QBR

[rochrist]

I've long hated the QBR, viewing it as ESPN's pathetic attempt to have a rating system to call their own. Plus, it's largely subjective.

But this, THIS stands out. Stroud's QBR yesterday was BETTER than Maye's. 21.1 to 20.8.

Meanwhile, via the ordinary QB rating system, Stroud was 28.0 versus Maye 100.7

 

[Pat the Pats Fan]

Well he did throw for more yards, and had more completions if you count the interceptions as completions.

 

[jmt57]

The fact that ESPN has never revealed the actual formula, referring to it as 'proprietary' as if it is some intellectual property that would cause them financial harm if revealed - the lack of transparency results in suspicion of bias and manipulation.

QBR gives a boost to quarterbacks who run and scramble, which in itself is questionable. I stopped looking at it early on when losing QBs with mediocre performances had higher QBRs than their opponent, who played a better game both statistically as well as by the eye test. For the conspiracy-minded, QBR came out at a time when Brady was being lauded, and tended to have low QBRs because he wasn't a running quarterback.

Charlie Batch's 186-yard, two-pick game has ESPN's best QBR ever

There's a risk of beating a dead horse when you talk about QBR, ESPN's proprietary statistic that was rolled out with great fanfare four years ago as the “one stat that measures the totality of a quarterback’s performance.”

www.nbcsports.com

But today I decided to dig a little deeper into what is, according to QBR, the greatest game any quarterback has ever played: Charlie Batch’s 186-yard, two-interception performance against the Buccaneers in 2010.

Really. Under “All-Time Best Games” on ESPN.com’s QBR page, the top game on the list is an utterly forgettable game that Batch played in place of the suspended Ben Roethlisberger in 2010. I couldn’t comprehend how 186 yards and two interceptions could add up to the greatest game ever played, with a 99.9 QBR on a scale of 0-100, but then again I didn’t remember exactly what Batch did in that game, and I know ESPN claims that QBR benefits from tape analysis that includes dropped passes and performance under pressure and other things that the traditional stats overlook. So I decided to re-watch Batch’s game and see how it looked.

Thanks to NFL.com’s Game Pass, it’s easy to go back and watch old games. So I did. And what I saw was not the greatest game any quarterback had in NFL history. Not even close. I’m sure Charlie Batch doesn’t think it was the greatest game in NFL history. I’m sure Charlie Batch’s mom doesn’t think it was the greatest game in NFL history. No sane person could possibly think it was the greatest game in NFL history. Only ESPN’s super-secret QBR formula could possibly arrive at the conclusion that it was the greatest game in NFL history.

Batch’s first pass of the game was absolutely terrible. He hit Aqib Talib in stride, right between the numbers -- which is not a good thing because Talib played for the Buccaneers. The Buccaneers’ offense couldn’t do anything with the great field position Batch gave up with his interception, but the Bucs did kick a field goal to take a 3-0 lead. Shortly after that, CBS showed a split screen of Talib and Batch and called them “the hero and the goat.”

So how did Batch go from goat to G.O.A.T., at least according to QBR? I have no earthly idea, because ESPN doesn’t make its QBR formula public. But there was certainly nothing in the rest of Batch’s play that day that suggested a “Greatest Of All Time” performance.

 

[cmasspatsfan]

It's a ridiculous system.

 

[Hammer of Thor]

The fact that ESPN has never revealed the actual formula, referring to it as 'proprietary' as if it is some intellectual property that would cause them financial harm if revealed - the lack of transparency results in suspicion of bias and manipulation.

QBR gives a boost to quarterbacks who run and scramble, which in itself is questionable. I stopped looking at it early on when losing QBs with mediocre performances had higher QBRs than their opponent, who played a better game both statistically as well as by the eye test. For the conspiracy-minded, QBR came out at a time when Brady was being lauded, and tended to have low QBRs because he wasn't a running quarterback.

Charlie Batch's 186-yard, two-pick game has ESPN's best QBR ever

There's a risk of beating a dead horse when you talk about QBR, ESPN's proprietary statistic that was rolled out with great fanfare four years ago as the “one stat that measures the totality of a quarterback’s performance.”

www.nbcsports.com

But today I decided to dig a little deeper into what is, according to QBR, the greatest game any quarterback has ever played: Charlie Batch’s 186-yard, two-interception performance against the Buccaneers in 2010.

Really. Under “All-Time Best Games” on ESPN.com’s QBR page, the top game on the list is an utterly forgettable game that Batch played in place of the suspended Ben Roethlisberger in 2010. I couldn’t comprehend how 186 yards and two interceptions could add up to the greatest game ever played, with a 99.9 QBR on a scale of 0-100, but then again I didn’t remember exactly what Batch did in that game, and I know ESPN claims that QBR benefits from tape analysis that includes dropped passes and performance under pressure and other things that the traditional stats overlook. So I decided to re-watch Batch’s game and see how it looked.

Thanks to NFL.com’s Game Pass, it’s easy to go back and watch old games. So I did. And what I saw was not the greatest game any quarterback had in NFL history. Not even close. I’m sure Charlie Batch doesn’t think it was the greatest game in NFL history. I’m sure Charlie Batch’s mom doesn’t think it was the greatest game in NFL history. No sane person could possibly think it was the greatest game in NFL history. Only ESPN’s super-secret QBR formula could possibly arrive at the conclusion that it was the greatest game in NFL history.

Batch’s first pass of the game was absolutely terrible. He hit Aqib Talib in stride, right between the numbers -- which is not a good thing because Talib played for the Buccaneers. The Buccaneers’ offense couldn’t do anything with the great field position Batch gave up with his interception, but the Bucs did kick a field goal to take a 3-0 lead. Shortly after that, CBS showed a split screen of Talib and Batch and called them “the hero and the goat.”

So how did Batch go from goat to G.O.A.T., at least according to QBR? I have no earthly idea, because ESPN doesn’t make its QBR formula public. But there was certainly nothing in the rest of Batch’s play that day that suggested a “Greatest Of All Time” performance.

When you have a proprietary model for anything, the first thing you need to do is prove that it works to whoever your buyer is. It needs to pass the smell test.

Doubly so when it’s a model that claims to improve on overall passer rating. Of all the traditional stats, passer rating, passer rating allowed, and net passer rating has the largest correlation to points scored, points allowed, and win/loss record, respectively.

QBR needs to show that it improves on it. We have no formula to work with, but there might be historical results to work with. I wonder if anyone’s tried to do that analysis with what’s out there? What I can say is that 21 points scored on offense having a lower QBR than 9 net points scored on offense (when you factor in the pick-6) does not bode well for the system.

 

[Wozzy]

Cumulative stats generally suck, DVOA is terrible also.

We can see points, turnovers and other stats… there’s no need to bunch them and give them new values. They are what they are.

 

[Nehalem]

I've long hated the QBR, viewing it as ESPN's pathetic attempt to have a rating system to call their own. Plus, it's largely subjective.

But this, THIS stands out. Stroud's QBR yesterday was BETTER than Maye's. 21.1 to 20.8.

Meanwhile, via the ordinary QB rating system, Stroud was 28.0 versus Maye 100.7

QBR might be better than QB rating in this case because Maye's 4 fumbles, 2 lost aren't factored into QBR. Maye also benefited from some great WR play (see Boutte's TD).

100.7 is basically Pro Bowl level play, especially given the defense he was up against. If you think Maye played at a Pro Bowl level yesterday... time to go to AA.

 

[Pape]

Qbr = garbage seat... iirc didn't it come out right around the time all those free lance football analytics sites (before they were all bought out then ultimately folded when the quality took a nosedive - buyers realized the fans left because they knew the difference between a work of passion and a work) [remember chff?, good stuff at least for awhile] were all touting new statistical modeling? not to be outdone, the world wide leader in sports dropped the turd known as the qbr on us... I actually wrote to profootball reference when they started adding it to their stat lines stressing them not to use it because there was no way it figure how the number was calculated... ..... they still add it in, but have curtailed its usage & offer no pop up explanation of the number...

 

[XLIX]

Here's how stupid QBR is.... I was looking up Stidham yesterday (for obvious reasons). In 2024, he was:

0-0, 0 yards, 0 TDs, 0 INTs. 4 rushes for 5 yards. QBR: 100