It’s Super Bowl week, which means the sporting news cycle is full of the same old shit it is every year. Empty interviews, retrospectives on Super Bowls past, endless predictions on who is going to win on Sunday…and one prediction in particular, which comes around every season like clockwork.
I get it, the Super Bowl is the biggest game of the season, and pundits have two whole weeks of anticipation to build up to, so trying to work out which one of two teams will emerge victorious is an easy ratings pull/pageview generator. Most of those calls are made by men and women making some calculations in their own head, though; only Madden tries to do it via the art of simulation.
As they have done since Super Bowl XXXVIII in 2003, EA Sports has run the teams and rosters for the big game through its system, played out a game and got a result. That result was a 31-17 win for the Eagles over the Chiefs.
Before you go taking that to the bank, though, either because you’re an Eagles fan or you have a devout belief in the sanctity of Madden’s player ratings and matchday authenticity, you should know that for all the attention EA Sports likes to lavish on these simulations, and for all the attention paid to them—this year’s prediction has its own trailer—the game’s track record isn’t much better than flipping a coin.
Since that first public attempt back in 2003, Madden has successfully predicted the winner of the game 11 times, meaning it has got it wrong on eight occasions (you can check out a full rundown of each score here). That 11-8 record is over .500, you might say, enough to at least make the playoffs and surely more accurate than most flesh-based pundits. Especially when you consider the game’s Super Bowl XLIX prediction was very accurate, nailing not just the winner but the exact score as well (NE 28 – 24 SEA).
But in recent years the wheels have fallen off! For the first decade of its simulations Madden nearly always got the winner right. For four of the last five years, though, and six of the last nine, its predictions have landed in the L column, meaning it simply cannot be trusted to be predicting the winner on Sunday with any more accuracy than tossing a coin, rolling some dice or asking an octopus to crawl towards a flag (actually, to be fair Paul was very good at his job).
All of which is to say, based on this 31-17 prediction I look forward to seeing the Chiefs win 24-10.