Applying Stats To Predict The Winner Of Super Bowl 47


Who will win Super Bowl 47? Will the San Francisco 49ers cover the 3.5 point betting line or will the Baltimore Ravens cause the upset? Today on the blog Mark Taylor applies his NFL predictive model and delivers his verdict.


The beauty of the NFL is that it combines the twists and turns of a league based season with all the sudden death excitement of a knock-out cup competition. The regular season, despite its faults of a truncated, unbalanced schedule, allows enough games for most of the cream of this year’s sides to make the post season. The post season then tests to breaking point the theory that on any given Sunday, any given team can win a single game.

In short, the best usually make it through to the post season, but then, even with the inbuilt advantage of bye weeks and home field advantage and because of the absolute requirement that the AFC plays the NFC in the final game, the Super Bowl contest can see that season’s great side play the merely good.

Superbowl XLVII takes place in New Orleans, late on Sunday evening UK time. The game will start around eleven-thirty in Britain and because of the extended halftime entertainment, won’t conclude until well past 3 am on Monday morning. The Baltimore Ravens, fourth seeds from the AFC represent their Conference and they face the NFC’s second seeded, San Francisco 49ers.

Preliminary Assessments

Baltimore’s 10-6 regular season record was bettered by eight of the eleven other teams which also made the playoffs and equalled by two sides which didn’t. So they can reasonably be described as a good, but not great team. While the 49ers overcame a partly self-imposed, mid-season switch of quarterback to post an 11-4 record with one game tied. So on mere games won the NFC, as represented by the 49ers look to have the initial edge.

We can try to confirm and quantify this advantage by referring back to how efficiently each side has moved the ball and prevented their opponents from gaining ground, both on the ground and through the air. Yards per attempt can tell us a great deal about a side’s strengths, especially when corrected for the strength of opponents faced.

The 49ers own the more impressive figures, so I’ll illustrate the efficiency statistics by describing their rushing efficiency. They ran the ball 492 times during the regular season for a total of 2491 yards at 5.1 yards per carry. The defences they faced allowed a total of 30427 yards on 7056 carries at 4.3 yards per carry. So the 49ers ran for 0.8 yards per carry more than a basket of teams managed in 256 games against San Francisco’s regular season opponents, or for convenience 115% of average.

The Offensive & Defensive Efficiency Stats For Baltimore & San Francisco.

TeamRaw Offense Rushing EfficiencyRaw Offense Passing EfficienyRaw Defense Rushing EfficiencyRaw Defense Passing Efficiency
Baltimore 1.04 1.01 0.96 0.97
San Francisco 1.15 1.15 0.86 0.86

The hinted at difference in quality from the win/loss/tie records of each side becomes more obvious when we look at these efficiency stats. San Francisco has a frighteningly good offense, which has continued to thrive even when second year quarterback, Colin Kaepernick replaced long time 49er’s passer, Alex Smith. Defensively, they are equally good, allowing teams to pass or rush for just 86% of their usual yardage. They are truly an excellent, balanced football team.

Statistically, the Ravens bring a reasonable, above average offense, more reliant on running the ball and a rounded, good, but not great defense. Certainly their defense, while above average, falls well short of the imposing units of old, when with many of the same, but younger legs, they were restricting teams to efficiencies which were comparable to those now being posted by the current 49ers team.

San Francisco’s all round efficiency numbers are good enough to place them at the head of the standings for all 32 teams. They occupied 2nd spot after week four when we first looked at these standings, a place behind Baltimore, who has since slipped down to eighth, mainly because of a less consistent offensive performance. So Sunday’s game doesn’t quite feature the NFL’s two best current sides over the course of the season, (Denver would have had to overcome Baltimore for that dream matchup to have occurred), but we do have a game that is typical of recent Super Bowl contests.

Super Bowl 47: An Analysis

To get to the real statistical heart of the game, we need to combine the respective defensive and offensive stats for each side. For example, Baltimore’s 96% run defense should slow down the 49ers 115% multi threat run offense, but the west coast side should still be able to move the ball around 111% of league average. The reverse matchups predict an evening of potential struggle for quarterback, Joe Flacco and the Ravens and overall the 49ers appear to have the game to win a competitive contest.

Game Day Matchups For Super Bowl XLVII

TeamGameday Rushing EfficiencyGameday Passing Efficiency
Baltimore 0.89 0.86
San Francisco 1.11 1.11

The Super Bowl is the unique game of each season because it invariably takes place at a neutral venue. Both sides have performed similarly on the road this year. Baltimore had a slightly inferior regular season road record, but won in Denver and New England, two extremely inhospitable places, during post season play. The 49ers were slightly better on the road in the regular season, including in New Orleans and then overcame an awkward, three time zone jump to defeat top seed Atlanta in the Championship round.

With travel doubts allayed and a full two weeks to prepare, we can compare the game day matchups to a database of such pairings from 2002 onwards. San Francisco’s above average expected offense should out point Baltimore’s expected struggles by just over four points based on each team’s expected game day efficiency stats.

Similar approaches based on points scored instead of yards gained, produce a very similar obstacle for Baltimore to overcome. The offenses faced by Baltimore average 23 points per game, but just 21.5 ppg when they played the Ravens. This confirms Baltimore as an above average defense, 1.5ppg better than average. Similarly, the 49ers opponents scored 22.8 ppg on average, but just 17ppg against the Niners. Consequently, the 49ers defensive unit is well above average and their ability is again quantified.

The Verdict

We can tally the figures for the offenses as well. San Francisco is again the better overall side, three points better than average on offense compared to just 2.5 ppg better for Baltimore. If we combine these relative points above average for each side, the 49ers emerge the overall superior outfit, this time by a total of five points. Currently, Baltimore is generally being given a 3.5 point start.

Statistics such as these are excellent at describing season long win/loss records, but individual games are sometimes decided by single events, such as an unconscious receiver fumbling the ball, as happened in New England during Baltimore’s Championship victory. There is also the not inconsiderable, common three point margin of victory to overcome. Points in the NFL are scored, most commonly, in chunks of three or seven points. However, San Francisco is the justifiable favourite on Sunday.

Don’t miss the event, even if you’ve never watched a game of American Football. Sport and entertainment collide to produce four hours of always engrossing spectacle.



Read more of Mark's work on his The Power Of Goals blog

And follow Mark of Twitter: @MarkTaylor0

NFL and football fan. I've seen my two favourite sides, Stoke and the San Diego Chargers play at the new Wembley....and both lost.