The best team lost the Super Bowl. The manner in which Super Bowl XLV played itself out provided a perfect demonstration of the problems facing football handicappers. Those problems prevent even the best handicappers from winning at better than a 60% long-term rate.
We all know the platitudes regarding NFL games and the Super Bowl in particular:
1.. The Super Bowl is won by the superior defense.
2. The team with the most rushing yardage wins in the NFL.
3. The team that has the longest times of possession will win.
In my pregame analysis I wrote in part:
"This is a very closely matched Super Bowl, with one of the most accurate lines I've ever seen in the big game. * * *
To make a prediction in this game required very deep analysis of the edges of both teams.
* * *
[Green Bay's passing] advantage is at least partially, if not completely, negated by the strong rushing defense of Pittsburgh. The Steelers have alowed a tiny 61 yards per game on the ground this season, and GB has only managed 103 yards per game rushing. The Steelers should be able to shut down whatever little running game GB possesses, and be able to key defensively on the pass. Expect coverage sacks, a lower completion rate than Aaron Rodgers excellent 66%, and shorter than normal passes.
* * *
The yardage translates to a 1 point edge for Pittsburgh before we consider the difficulty that GB will have attempting to pass if they can't run. The expected reduction in passing yards for GB gives the Steelers a 3-point advantage."
Pittsburgh gained 50 more net yards in the game, which equates to the 3-point edge handicapped before the game. As predicted, the Steelers completely shut down Green Bay's rushing offense, permitting them an infinitesimal 50 yards on the ground for the entire game. That's 50% less than Green Bay's average rushing yardage, and 76 fewer yards than the Packer's defense allowed to the Steelers.
Pittsburgh held the ball for 33:25 minutes compared to only 25:35 for the Packers -- a huge 10% advantage in time of possession. Given the difference in time of possession, it is no surprise that the Steelers led 19-15 in first downs and 64-55 in total plays.
Packer QB Rodgers was held to 61% completions, down from his 66% average and lower than Pittsburgh's 62.5% completion rate on more attempts. Pittsburgh's defense recorded three times more sacks than Green Bay's defense.
The Packers were forced to punt twice as many times, and averaged just 40.5 yards per punt compared to an average of 51 yards per punt for the Steelers. Neither team gained any meaningful yards on punt returns, but the Packer's average kickoff return was only 13.5 yards, compared to an average kickoff return of 18.5 yards for Pittsburgh.
In short, Pittsburgh was superior in almost every category capable of being handicapped. The only category in which Green Bay had better numbers was yards per pass. Despite being held below his 8.3 yard per pass attempt average, Rodgers had a 7.8 to 6.6 yard per attempt advantage in the Super Bowl. Aggregating all categories, Pittsburgh was clearly the better team. Still, they lost.
The difference was three turnovers, all of which resulted in touchdowns, and one of which was returned for a touchdown. If shown the post game statistics without the turnover numbers and asked which team won, every statistical handicapper would choose the Steelers. The difference between Pittsburgh winning the game outright and losing by 6 points was the 7 points represented by the interception run back for a touchdown.
Turnovers are one of the most important factors in football the effect of which are completely subject to luck, and cannot be handicapped in advance. That is not to say that the tendency of a team to fumble or the tendency of a quarterback to be intercepted, as well as the ability of a defense to force fumbles and interceptions cannot be handicapped. Such factors can be handicapped. What is subject to luck is:
1. Whether a fumble will result in a turnover, which is, to a great extent dependent on the random bounces of a football; and
2. The location on the field where any turnover occurs. A turnover between a team's own goal line and midfield will usually result in a touchdown for the opposition, while a turnover in the opponent's red zone may prevent a score but otherwise is no more damaging than a good punt.
Every punt is also a turnover, and Green Bay was forced to punt twice as often as Pittsburgh, with each punt being 10.5 yards shorter than Pittsburgh's punts. The point is that punts don't hurt a team to the same extent as turnovers in their territory. Turnovers near the opponent's goal don't hurt any more than punts.
The inability to handicap the number of fumbles that will be lost, and the location of turnovers is one of the biggest dangers to good handicappers. The other two are in-game injuries, and players simply not playing up to their abilities.
In the Super Bowl, Pittsburgh was not only subject to a turnover deep in their own territory, but two more at midfield. Each team had one fumble. The ball bounced back to Green Bay on their fumble, while Pittsburgh lost their fumble.
In addition, Roethlisberger choked and did not play up to his ability demonstrated during the season. From a handicapping standpoint, Roethlisberger had thrown 7 interceptions in 12 games played, for an average of only .6 interceptions per game, while Rodgers had thrown 13 interceptions in 15 games for a 50% higher .9 interceptions per game. In terms of pass attempts, 2.3% of Rodgers' attempts were intercepted this season, almost twice as many as the 1.2% of attempts resulting in interceptions for Roethlisberger. Green Bay's defense achieved 1.6 interceptions per game compared to 1.3 per game for Pittsburgh's defense. Based on the offensive and defensive numbers, the probability of an interception had to be considered higher for Rodgers than for Roethlisberger. Roethlisberger's two interceptions in a single game to none for Rodgers could not have been foreseen from an examination of the past performances this season.
For those who believe the side the bookmaker needs always wins, the bookmakers needed Pittsburgh. They were loaded with Green Bay. The public bet heavily on Green Bay, sending the line up from -2.5 to -3. The number "3" occurs with more frequency than any other final score differential in the NFL. There's a huge difference between getting +3 and getting +2.5. The move onto 3 from 2.5 requires a great deal of money to be bet on the favorite. Wise guy money came in on Pittsburgh late. The public won, but the win couldn't be handicapped in advance.
It doesn't happen often, but the better overall team lost the championship. Given a chance, I would pick Pittsburgh again, and bet against another confluence of bad luck.