Over the past six months Marty and I have penned several columns for The New York Times. Five of these columns have been part of the Keeping Score feature. Frequently the Keeping Score column is written by Alan Schwartz [The Numbers Game author and the person who described our book as “Freakonomics meets ESPN.”]. Other people, though, contribute to this feature.
Today’s column was penned by Aaron Schatz. I discussed Schatz’s excellent piece at The Sports Economist earlier today. Before repeating what I said in that forum, I want to add that it is interesting that Schatz finds that kickers – like quarterbacks – are also quite inconsistent. At least we see inconsistency with respect to field goal accuracy.
With respect to distance on kick-offs Schatz finds that kickers are relatively consistent. In fact, Schatz argues that distance on kick-offs is “one of the most predictable individual stats in the entire NFL.” The correlation coefficient for kick-off distance is only 0.60.
To put that in perspective, a basketball player’s Win Score per minute this season has a 0.82 correlation with what he did last year. From what Schatz says, it appears no aspect of an NFL player’s performance approaches the consistency we see in the NBA across time. Clearly part of this is that NFL teams play such a small sample of games. And part of this is that a player’s performance in the NFL depends so much on his interaction with his teammates – although this issue should not be relevant for field goal accuracy.
People argue that the interaction effects in the NBA imply that the statistics tracked for individuals are meaningless. The relatively high correlation we see in performance across time in the NBA, though, suggests that these interaction effects are exaggerated. Although player performance in the basketball does exhibit some variation, the level of variation is less than what we see in football and baseball. Furthermore, it is not clear that the variation we observe is entirely due to the interactions with a player’s teammates. Coaching and injuries also cause player performance to change.
In sum, before one should advocate dismissing all the statistics the NBA tracks there should be stronger evidence that this step is necessary.
Okay, enough on that topic for for now. My purpose today is to focus on the work of Schatz. I suspect that people who read our work in this forum do not always check out The Sports Economist. This is a shame since the level of talent at The Sports Economist is quite a bit higher than what we offer here at The Wages of Wins. Looking over the roster of economists at The Sports Economist makes that point pretty clear.
Still, for some clicking over to a new website might be difficult. So to save people from clicking, here is what I said about Schatz’s wonderful column today:
The NFL Evaluate Kickers Incorrectly
Aaron Schatz – of Football Outsiders fame – has penned an interesting Keeping Score column in today’s New York Times. The column makes a simple, yet bold assertion in the title: “NFL Kickers are Judged on the Wrong Criteria.”
Schatz argues there are two aspects to a kicker’s performance – accuracy on field goal attempts and distance on kick-0ffs. Of these two characteristics, field goal accuracy gets you paid in the NFL. To bolster this argument he notes the $5.4 million the Dallas Cowboys are scheduled to pay Mike Vanderjagt over three seasons. Vanderjagt is the NFL career leader in field goal accuracy. He is also, according to Schatz, one of the NFL’s worst kickers in distance on kick-offs.
Given the money paid Vanderjagt’s, it appears accuracy on field goals is what drove the Cowboy’s decision. Yet Schatz says that focus is misplaced.
“There is effectively no correlation between a kicker’s field-goal percentage one season and his field-goal percentage the next. But average kickoff distance shows more consistency from season to season than almost any other individual statistic in the N.F.L.”
At Football Outsiders Schatz expands on this observation:
“Measuring every kicker from 1999-2005 who had at least 10 field goal attempts in two consecutive years, the year-to-year correlation of field goal percentage is .03. … On the other hand, the year-to-year correlation of average kickoff distance — same time period, same minimum of 10 kickoffs — is .60. That makes average kickoff distance one of the most predictable individual stats in the entire NFL, at any position.”
Still, teams appear to be paying for accuracy. In essence, Schatz argues that the NFL is Fooled by Randomness. Field goal accuracy defies prediction, so noting past accuracy is not useful in determining wages. Yet NFL teams are incorporating this information in negotiating salaries.
Although this story is interesting by itself, one could argue that this is just one more example of the Moneyball phenomenon in sports. Is anyone compiling a list of such examples?