We often hear talk in sports about regression to the mean. In most cases the talk centers around a player having a career year falling back down to earth, but regression doesn’t have to be a negative thing. Regression to the mean simply refers to the phenomenon that, all things being equal, the best bet for any given player with a long enough track record to establish some notion of said player’s abilities is that he will perform at a level close to the level his track record has established as normal for him. No, that isn’t a mathematical definition, but it’s accurate enough for our purposes.
The more a veteran player deviates from his average level of play in one year, the better the odds that year will prove to be an aberration, a statistical blip, if you will. So, for example, Falcons fans should probably not be making their Super Bowl reservations just yet, as quarterback Matt Ryan last season suddenly morphed from a long career of good but not great play into the best quarterback in the NFL, at least from a passer rating perspective. Ryan made an enormous jump in his level of play fairly late in his career. It is possible he can sustain that level of play, but that’s not the way to bet. The better bet is that Ryan will regress back towards what he has been every other year of his career. This isn’t meant to dump on Ryan or Falcons fans, it was just the first example that came to mind.
What does this have to do with the Jets, you might ask? Good question. I’m so glad you brought it up. As it turns out, regression in its daily usage is strictly a negative thing. If we see improvement in something in our every day lives we don’t say things have regressed, we say they’ve progressed. But in statistics there is only one word describing the phenomenon of a tendency to return to established levels of performance: regression. So if, for example, Muhammad Wilkerson had the worst year of his career in 2016, we might expect regression to the mean in 2017, but that would be a positive thing for Wilkerson and the Jets. Positive regression, so to speak.
Now quick, name a Jet who had a career year in a positive way last year. Maybe Brian Winters? Even with Winters, he’s still seemingly on an upward career arc, so it’s difficult to say if we should expect regression from his 2016 level of play or if 2016 represents something close to his long term baseline level of play. It’s difficult to think of a single other Jet that represents a significant risk of his play coming back down to earth based on the regression to the mean principle. Perhaps Morris Claiborne would be an example, but he wasn’t with the Jets last year and in any event only played eight games.
Now, is it equally difficult to think of Jets who had an uncharacteristically dismal year in 2016? Not really, right? Muhammad Wilkerson, Sheldon Richardson, Kelvin Beachum; these are all players who have in the past been better than average players at their respective positions, and all of them play major roles in the Jets’ plans for 2017. Each of those three players would appear to be candidates to benefit greatly from a regression to the mean in 2017. To the extent there is hope for the Jets in the 2017 season, it revolves primarily around hopes for improvement in the many young players expected to fill major roles on this team, and around bounce back years (positive regression) from some of the team’s best players. Regression to the mean could be a fairly powerful force for the 2017 Jets if things go right. Perhaps we can start a new chant for our hoped for bounce back boys: be regressive, be be regressive!