My Issue with Colts Statistical Analysis

Over summer, football fans are hungry for news.  There is very little going on, particularly between organized team activities and training camp, so many will find numerous sources using statistical analysis to analyze NFL teams — and predict the likelihood of future success.

I think statistical analysis is a great tool for reviewing performance on a large scale.  It allows fans to get a panoramic view of the kind of team the Colts, or other teams, were in the previous season.  Statistics also serve as a tool for loose projections about future success.

The issue I have with statistics is that they do not go far enough to tell the whole story.  Very rarely are they broken down into enough cross-sections or variables for an accurate perspective to form.  Beyond that, no matter how many different ways they are broken down they can be misleading.

Crunch those numbers!

Take the following statistical analysis I did following the 2009 season.

Colts Defense

Behind < 7
14 Poss│64 Rush Yd│285 Pass Yd│2 TD│2 Int│2 FTO│1 ToD│7 Punt
14% Score│36% TO│50% Punt

Down less than a touchdown the defense either created a turnover or forced a punt 86-percent of the time in 2009.  What happened when the score was tied?

Tied
19 Poss│368 Rush Yd│315 Pass Yd│3 TD│6 FG│2 FTO│7 Punt
50% Score│11% TO│39% Punt (1 Possession Ended Half)

Something happens here.  What happens cannot be understood or explained by statistical analysis.  For one reason or another, the Colts defense in 2009 really stiffened up when failing would put the team down two scores — but without that kind of threat, the defense played a very traditional “bend but don’t break” style.

While the causes for this statistical anomaly are varied and far too detailed to explain with numbers, there is an explanation that numbers themselves cannot provide.  No matter what the numbers say, the Colts defense was a better defense in 2009 than it was in 2008.  No matter what analysis Football Outsiders or other sources provide regarding the statistical performance of Daniel Muir and Antonio Johnson last year compared to the performance of our defensive tackles in 2008, our defensive interior was far better in 2009.

I know it was better because I watched the team play every game, every down.  There were times when the Colts defense reared an old face and underperformed, but generally, when the game was on the line and things counted, the defense stepped up and helped the Colts win several very close games last season.

On the one hand, it is easy to praise Peyton Manning for his brilliance in engineering seven come-from-behind victories last year.  On the other hand, dismissing the reality that it also took a defense capable of shutting down the opposing offense in those situations to provide Manning with his opportunity seems kind of short-sighted.

Don’t get me wrong, I think statistics are a great tool in a repertoire of data to consider when forming opinions about NFL teams and their likelihood to succeed.  I just think statistics are not a replacement for real-time observations or knowledge about how a team performs when the game is on the line.

The Colts defense was much improved last season.  There is every indication to believe that this season they will be even better, after some first year starters put a year of experience under their belts, and now that the team has had a year to get acclimated to Larry Coyer’s new defensive wrinkles.  Football Outsiders or other statistical analysis sources may tell you that the statistics do not bear me out, but I will trust my eyes on game-day more than a collection of numbers months later.

Quantcast