Another Favre Column
The big story in sports is Jet Favre. Brett Favre unretired and discovered the Packers didn't want him anymore. After much anguish, the Jets sent a conditional fourth round draft choice to the Packers. The Jets then proceeded to cut Chad Pennington. For those keeping score...
the Jets have said that Brett Favre is better than Chad Pennington.
the Packers prefer Aaron Rodgers over Favre.
and the Buccaneers, who apparently were a candidate to acquire Favre, have essentially said they prefer Jeff Garcia.
Given all these statements I thought I would pretend for a moment that this is a basketball story. In other words, I am going to measure the Wins Produced of all the players in the story and simply note which team is making the best decision. The Wins Produced story Table One presents such analysis for Favre, Pennington, and Garcia.
Table One: Comparing Brett Favre, Chad Pennington, and Jeff Garcia
Pennington began playing in 2000. As we can see in Table One, Pennington's career averages currently stand as follows: Wins Produced: 10.5 Net Points: 406.7 Wins Produced per 100 plays: 0.478 Net Points per play: 0.185 To put these numbers in perspective, the average quarterback posts a Wins Produced per 100 plays of 0.406 and a Net Points per play of 0.155. So Pennington appears to be an above average quarterback. But of course, he is no Favre. Or is he? Here is what Jet Favre has done since 2000: Wins Produced: 22.1 Net Points: 862.3 Wins Produced per 100 plays: 0.463 Net Points per play: 0.181 When we look at the per-play metrics (and I will talk about the aggregate metrics in a moment) we see that Favre, like Pennington, is above average. But Favre - again on a per-play basis - is not quite as good as Pennington. And the story doesn't change if you look at Favre's entire career. In Favre's first nine seasons (prior to 2000) his Wins Produced per 100 plays was 0.460 while his Net Points per play was 0.179. Again, above average marks. And again, not quite what Pennington has done in his career. Okay, Favre - on a per-play basis - is not quite as productive as Pennington. What about Garcia? Here is what Garcia has done since 2000: Wins Produced: 18.7 Net Points: 719.5 Wins Produced per 100 plays: 0.543 Net Points per play: 0.208 Once again we see an above average quarterback. And once again we see per-play measures that exceed what is offered by Favre. When we look at these per-play measures it looks like the Bucs made a good decision. The Jets decision, though, appears somewhat questionable. And the Packers...well, I will get to them in a moment. Some Problems For now, I want to focus on two problems with the analysis.
My comparison of the three quarterbacks emphasized the per-play measures. But that focus is misleading. Pennington - as any Jets fan would know - has trouble making it on to the field. In only one season has Pennington accumulated more than 500 plays. Meanwhile Favre has accumulated more than 500 plays every year he played in Green Bay. And nine times his plays topped the 600 mark. As a result, Favre has produced far more wins than Pennington. And the same goes for the comparison between Favre and Garcia.
Favre is certainly more consistent taking the field. But as I have emphasized in the past, consistency is not a word we should use in discussing quarterbacks. Past performance of quarterbacks is not a very good predictor of future performance. In The Wages of Wins we discussed this observation explicitly with respect to Favre and also with respect to quarterbacks in general.
This second point was also made in the following posts. The Value of Player Statistics in the NFL Football Outsiders and QB Score And it is this second point that makes the discussion of football players different from basketball players. Basketball players - relative to what we see in baseball and football --- are quite consistent. Hence once we measure what a player has done in the past we have a pretty good idea what he will contribute in the future. Such a statement, though, cannot be made about football players. Relative to last year, Favre now has a completely different set of receivers. He also has a different offensive line, different running backs, and most importantly, a different offense (and coaches calling the plays). All of these differences should make a difference. How much of a difference? The stats tracked for quarterbacks don't seem to answer this question. We do see inconsistency across time in the quarterback's numbers. And so we suspect all the stuff changing around the quarterbacks is changing the stats we see. This also tells us that the stats used to evalate quarterbacks are not just about the quarterback, and therefore can't tell us who is "better" or "worse". As a consequence, we can't really say that the Jets are better off with Favre. Or whether the Packers or Buccaneers made a good or bad decision. Winners and Losers? Actually, let me amend this statement. A few weeks ago I commented on Favre's return. I want to build upon that post by identifying the winners from all that has transpired:
The Packers, I think, are winners. Favre retired back in March. At that point the Packers committed to Aaron Rodgers (something they were going to have to do eventually). When Favre returned the Packers decision-makers suffered some anguish. But when all is said and done, they still have Rodgers at quarterback, which is where they were when Favre retired. Now, though, they also get a high draft choice from the Jets. So the Packers are better off than they were when Favre first left the Packers last March.
Favre, I think, is also a winner. He gets to keep playing football. And he gets to continue collecting his hefty paycheck. So that's a winning scenario for Favre.
Sportswriters are also winners. Sportswriters seem to delight in writing stories about how players are threatening their "legacy." Such columns must be easy to write since we see many of these as a player's career approaches its end. Now with every bad game Favre plays with the Jets members of the media can knock out another "legacy" column. In fact, I imagine sportswriters have probably already written these columns, thus reducing their workload in the future.
How about the Jets and Bucs? Again, we can't forecast what a quarterback will do in the future. So it's hard to see whether these teams will ultimately be winners. And the same story is told about the Miami Dolphins, the new employer of Pennington. Yes the past data suggests that Pennington will be an upgrade over Josh McCown and John Beck. But as I have said repeatedly, quarterbacks are really like mutual funds. Past performance really doesn't tell us much about the future. Consequently, every Lions fan (and I am one) can enter every season hopeful about the future (and this makes football great). - DJ For more on The Wages of Wins football metrics see The New QB Score Consistent Inconsistency in Football Football Outsiders and QB Score The Value of Player Statistics in the NFL