| |
| by Danny Internets | January 26th, 2009 - 7:54 pm
|
|
| While most serious gamers look down on the NFL Madden franchise with disdain, it’s hard to ignore the game’s level of realism. Not only does it look true to life, but apparently its assessment of team rosters and individual athletes has significant predictive power. Beginning in 2004, EA Games has annually used the latest NFL Madden game to simulate the Superbowl matchup after the post-season outcome has been decided. Throughout the five annual simulations, the game has successfully predicted all but one of the winners (and let’s be honest, no one really expected the NY Giants to take home the trophy last year). |
|
| Year |
Actual winner |
Predicted winner |
Actual score |
Predicted score |
| 2004 |
Patriots |
Patriots |
32-29 |
23-20 |
| 2005 |
Patriots |
Patriots |
24-21 |
47-31 |
| 2006 |
Steelers |
Steelers |
21-10 |
24-19 |
| 2007 |
Colts |
Colts |
29-17 |
38-27 |
| 2008 |
Giants |
Patriots |
17-14 |
38-30 |
| 2009 |
? |
Steelers |
? |
28-24 |
|
While it’s still entirely plausible that the string of successful guesses is due to chance, the potential accuracy of NFL Madden to predict real game outcomes based solely on statistics raises interesting questions about the importance of coaching strategy and team matchups in professional sports. Are the leadership roles and managerial functions of coaches and quarterbacks simplistic enough to be carried out by rudimentary AI? The jock in me grunts “no!”, but the nerd in me squeals “yes!”
So who will it be this year, virtual John Madden?
BOOM–Steelers, 28-24.
Similar Posts:
Loading ...
Leave a Reply