Had we trusted the original simulation against Northwestern, the X-Factor record would have perfectly matched KU's real 4-7 mark. The original prediction was a 42-40 Wildcat victory, but we tried again because we didn't think Northwestern had a chance of winning the game. Sorry. Our bad.
The X-Factor simulations were correct in picking KU's victories in the first - Tulsa - and last - Missouri - games of the season, though the virtual scores were a little higher than the real ones.
The other games that our simulation came closest to were losses to Colorado and Texas - the two games before the season finale against Missouri - thanks to loyal reader and KU student Justin and his suggested tweaks to the game's settings.
The X-Factor picked the score of the CU game within five points, and it correctly predicted a last-minute Longhorn victory - though it was a virtual field goal as opposed to a real-life touchdown pass.
Iowa State and Oklahoma were the only other two games in which the X-Factor predicted the correct result - both losses in real life and the virtual one. Our simulation also correctly predicted that OU quarterback Jason White would have a big game against the Jayhawks, though the virtual White had a game that was just a bit better than his real-life counterpart.
Our simulation was only way off on two games - KU's victories over Toledo and Kansas State. The X-Factor simulation said KU would lose the Toldeo game by nearly three touchdowns, while the real-life game ended up a 49-point KU victory.
We blame the difference in the K-State game on the X-Factor not being able to adequately predict either the Jayhawks' hunger to defeat the Wildcats for the first time in more than a decade or Mark Mangino's ability to outcoach Bill Snyder.
Had the Jayhawks held on to second-half leads against Northwestern and Texas Tech, their real-life record would have turned out a game better than the X-Factor's 5-6 mark. We'll count those as half-victories when it comes to how realistic our simulation turned out.
When it comes to exact plays, the X-Factor came pretty close to a couple real-life examples:
Unlike the real Jayhawks, our virtual KU football team didn't suffer any major injuries in games. When real players missed games because of injury, though, we did substitute them in the simulations, and sometimes - like Brian Luke against Missouri - the virtual back-ups performed at nearly the same unexpected level as the real ones.
All in all, we're pretty happy with the final results of our first X-Factor season. Sure, there were some big differences between the virtual and real seasons, but we expected as much in our first attempt to predict KU's football future. And the problems we encountered this year will only help us do an even better job next year.