How Passer Rating Works

[percyhoward]

Because of the confusion over whether the Cowboys passing so often inflates team passer rating:

There are only 4 factors in passer rating: completion percentage, yards per attempt, TD%, and INT%.

However many passes you throw, as long as those 4 factors stay the same, passer rating stays the same.

6 of 10 for 100 yards 1td 1int 87.5
12 of 20 for 200 yards 2td 2int 87.5
18 of 30 for 300 yards 3td 3int 87.5
24 of 40 for 400 yards 4td 4int 87.5

 

[burmafrd]

Because of the confusion over whether the Cowboys passing so often inflates team passer rating:

There are only 4 factors in passer rating: completion percentage, yards per attempt, TD%, and INT%.

However many passes you throw, as long as those 4 factors stay the same, passer rating stays the same.

6 of 10 for 100 yards 1td 1int 87.5
12 of 20 for 200 yards 2td 2int 87.5
18 of 30 for 300 yards 3td 3int 87.5
24 of 40 for 400 yards 4td 4int 87.5

that formula reallly needs to be overhauled in the current NFL

 

[AbeBeta]

that formula reallly needs to be overhauled in the current NFL

Why? The formula is consistent.

What needs to be improved is fans' understanding of statistics and their interpretation.

 

[percyhoward]

that formula reallly needs to be overhauled in the current NFL

You hear that a lot, but nobody has been able to come up with another rating that has a higher correlation to winning.

 

[AbeBeta]

You hear that a lot, but nobody has been able to come up with another rating that has a higher correlation to winning.

Is that really true? I hear people say that all the time but I've never seen an actual analysis. Of course, it stands to reason that a statistic that addresses accuracy, yardage, tds, and interceptions is going to be more predictive than any one of those values (or anything else) individually --- that's just basic statistics. What we don't know is whether an adjusted statistic -- for example one that includes QB rushing yards and fumbles as part of the equation is going to be a better measure that more strongly correlates

 

[burmafrd]

The emphasis needs to be increased as regards YPA and decreased as regards Completion percentage. If that were done it would be more in line with the NFL of today.

Factoring in fumbles and such- I think it would just muddy the waters

 

[AbeBeta]

The emphasis needs to be increased as regards YPA and decreased as regards Completion percentage. If that were done it would be more in line with the NFL of today.

Factoring in fumbles and such- I think it would just muddy the waters

If the goal is to have a statistic that is predictive of winning then there is really no way that factoring in fumbles would muddy the waters.

 

[Yakuza Rich]

The problem with the formula is that it favors more WCO and dink and dunk passing offenses. Particularly those that throw the ball a ton in the red zone. This was one of the things that Favre used to do all of the time. He would throw for a TD inside the 5-yard line. However, he was really great at being able to do that (it's hard to do than it sounds). And if we look at the QB ratings over the years, the modern QB has a much higher QB rating. That's where you get all of the 'Mark Sanchez was better than Johnny Unitas at this point in his career talk.'

Well, no...he wasn't.

Sanchez just played in a scheme that helps boost QB rating whereas Unitas had to rely on 7-step drop pass plays and threw it deeper on a more consistent basis.

I like the QB rating metric when you take my simple formula of QB Rating Differential (QBRD) and simply take a team's QB Rating and subtract their QB Rating Allowed on defense. It still leaves out some key metrics that factor in forecasting a team's success, but it's a quick-n-easy way to figure out how well a team is playing.







YR

 

[Tricericon]

Is that really true? I hear people say that all the time but I've never seen an actual analysis. Of course, it stands to reason that a statistic that addresses accuracy, yardage, tds, and interceptions is going to be more predictive than any one of those values (or anything else) individually --- that's just basic statistics. What we don't know is whether an adjusted statistic -- for example one that includes QB rushing yards and fumbles as part of the equation is going to be a better measure that more strongly correlates

It is emphatically not true. The far simpler (pass yards + 20*(pass TD) - 45*(interceptions thrown) - sack yards)/(passing attempts + sacks) is superior.

There are other options as well.

Poke around Pro Football Reference, Football Outsiders, Advanced NFL Stats a bit.

 

[Yakuza Rich]

If the goal is to have a statistic that is predictive of winning then there is really no way that factoring in fumbles would muddy the waters.

That muddies the waters only slightly. Most turnovers in the league are interceptions. Where things get muddied is when a QB throws a 3 yard pass for a TD versus a QB that goes 6 for 7 on a drive and doesn't get a TD score.



YR

 

[AbeBeta]

That muddies the waters only slightly. Most turnovers in the league are interceptions. Where things get muddied is when a QB throws a 3 yard pass for a TD versus a QB that goes 6 for 7 on a drive and doesn't get a TD score.

YR

It really only muddies it if a player constantly does this. Over the course of a year things like that even out and you get a reasonable estimate of the player's impact on TDs in the scoring game.

And again, adding fumbles would certainly not muddy the rating if you truly want to correlate to winning. One more variable in the mix is going to strengthen the correlation.

 

[AbeBeta]

It is emphatically not true. The far simpler (pass yards + 20*(pass TD) - 45*(interceptions thrown) - sack yards)/(passing attempts + sacks) is superior.

There are other options as well.

Poke around Pro Football Reference, Football Outsiders, Advanced NFL Stats a bit.

Again, what I'm saying is that no one is providing evidence of superiority.

 

[Califan007]

If you're wanting to predict winning then yards per attempt and INT% should hold more importance than they currently do.

If you're wanting to show individual production in an efficient manner, then you more or less keep it as it is. QB rating shows quarterback efficiency moreso than anything having to do with winning.

 

[Sarge]

All I know is passer rating doesn't necessarily mean playoff wins, let alone Super Bowl Wins.

W's are the bottom line for me. .02

 

[superpunk]

The problem with the formula is that it favors more WCO and dink and dunk passing offenses. Particularly those that throw the ball a ton in the red zone. This was one of the things that Favre used to do all of the time. He would throw for a TD inside the 5-yard line. However, he was really great at being able to do that (it's hard to do than it sounds). And if we look at the QB ratings over the years, the modern QB has a much higher QB rating. That's where you get all of the 'Mark Sanchez was better than Johnny Unitas at this point in his career talk.'

Well, no...he wasn't.

Sanchez just played in a scheme that helps boost QB rating whereas Unitas had to rely on 7-step drop pass plays and threw it deeper on a more consistent basis.

I like the QB rating metric when you take my simple formula of QB Rating Differential (QBRD) and simply take a team's QB Rating and subtract their QB Rating Allowed on defense. It still leaves out some key metrics that factor in forecasting a team's success, but it's a quick-n-easy way to figure out how well a team is playing.

tbh if Sanchez could time travel and remain at the same skill level he'd probably be way better than unitas. the league as a whole is just way way way better. Roger Staubach would be a bum today.

 

[Idgit]

tbh if Sanchez could time travel and remain at the same skill level he'd probably be way better than unitas. the league as a whole is just way way way better. Roger Staubach would be a bum today.

Everybody knows you can't time-travel without your football skills diminishing due to the temporal latency problem, so this is just stupid.

 

[Nightman]

I know ESPN has come up with their QBR that takes a lot more variables into account, including late game throws and TDs and such. I don't know how much more predictive of wins it is, but I always assume the more info the better. I like the old Passer Rating, but statistical analysis is a growing field in sports.

 

[percyhoward]

Is that really true? I hear people say that all the time but I've never seen an actual analysis.

Yeah, it's really true. Two of the better-known ones who have come up with their own ratings freely admit this, and explain that finding a rating that correlates to winning is not their goal.

Does your passer rating correlate to winning?

ESPN
"QBR is meant to correlate to offense, moving the ball downfield, turning good field position into points, avoiding giving it back to the defense. Good offensive performance is not the same as winning. To the degree that offense correlates to winning, QBR should be helpful."

PFF
"You're looking for statistics to correlate well to winning, where we are trying to find the best methods to evaluate players, and I don't believe those two are the same thing."

 

[Yakuza Rich]

It really only muddies it if a player constantly does this. Over the course of a year things like that even out and you get a reasonable estimate of the player's impact on TDs in the scoring game.

And again, adding fumbles would certainly not muddy the rating if you truly want to correlate to winning. One more variable in the mix is going to strengthen the correlation.

And in today's game that does happen frequently. Favre is a good example. So were Andy Reid's offenses in Philly. The short passes allowed for a higher completion rate and throwing the ball in the end zone while a superior QB who passes the ball more effectively has a lesser QB rating if they throw the ball less inside the 10 yard line and throws longer passes.

Aikman was a perfect example of a QB that the scheme hurt his QB rating. Long passes and we ran the ball with Emmitt inside the 10 yard line. Meanwhile QB's like Chris Miller could achieve a better QB rating thru their scheme. Do I believe the Falcons were more effective passing the ball than Dallas in the early 90's?

Not in your life.




YR

 

[percyhoward]

It is emphatically not true. The far simpler (pass yards + 20*(pass TD) - 45*(interceptions thrown) - sack yards)/(passing attempts + sacks) is superior.

Not a passer rating per se, this is actually just the formula for adjusted net yards per attempt.

In the largest sample I've found (22 seasons) comparing the two stats, the difference in the correlation coefficients is .04, with ANY/A edging out passer rating. (This is more or less the result in studies with smaller samples too).

How significant is that .04 in practical use when comparing stats to measure the Cowboys' rankings with Romo? Not very.
Cowboys Ranks in Passer Rating, 2006-12
Passer Rtg: 8th, 3rd, 13th, 6th, 12th, 4th, 9th 7.9 avg.
Adj Net YPA: 7th, 3rd, 12th, 4th, 11th, 5th, 11th 7.6 avg.

Whichever method you use, during the Romo era, Dallas has ranked between 7th and 8th, and closer to 8th.