Super Bowl Prediction

After correctly predicting the winners of the last four Super Bowls (which is as long as the site has been simulating NFL games) and, coming off of a weekend in which it predicted a (26-23) victory by the New York Giants and a (24-17) win by the New England Patriots in the Conference Championship games, sports simulation website WhatIfSports.com likes the Patriots to defeat the Giants by an average score of 30-16. Included in this feature are an explanation of the analysis, a representative boxscore with play-by-play, a brief game story related to the boxscore, stats, stats and more stats from the simulations. If that is not enough, anyone can simulate this game (or the 1972 Dolphins vs. 2007 Patriots, etc.) for free by using the SimMatchup feature.

WhatIfSports used its award-winning SimLeague Football simulation engine to "play" the Super Bowl 10,001 times. For each game, we make use of as many strength-of-schedule-adjusted statistics from the regular season and playoffs as we can find to determine the likelihood and potential outcome of each play. With 10,001 games looking at every play and ever scenario, this is the most thorough way to analyze any matchup. With the Super Bowl, we are pulling out all the stops, using the results of those simulations to come up with more data than just about anyone could ever use (below). In short though, New England wins an impressive 77.2% of those 10,001 games. The average score is 30-16, while the most common score is 31-21 Patriots. What follows is just one example of the play-by-play of the Super Bowl from those 10,001 games that just so happened to have the exact same score as the average game.

The Game: Super Bowl XLII
New England 30 vs. New York 16 - The Patriots complete their perfect season in almost perfect fashion. New England's offense rolls, while its defense shuts out the Giants for three quarters. In fact, the only score that New York can muster before New England has a 28 point lead is a second quarter safety with 1:53 left in the half when Kawika Mitchell gets a sack on Tom Brady in the end zone, just Mitchell's fourth sack on the season. The score was still 10-2 at that time with the Pats' previous TD coming on a Randy Moss 23-yard reception eight and a half minutes earlier. Getting the ball back in the final two minutes of the half and trailing by just eight and with two timeouts, New York misses a "giant" opportunity. A three-and-out forces a punt with 1:06 remaining and 69 yards for Brady and Patriots to go. It only takes 43 seconds before Brady finds Jabar Gaffney for a four yard score. What could have been 10-9 at halftime, becomes 17-2. The Patriots are never tested again, as the Giants merely score in the final quarter with the game out of reach. After succeeding with the running game for the last two weeks, New England's lone weak spot is on the ground. Laurence Maroney gets just 58 yards on 25 carries. But, unlike what the Packers were able to do last game, Brady throws the ball effectively and the Patriots stick to the run when the situations call for it. For the game, Brady completes 62.5% of his passes for 234 yards and three touchdowns with no interceptions. 99 yards and a touchdown go to Randy Moss, who had been relatively quiet in the playoffs thus far. Brady's counterpart, Eli Manning does fairly well, yet throws his first two interceptions of the post-season in the loss. Lawrence Tynes is only heard from on two late, insignificant extra points. All season long, the Patriots always had more weapons than their opponents on both sides of the ball. If anyone performed poorly, the rest of the team could pick up his slack and find a way to win. Fittingly, New England must overcome obstacles, like an utimely safety and a lack of a running game, yet everything still appears to click in the dominant performance. (And, just for fun - 1972 Dolphins vs. 2007 Patriots.)

Super Bowl MVP - Tom Brady - 20-32 for 234, 3 TDs, 0 INTs

Team Stats: Results from 10,001 Super Bowl Simulations

Before we get too deeply into the stats, please note these few items. The difference between the average score and most common score is significant. As you can see, even though the Patriots win by 13.2 "on average," the Giants lose by less than 12 more often than not. The actual likelihood of either team winning the toss is (and should always be) 50%; however, over the course of the simulations, the Giants won the toss a few more times than the Patriots. L2MinH signifies the likelihood that the team scores in the last two minutes of the first half. L2MinG is the same thing for the end of the game. And yes, we probably spent too much time coming up with a mathematical model to predict the length of Jordan Sparks' performance of the National Anthem.

 Average Scores After Each Quarter

 Most Common Scores After Each Quarter

 Miscellaneous Team Percentages

Game Stats: Results from 10,001 Super Bowl Simulations

 Average Game Results

Player Stats: Results from 10,001 Super Bowl Simulations

 Average QB Results

 Average RB Results

 Average WR/TE Results

 MVP Likelihood

We hope that you enjoy our Beyond the Boxscore feature. As with all of our free features, games and content we are willing to listen to feedback and improve this page and content to make it as interesting and user-friendly as possible. With any comments, questions or topic suggestions, please contact Customer Support. Thanks!