Quarterback Rating

Galian Beast

Well-Known Member
Messages
14,735
Reaction score
7,457
Quarterback rating is such a difficult stat. While I think it is still the best stat available to us and superior to QBR, it's clear that QB Rating leaves too much information out of it.

If we were to create our own rating system what would it look like?

Would it be weighted? i.e. Does a QB with two passing touchdowns against the best pass defense in the NFL equate to a QB with two passing touchdowns against the worst pass defense in the NFL?

Would efficiency and effectiveness on 3rd downs, 2-minute warnings in both halves, and the 4th quarter be weighed differently as well?

How would turnovers be weighed? Are interceptions as bad as fumbles lost? Does it matter how much field position is lost? If a QB throws a hail mary on 4th down and the position the opposing team takes is equal to that of a punt, should that count against their stats versus a fumble picked up on the QBs 20? What about a pick 6?

What about a QB who takes a sack on 3rd down pushing you out of field goal range?

I think we need situational based statistics rather than stats from 1950. I'm not sure how we're in such a data-driven world and we're still using such simplistic measurables. Am I alone here?
 
Another thing that I hate that we don't look at:

The team is on 3rd and 30 and the defense gives them 25 yards. This is a failed play, yet we would reward it statistically.
 
Quarterback rating is such a difficult stat. While I think it is still the best stat available to us and superior to QBR, it's clear that QB Rating leaves too much information out of it.

If we were to create our own rating system what would it look like?

Would it be weighted? i.e. Does a QB with two passing touchdowns against the best pass defense in the NFL equate to a QB with two passing touchdowns against the worst pass defense in the NFL?

Would efficiency and effectiveness on 3rd downs, 2-minute warnings in both halves, and the 4th quarter be weighed differently as well?

How would turnovers be weighed? Are interceptions as bad as fumbles lost? Does it matter how much field position is lost? If a QB throws a hail mary on 4th down and the position the opposing team takes is equal to that of a punt, should that count against their stats versus a fumble picked up on the QBs 20? What about a pick 6?

What about a QB who takes a sack on 3rd down pushing you out of field goal range?

I think we need situational based statistics rather than stats from 1950. I'm not sure how we're in such a data-driven world and we're still using such simplistic measurables. Am I alone here?
Based on a 10-season study, there's a combination of EPA, CPOE, and Success Rate (basically, all those things you mentioned) that far outperforms the predictive ability of ESPN's QBR.

As with most QB metrics, the larger the sample, the more the resulting rankings resemble traditional passer rating.
 
I'm thinking score and win/loss record are still the best indicators because they take into account all the intangibles that cannot be measured by statistics.
A wise man once said "you are what your record says you are."
.
 
Last edited:
I'm thinking score and win/loss record are still the best indicators because they take into account all the intangibles that cannot be measured by statistics.
A wise man once said "you are what your record says you are."
As a team, yes. But team stats like W-L don't even give a clear picture of each unit's strengths/weaknesses, let alone each player's.

Cowboys NFL Rank
Points per drive

offense 2.64 (3rd)
defense 1.77 (10th)

Drive Success Rate
offense 77.8% (2nd)
defense 70.9% (18th)

T.O.P. per drive
offense 2:52 (12th)
defense 2:59 (24th)


Drive success rate is the percentage of down series that result in a first down.
 
Quarterback rating is such a difficult stat. While I think it is still the best stat available to us and superior to QBR, it's clear that QB Rating leaves too much information out of it.

If we were to create our own rating system what would it look like?

Would it be weighted? i.e. Does a QB with two passing touchdowns against the best pass defense in the NFL equate to a QB with two passing touchdowns against the worst pass defense in the NFL?

Would efficiency and effectiveness on 3rd downs, 2-minute warnings in both halves, and the 4th quarter be weighed differently as well?

How would turnovers be weighed? Are interceptions as bad as fumbles lost? Does it matter how much field position is lost? If a QB throws a hail mary on 4th down and the position the opposing team takes is equal to that of a punt, should that count against their stats versus a fumble picked up on the QBs 20? What about a pick 6?

What about a QB who takes a sack on 3rd down pushing you out of field goal range?

I think we need situational based statistics rather than stats from 1950. I'm not sure how we're in such a data-driven world and we're still using such simplistic measurables. Am I alone here?

Oh passer rating is without question worse than QBR. Passer rating is QBR without context.
 
QB Rating (20%) + Win % (40 %) + Total Points (40%).
 
How is it fair to tank a single player by the teams win percentage? Not that it happens often, but a QB can cave a damn near perfect game and the team still loses.

You can't flat out rate them by win percentage, imho
 
Mahomes in his final season at Tech passed for 5k yards 41 TD’s to only 10 picks. Yet his team went 5-7. QB’s need help. Did that make him a bad QB?
 
Another thing that I hate that we don't look at:

The team is on 3rd and 30 and the defense gives them 25 yards. This is a failed play, yet we would reward it statistically.

In terms of QBR, that throw wouldn’t be weighted higher than a 12 yard throw that resulted in a first down on 3rd and 10.

QBR isn’t just about yardage. It takes into account every situation and down/distance
 
Quarterback rating is such a difficult stat. While I think it is still the best stat available to us and superior to QBR, it's clear that QB Rating leaves too much information out of it.

If we were to create our own rating system what would it look like?

Would it be weighted? i.e. Does a QB with two passing touchdowns against the best pass defense in the NFL equate to a QB with two passing touchdowns against the worst pass defense in the NFL?

Would efficiency and effectiveness on 3rd downs, 2-minute warnings in both halves, and the 4th quarter be weighed differently as well?

How would turnovers be weighed? Are interceptions as bad as fumbles lost? Does it matter how much field position is lost? If a QB throws a hail mary on 4th down and the position the opposing team takes is equal to that of a punt, should that count against their stats versus a fumble picked up on the QBs 20? What about a pick 6?

What about a QB who takes a sack on 3rd down pushing you out of field goal range?

I think we need situational based statistics rather than stats from 1950. I'm not sure how we're in such a data-driven world and we're still using such simplistic measurables. Am I alone here?
I definitely wouldnt factor in who you are facing

I think td passes counts for too much. I used to hate how Aikman and Irvin would consistently do so much just for Emmit to pop it in for a 1 yard score. I used to think thats actually going to hurt Aikmans quaterback rating
 
Quarterback rating is such a difficult stat. While I think it is still the best stat available to us and superior to QBR, it's clear that QB Rating leaves too much information out of it.

If we were to create our own rating system what would it look like?

Would it be weighted? i.e. Does a QB with two passing touchdowns against the best pass defense in the NFL equate to a QB with two passing touchdowns against the worst pass defense in the NFL?

Would efficiency and effectiveness on 3rd downs, 2-minute warnings in both halves, and the 4th quarter be weighed differently as well?

How would turnovers be weighed? Are interceptions as bad as fumbles lost? Does it matter how much field position is lost? If a QB throws a hail mary on 4th down and the position the opposing team takes is equal to that of a punt, should that count against their stats versus a fumble picked up on the QBs 20? What about a pick 6?

What about a QB who takes a sack on 3rd down pushing you out of field goal range?

I think we need situational based statistics rather than stats from 1950. I'm not sure how we're in such a data-driven world and we're still using such simplistic measurables. Am I alone here?
You have no clue how either one works.

You should go back to some on more on your level of understanding.

 
Quarterback rating is such a difficult stat. While I think it is still the best stat available to us and superior to QBR, it's clear that QB Rating leaves too much information out of it.

If we were to create our own rating system what would it look like?

Would it be weighted? i.e. Does a QB with two passing touchdowns against the best pass defense in the NFL equate to a QB with two passing touchdowns against the worst pass defense in the NFL?

Would efficiency and effectiveness on 3rd downs, 2-minute warnings in both halves, and the 4th quarter be weighed differently as well?

How would turnovers be weighed? Are interceptions as bad as fumbles lost? Does it matter how much field position is lost? If a QB throws a hail mary on 4th down and the position the opposing team takes is equal to that of a punt, should that count against their stats versus a fumble picked up on the QBs 20? What about a pick 6?

What about a QB who takes a sack on 3rd down pushing you out of field goal range?

I think we need situational based statistics rather than stats from 1950. I'm not sure how we're in such a data-driven world and we're still using such simplistic measurables. Am I alone here?
1. The most difficult thing about Passer Rating is that nobody ever gets its name right. (It's Passer Rating and not Quarterback Rating for a reason).

2. Everything you're talking about exists. There's Football Outsiders' DVOA and DYAR. There's Expected Points Added, which, is there a place to find that easily?
 

Staff online

Latest posts

Forum statistics

Threads
474,003
Messages
14,505,658
Members
24,207
Latest member
TomGiantsfan
Back
Top