The Super Bowl Prediction System

January 26, 2013

It has been six years since a Super Bowl failed to feature at least one elite passing offense. Since Super Bowl XL between the Steelers and Seahawks, twice we’ve seen Manning’s Colts, twice we’ve seen Brady’s Patriots, once we’ve seen Warner’s Cardinals, once we’ve seen Brees’ Saints, and once we’ve seen Rodgers’ Packers. This year, we get a matchup between a pair of defensive-minded teams, the 49ers and the Ravens, that each earned their ticket by eliminating one or more elite passing teams.

Hopefully, this throwback game will be just what the Super Bowl Prediction System needs to get back to its winning ways. The system has made the incorrect prediction the last two years and four of the last six. However, the system continues to have a positive overall record of 28-14 (with four non-selections) and is 16 of the last 22, which is 73 percent.

The Super Bowl Prediction System tends to favor defense over offense, in general, so you might expect the Ravens and 49ers to be evenly matched. The oddsmakers share that sentiment, having the 49ers favored by a small margin, a little bit more than a field goal at -3.5. However, the prediction system indicators show the 49ers to be overwhelming favorites:

Category Win% Team with Advantage
Points Scored .543 Ravens
Points Allowed .630 49ers
Point Differential .630 49ers
Fewer Net Passing Yards .609 49ers
Rushing Yards .543 49ers
Rushing Yards/Carry .565 49ers
Opponent Net Passing Yards .565 49ers
Opponent Rushing Yards .609 49ers
Opponent Rushing Yards/Carry .587 49ers
Opponent Total Yards/Game .652 49ers
Turnover Differential .587 Tie
Regular Season Record .543 49ers


We use 12 indicators to predict the winner. As you can see from the Win Percentage column, each one of these indicators predicts the winner more than half the time. But taken together, the Super Bowl Prediction System is even more successful. The 49ers have the edge in all but two of the indicators. The lone Ravens advantage is in points scored during the regular season, and they only beat the 49ers by a single point, 398-397. The 10 categories the 49ers won were mostly convincing victories, including a 350.9 to 294.4 edge in opponent total yards per game and 2,491 to 1,901 edge in rushing yards for the season. The Ravens did throw for more passing yards for the season with 3,739 to 3,298, but teams with more passing yards lose the Super Bowl more often than they win it.

Here’s the most significant aspect of this year’s prediction. Prior to this matchup, 10 Super Bowl teams had 10 or more of the 12 indicators in their favor. Nine of those 10 teams went on the win the game, most recently in the 2008-2009 season when the Steelers beat the Cardinals. Interestingly, the one team that failed to win a game with a 10-category advantage was the first team with that advantage, the Vikings in the 1969-1970 season’s Super Bowl. The Ravens will attempt to be the first team to beat those odds in 43 years.


COMMENTS (6 Comments, most recent shown first)

So you're basically saying that when the signal is very strong, as it was this year, the prediction system is very reliable. I'd say you shouldn't give up your day job.

1:34 PM Feb 4th
Baltimore 34 San Francisco 31
7:20 AM Feb 4th
Craig Jolley- Without looking, I would suspect regular season field goal kicking might not be that predictive because, according to Football Outsiders, it is not a skill that repeats itself well year-to-year. I assume the same would hold true regular season to playoffs.

Here's one: The last 7 Super Bowls have been won by the quarterback whose 1st name came 1st in the alphabet.....
12:29 PM Jan 28th
This is fine, but will they cover?
12:16 AM Jan 28th
Have always had the sense that the superior team wins the Super Bowl more often than in other games -- and that much superior teams win blowouts more often.

If so, it could be because it's a game in which both teams load for bear, and play close to their maximum abilities, so that "noise" decreases as a factor.

Suspect that the Super Bowl is one of the "purest" contests in sports, in terms of finding out who's the best team. Just an impression.

But if that's accurate, it would explain why this prediction system is more accurate than experts would expect that it would be. You're really just asking who's the better team, and then leaning more heavily than usual on that identification.

Or not :- )

6:28 PM Jan 26th
A couple of questions:

1. Does the prediction system work equally well throughout the playoffs?

2. My sense is field goal efficiency (not sure about the best way to measure) might take on added importance in what could turn into a close, low-scoring game as suggested. Any evidence to confirm or deny FG as a valid indicator in defensive match-ups?
12:02 PM Jan 26th
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by|Terms & Conditions|Privacy Policy