From Gruden to Belichick: The AI of John Madden Football David Kent March 16, 2004 STS 145A: The History of Computer Game Design"They knew where we were going, what we were doing...They were in our playbook." -Oakland guard Frank Middleton, on the Tampa Bay Buccaneers’ defense (AP) "[Belichick's] game plans change drastically from week to week and are particularly effective at exploiting mismatches in pass protection and coverage. He is known to help design a defensive strategy for any given week and then disguise it with a new package of blitzes and stunts." -The Houston Chronicle on New England Patriot's coach Bill Belichick.(Duarte) Super Bowl XXXVII pitted the NFL's #1 rated offense, the Oakland Raiders, against its #1 rated defense, the Tampa Bay Buccaneers. In the end, defense prevailed, with the Buccaneers intercepting Oakland quarterback and regular season MVP Rich Gannon a super bowl record five times. After the game, many of the Oakland players commented on how it seemed the Tampa defense knew what play they were going to run before they hiked the ball. Tampa’s coach, Jon Gruden, was the coach of the Raiders the year before, and thus had an extensive knowledge of their offense. Tampa safety John Lynch even stated after the game, "Every play they ran, we've run in practice."(AP) The next year, however, a new defense would storm through the NFL playoffs, disrupting playoff record-setting quarterback Peyton Manning and winning the super bowl in an entirely different way. While Tampa's defense was certainly good, their knowledge of their opponent’s playbook played a large role in the dominance they displayed in the super bowl. New England, on the other hand, relied on complicated defense designed by their coach, Bill Belichick. These two varied approaches provide an excellent illustration of a dichotomy that exists in gaming AI; while some developers allow the computer opponent to “cheat” by using information and movements unavailable to the player, others try their best to createa truly intelligent opponent. The use of “cheating” AI, though it is present in nearly every computer game, has been somewhat controversial. At the Game Developer’s Conference (GDC) in 1998, Stephen Woodcock, the moderator for the artificial intelligence roundtable, noted that “The foremost topic on the minds of most of the people present was that of building a ‘cheating AI.’” He states that the general consensus they reached was that this type of cheating should be avoided at all costs. In another discussion, however, it was mentioned that “a vocal minority of developers felt the move towards developing better strategic AIs was primarily a waste of time.” (Woodcock 1) These developers felt that “well-placed but thoroughly plausible unit placements (via judicious cheating on the part of the AI) would go a long way towards providing the player with an enjoyable gaming experience.” (Woodcock 1) The presence of “cheating” in game AI has not been the only issue discussed at these conferences. Many developers at the 1998 GDC felt that “too many games are still being developed without much consideration to the implementation of the computer opponents until very late in the production cycle.” (Woodcock 2) Artificial intelligence and its implementation has been an issue that has come to the forefront, especially with the recent improvements to graphics hardware have taken more and more of the visual processing load off of the CPU. However, while many game developers have looked at the progression in AI made by others in strategy and first-person shootersi, or even in real-world applications such as robots, one genre has been entirely ignored: the sports game. Sports games have, on the whole, been ignored by many in the gaming industry. Ernest Adams states that “sports games don’t get the respect from the developer community that they deserve.” He elaborates by saying that “With most other kinds ofgames, if…the AI isn’t smart enough to beat the player, it’s possible to change the game itself to compensate for these problems. With sports games you can’t do that.” One title in particular stands out among sports games for the level of its artificial intelligence. EA Sports' Madden series has made a few significant innovations in the field of artificial intelligence that other games could certainly learn from. A good balance involving customizability and a limited amount of cheating AI has enabled the Madden series to please its community with a challenging, but not altogether impossible "hardest" difficulty level. By examining the development of the Madden series’ AI, other genres could learn a great deal from how EA has learned from its mistakes by listening to its community in the development of the game's artificial intelligenceii. In looking at the development of Madden’s artificial intelligence, we will first examine the development of the Madden series on the whole, and then look at the improvements in artificial intelligence over the years. Finally, it is important to take a look at the community's view of the "cheating" present in Madden's artificial intelligence and listen to EA's response concerning how much of this "cheating" is actually present, and how much is myth. The Development of John Madden Football In 1986 Electronic Arts decided that it wanted to produce a football game. However, Trip Hawkins, one of the founders of EA knew that they needed some guidance from a football guru. Even in 1986, John Madden was known as one of the preeminent football commentators. The story of how Trip was able to get in contact withJohn Madden varies depending on which source is used, with some indicating they met on an Amtrak train (Todd), while others insist that Trip had to track down John Madden in a parking lot as he left work (Holmes). Regardless of how they met, Madden agreed to help EA develop its football game on one condition: it had to be 11 on 11 football. "'If it's not 11-on-11, it's not real football,' Madden said."(McClusky) EA had planned to release a football game that was seven on seven, but Madden simply stated that he wouldn't put his name on a product that wasn't real football. His inflexibility would serve as the impetus for EA Sports' later slogan "If it's in the game, it's in the game", denoting its attention to detail and realism within the sports genre.
View Full Document
Unlocking...