"You are what your record says you are."
Bill Parcells' famous quote about the NFL brilliantly expressed the importance of not making excuses and playing the "what if" game in an NFL locker room. Teams win and lose games, and all the justification in the world doesn't change that.
Unfortunately, Parcells' truism is also completely useless for understanding what is going to happen in the future.
This week's advanced stat is Pythagorean wins.
The concept is simple, even though the math is tricky. The idea is that over time, a team's point differential (points scored minus points allowed) is a better indicator of how that team will perform in the future than their record is.
For example, a team that is 6-2 but has been outscored by five points on the season because of a couple of blowout losses is more likely to play like a 4-4 team going forward than like a 6-2 team.
While fans like to point to a team's record and crow, oftentimes there are warning signs that the luck is about to run out. The same can be true for a struggling team that has actually outscored opponents on the season. This relationship between future wins and point differential is not debatable.
For the purposes of projection into the future, you are better off looking at Pythagorean wins than real wins. The 2010 Packers won 10 games on the field, but were dominant on the field, leading the NFC in point differential. They weren't really a 10-win team sneaking into the playoffs on tiebreakers—they were really a juggernaut with some bad luck. The playoffs revealed the truth.
There are many different ways to calculate this number, but they all are based on essentially the same components: "points for" and "points against." This can be a handy tool for predicting a big fall from a team with a good record, or a big leap forward from a team that had some bad luck. No statistical tool is right all the time, but Pythagorean wins has a good track record and is far more useful than actual wins for predicting the future of a team.
For a decade, the Indianapolis Colts were the kryptonite of expected wins. Pythagorean wins tend not to work well at the edges of the NFL. The worst teams always show up as having underperformed and the best teams typically over-perform. That's just the nature of the metric. The Colts under Peyton Manning threatened to blow the whole thing out of the water, however.
The Colts outperformed their expected win total every year from 2002 to 2010. The odds of this being just luck are extraordinary. Under Peyton Manning, the Colts could be counted on to pull out an extra game or two more each year than what a normal team with the same talent would. It was a remarkable run that is unlikely to be matched.
The 2011 Colts were the exact opposite of everything that happened to the Colts in the past decade. They were obviously terrible, but they also showed a unique ability to fail in the clutch. The Colts had the ball with a chance to tie or take the lead in the fourth quarter of games ten different times, and only succeeded in three of them. That includes the Pittsburgh game which they eventually lost despite tying the score. The 2011 Colts were a terrible team, but they could just as easily have been a five or six win team rather than a two win team. The numbers say they should have been at least a win better than they were.