Passer rating flaws
The NFL has this stat for quarterbacks called the passer rating. It's formulation is nonintuitive, and while the lowest possible rating is 0.0, the highest possible rating is 158.3. The league average rating is usually in the high 70s, with the leader being around 100.
The passer rating has four components, with a quarterback being rewarded for 1) completing a high percentage of passes, 2) throwing for a large number of yards per attempt, and 3) throw a large number of touchdowns per attempt, and being penalized for 4) throwing lots of interceptions.
There are several quirks in the formula. One for example is how one might think that throwing 10 passes and completing none of them should result in the lowest possible score, but no, the passer rating rewards this performance for not throwing it to the other team, and gives it a 39.6. (One interception in ten attempts is however 0.0.) And then there are questions about its relevance because quarterbacks can do very important things to help their teams win by running the ball, which is completely ignored by the stat and contributes in a small way to lowering the rating.
However, the quirk I choose to focus on in this post is how the rating for a season is not the average or sum of the ratings of individual games. This is not completely related to the cutoffs where a quarterback does not get any more points for increasing his completion percentage from 78% to 81%, but this can really be taken advantage of to create some interesting scenarios. For example, please examine the following four games, which are not totally out of the realm of plausibility in the NFL:
Game 1:
7 comp/25 att, 188 yards, 2 touchdowns, 0 interceptions->Rating: 85.1
Game 2:
18/25, 295 yds, 1 TDs, 3 INTs->Rating: 85.0
Game 3:
17/25, 235 yds, 2 TDs, 4 INTs->Rating: 85.0
Game 4:
13/25, 100 yds, 3 TDs, 1 INTs->Rating: 85.0
The total of these four games:
55/100, 818 yds, 8 TDs, 8 INTs->Rating: 75.3
Does this make the passer rating a bad stat? I tend to think not when used for a season. The problem I feel is how the passer rating for a single game is being used more frequently. This should stop. Only Rams QB Marc Bulger had more single games (8) this season with rating over 100 than Chicago QB Rex Grossman's 7, despite the fact that Grossman gets all sorts of flak from how he's terrible because he had games with ratings of 36.8, 23.6, 10.2, 1.3, and a season ending 0.0. But the Bears did win 2 of these 5 games.
The passer rating has four components, with a quarterback being rewarded for 1) completing a high percentage of passes, 2) throwing for a large number of yards per attempt, and 3) throw a large number of touchdowns per attempt, and being penalized for 4) throwing lots of interceptions.
There are several quirks in the formula. One for example is how one might think that throwing 10 passes and completing none of them should result in the lowest possible score, but no, the passer rating rewards this performance for not throwing it to the other team, and gives it a 39.6. (One interception in ten attempts is however 0.0.) And then there are questions about its relevance because quarterbacks can do very important things to help their teams win by running the ball, which is completely ignored by the stat and contributes in a small way to lowering the rating.
However, the quirk I choose to focus on in this post is how the rating for a season is not the average or sum of the ratings of individual games. This is not completely related to the cutoffs where a quarterback does not get any more points for increasing his completion percentage from 78% to 81%, but this can really be taken advantage of to create some interesting scenarios. For example, please examine the following four games, which are not totally out of the realm of plausibility in the NFL:
Game 1:
7 comp/25 att, 188 yards, 2 touchdowns, 0 interceptions->Rating: 85.1
Game 2:
18/25, 295 yds, 1 TDs, 3 INTs->Rating: 85.0
Game 3:
17/25, 235 yds, 2 TDs, 4 INTs->Rating: 85.0
Game 4:
13/25, 100 yds, 3 TDs, 1 INTs->Rating: 85.0
The total of these four games:
55/100, 818 yds, 8 TDs, 8 INTs->Rating: 75.3
Does this make the passer rating a bad stat? I tend to think not when used for a season. The problem I feel is how the passer rating for a single game is being used more frequently. This should stop. Only Rams QB Marc Bulger had more single games (8) this season with rating over 100 than Chicago QB Rex Grossman's 7, despite the fact that Grossman gets all sorts of flak from how he's terrible because he had games with ratings of 36.8, 23.6, 10.2, 1.3, and a season ending 0.0. But the Bears did win 2 of these 5 games.
0 Comments:
Post a Comment
<< Home