To all of the EJ bashers

[FireChan]

 

Invalid no, but it does make it so they can give a hand up to who they feel like. Example, Tom Brady is 22nd on the list with a 2-1 record and brees is 7th with a 1-2 record.

 

I know there are other factors but that's just going by qbr.

 

Brady put up 12 against the Raiders. Brees played way better in his games. I don't see the problem.

[Orton's Arm]

I think the QBR vs. the old rating is an interesting debate. The new system definitely factors in some important elements that had previously been ignored. However, the thing about the old rating is that while it may not be great for one or two games, it's fabulous for an entire season. Outside of points scored and given up, is there a better predictor of success than the differential between a team's offensive and defensive (old-style) QB ratings?

 

To be honest, I'm not a fan of the old quarterback rating system. John Elway's career quarterback rating was 79.9. Kelly Holcomb's was 79.2. Holcomb emphasized short, high percentage passes; thereby boosting his completion percentage. Elway didn't. Quarterback rating is going to make Holcomb look better than he should vis-a-vis Elway; because it takes completion percentage into account.

 

Jim Kelly had a career quarterback rating of 84.4. Rob Johnson's career quarterback rating was 83.6. A QB who throws the ball away to avoid a sack hurts his rating; whereas the guy who takes the sack protects his QB rating.

 

A third problem with quarterback rating is that a 5 yard pass + 45 yards of YAC is treated exactly the same as a 50 yard pass.

 

YAC (yards after the catch) is an interesting dilemma for any quarterback rating system. Joe Montana would hit his receivers in perfect stride. A five yard pass could turn into a nine or ten yard gain. I could easily see giving Montana credit for a significant portion of that YAC. But if a QB throws a screen pass to a stationary WR, and if that WR then rips off 45 yards of YAC, it's not clear why that YAC should be treated as though it was air yards. (Which is how quarterback rating treats it.) Unlike quarterback rating, QBR attempts to differentiate between YAC for which the QB was at least partially responsible, and YAC that was purely the result of the efforts of the receiver and his blockers.

 

Unlike quarterback rating, QBR takes down and distance into account. Suppose a quarterback completes a 5 yard pass on 3rd and ten. That completion will be pretty good for his quarterback rating. But it will hurt his QBR.

Edited September 27, 2014 by Edwards' Arm

[Kelly the Dog]

Yes, I know...I'm having a moment of hopefulness. It'll pass in about 15 hours, at the latest.

I assume when you say 15 hours you mean 15 minutes, yes?

[Boolay]

 

 

To be honest, I'm not a fan of the old quarterback rating system. John Elway's career quarterback rating was 79.9. Kelly Holcomb's was 79.2. Holcomb emphasized short, high percentage passes; thereby boosting his completion percentage. Elway didn't. Quarterback rating is going to make Holcomb look better than he should vis-a-vis Elway; because it takes completion percentage into account.

 

Jim Kelly had a career quarterback rating of 84.4. Rob Johnson's career quarterback rating was 83.6. A QB who throws the ball away to avoid a sack hurts his rating; whereas the guy who takes the sack protects his QB rating.

 

A third problem with quarterback rating is that a 5 yard pass + 45 yards of YAC is treated exactly the same as a 50 yard pass.

 

YAC (yards after the catch) is an interesting dilemma for any quarterback rating system. Joe Montana would hit his receivers in perfect stride. A five yard pass could turn into a nine or ten yard gain. I could easily see giving Montana credit for a significant portion of that YAC. But if a QB throws a screen pass to a stationary WR, and if that WR then rips off 45 yards of YAC, it's not clear why that YAC should be treated as though it was air yards. (Which is how quarterback rating treats it.) Unlike quarterback rating, QBR attempts to differentiate between YAC for which the QB was at least partially responsible, and YAC that was purely the result of the efforts of the receiver and his blockers.

 

Unlike quarterback rating, QBR takes down and distance into account. Suppose a quarterback completes a 5 yard pass on 3rd and ten. That completion will be pretty good for his quarterback rating. But it will hurt his QBR.

 

You are way too smart and logical for this thread. We only care about cherry picking statistics and timeframes that make EJ look good. I actually saw some nitwit post the following on the chat. "Haters can no longer complain about EJ, he has a 90 passer rating in TODAYS game" . Do I need to dissect how patently absurd that statement is? It is akin to saying that we can't complain about a lifetime .200 hitter in baseball cause he is 2-4 today which makes him a .500 hitter on the day.

[thebandit27]

 

 

You are way too smart and logical for this thread. We only care about cherry picking statistics and timeframes that make EJ look good. I actually saw some nitwit post the following on the chat. "Haters can no longer complain about EJ, he has a 90 passer rating in TODAYS game" . Do I need to dissect how patently absurd that statement is? It is akin to saying that we can't complain about a lifetime .200 hitter in baseball cause he is 2-4 today which makes him a .500 hitter on the day.

 

There's a fair amount if cherry picking coming from both sides.

 

One could take a bottom line stance on EJ, and simply say he's 6-7 as a starter, 2-1 right now, and coming off a rough outing, so we'll see how he responds...

 

...then again what fun would that be?

[The Wiz]

Brady put up 12 against the Raiders. Brees played way better in his games. I don't see the problem.

I agree with Brees having better games but some of the metrics they use that skew the numbers. Staying with these two guys vs vikings. Brees had a huge game 27/35, 293 yards and 77% completion. Passer rating of 120.3 but a QBR of 79.8. Brady against the vikings went 14/21, 149 yards and 66% completion. Passer rating came out to be 103 but a QBR of 86.5.

 

That's my issue with the "clutch index" and "probability to win" based on the passing game. If you are winning 24-7 and throw a 4 yard dump off it increases the probability to win. I understand it should increase because keeping the ball when you're winning increases your chances of finishing the game that way. But then compounding the probability to win with the clutch factor when you're already winning increases the total QBR.

 

 

The QB isn't doing anything more or less than he has all game but is getting additional points because of it. JMO.

 

EDIT: Also Brees threw 2 TD, 0 INTs and Brady threw 1 TD, 0 INTs.

Edited September 27, 2014 by The Wiz

[Boolay]

All these fabricated QB ratings are useless. Bottom line you want your QB to get yards passing, throw TD's, avoid interceptions and generate first downs on passing situations consistently. Any system that rewards a QB for a 2 yd pass on 3rd and 9 is worthless.

[FireChan]

 

I agree with Brees having better games but some of the metrics they use that skew the numbers. Staying with these two guys vs vikings. Brees had a huge game 27/35, 293 yards and 77% completion. Passer rating of 120.3 but a QBR of 79.8. Brady against the vikings went 14/21, 149 yards and 66% completion. Passer rating came out to be 103 but a QBR of 86.5.

 

That's my issue with the "clutch index" and "probability to win" based on the passing game. If you are winning 24-7 and throw a 4 yard dump off it increases the probability to win. I understand it should increase because keeping the ball when you're winning increases your chances of finishing the game that way. But then compounding the probability to win with the clutch factor when you're already winning increases the total QBR.

 

 

The QB isn't doing anything more or less than he has all game but is getting additional points because of it. JMO.

 

EDIT: Also Brees threw 2 TD, 0 INTs and Brady threw 1 TD, 0 INTs.

 

Who won?

 

Like I said, personally I would tinker with it a little bit. Because it does have it's flaws . Still, I like how it attempts to be contextually relevant, rather than rewarding dump offs for 5 yards on 3rd and 12.

Edited September 27, 2014 by FireChan

[Orton's Arm]

There's a fair amount if cherry picking coming from both sides.

 

One could take a bottom line stance on EJ, and simply say he's 6-7 as a starter, 2-1 right now, and coming off a rough outing, so we'll see how he responds...

 

...then again what fun would that be?

 

There was a 16 game stretch during which the Ravens went 15-1, with Trent Dilfer at the helm. Those 15 wins included playoff wins and a Super Bowl championship. On the other hand, there were plenty of times when the Broncos went 7-9 or 6-10 with John Elway under center. I do not believe that Trent Dilfer was a better quarterback than Elway. Nor do I believe that he played at an above-Elway level during that 16 game stretch. What I do believe is that the Ravens were much better at non-QB positions than most of the Broncos teams for which Elway played. Citing a QB's "record as a starter" ignores the fact that football is a team sport, and that non-QBs can also affect the outcome of games.

 

I agree with Brees having better games but some of the metrics they use that skew the numbers. Staying with these two guys vs vikings. Brees had a huge game 27/35, 293 yards and 77% completion. Passer rating of 120.3 but a QBR of 79.8. Brady against the vikings went 14/21, 149 yards and 66% completion. Passer rating came out to be 103 but a QBR of 86.5.

 

That's my issue with the "clutch index" and "probability to win" based on the passing game. If you are winning 24-7 and throw a 4 yard dump off it increases the probability to win. I understand it should increase because keeping the ball when you're winning increases your chances of finishing the game that way. But then compounding the probability to win with the clutch factor when you're already winning increases the total QBR.

 

 

The QB isn't doing anything more or less than he has all game but is getting additional points because of it. JMO.

 

EDIT: Also Brees threw 2 TD, 0 INTs and Brady threw 1 TD, 0 INTs.

 

I agree that on the surface, it seems odd that Brady's QBR against the Vikings was higher than Brees'. I'd have to dig deeper into the stats before concluding why this was the case. If (for example) Brady was much better at converting third downs, or avoiding fumbles, or something of that sort, then his higher QBR would not necessarily reflect a flaw in that metric. But if Brady was merely being rewarded for the fact his team was in the lead, then it probably does demonstrate a need to improve QBR.

 

My impression is that the much-discussed "clutch factor" weights how heavily each play is graded. If the game is on the line, your plays will count three times as much as plays in garbage time. Also, if your team has pretty much won the game, then that reduces the "clutch factor" associated with whatever short, safe passes you throw after your team has secured victory. Meaning your early game plays (when both teams still had a chance of winning) are more heavily weighted than your late game plays.

 

But if you're still not sold on QBR, there's another, much simpler statistic which reveals many of the same things as QBR. This statistic is air yards per attempt. (After clicking on that link, it will be necessary to sort the list by air yards per attempt.)

 

Air yards per attempt isn't the right tool with which to measure every quarterback. Suppose that Joe Montana throws a 5 yard pass to Jerry Rice, hitting him in perfect stride. Because Rice didn't have to slow down and wait for the ball, he was able to get 4 yards of yards after the catch (YAC). Montana deserves his share of credit for that YAC; and air yards per attempt wouldn't give him any of that credit.

 

But let's say a QB doesn't have Montana's gift for hitting receivers in perfect stride. In that case, air yards per attempt is a perfectly valid measurement tool. It strips away the things the receiver is able to do on his own, after the catch. It gives the QB credit only for the portion of the play for which he was personally responsible.

 

67.7% of Manuel's total passing yards have come because of YAC. That's the third highest percentage of any starting quarterback in the league. Through the first three games of the season, Manuel is averaging 2.28 air yards per attempt--the third worst average of any starting quarterback. There are eleven quarterbacks averaging over four air yards per attempt. One of whom is Ryan Fitzpatrick.

 

Manuel is not currently playing like a top-25 QB--at least not if his stats through the first three games are any indication. I assume he'll be given at least the rest of this season to try to improve.

[K-9]

There's a fair amount if cherry picking coming from both sides.

 

One could take a bottom line stance on EJ, and simply say he's 6-7 as a starter, 2-1 right now, and coming off a rough outing, so we'll see how he responds...

 

...then again what fun would that be?

There's a fair amount if cherry picking coming from both sides.

 

One could take a bottom line stance on EJ, and simply say he's 6-7 as a starter, 2-1 right now, and coming off a rough outing, so we'll see how he responds...

 

...then again what fun would that be?

 

More reasoned thought, less fun. That's fine with me.

 

GO BILLS!!!

[San-O]

You are way too smart and logical for this thread. We only care about cherry picking statistics and timeframes that make EJ look good. I actually saw some nitwit post the following on the chat. "Haters can no longer complain about EJ, he has a 90 passer rating in TODAYS game" . Do I need to dissect how patently absurd that statement is? It is akin to saying that we can't complain about a lifetime .200 hitter in baseball cause he is 2-4 today which makes him a .500 hitter on the day.

 

To anyone, watch the first half of the last game. Small sample, but good representation IMO.

[thebandit27]

 

 

To anyone, watch the first half of the last game. Small sample, but good representation IMO.

 

Serious question: why is that half of football a good representation in comparison to the other 25 halves he's played in the NFL?

[thebandit27]

 

 

There was a 16 game stretch during which the Ravens went 15-1, with Trent Dilfer at the helm. Those 15 wins included playoff wins and a Super Bowl championship. On the other hand, there were plenty of times when the Broncos went 7-9 or 6-10 with John Elway under center. I do not believe that Trent Dilfer was a better quarterback than Elway. Nor do I believe that he played at an above-Elway level during that 16 game stretch. What I do believe is that the Ravens were much better at non-QB positions than most of the Broncos teams for which Elway played. Citing a QB's "record as a starter" ignores the fact that football is a team sport, and that non-QBs can also affect the outcome of games

 

So we cannot use record as a starter to assess whether or not the QB is playing well enough to win, despite the fact that the team's record is, quite literally, the one stat that tells us that?

 

However, we CAN use QBR, the same stat that tells me that Derek Anderson is the best QB in the NFL, and that Austin Davis is better than Aaron Rodgers?

 

Gotcha.

 

Seriously folks, the total abandonment of reason for the sake of picking a side and hoping to be right is higher with regard to EJ than with any other player I've ever seen discussed on this board.

 

It's crazy.

 

I won't be move

 

 

 

There was a 16 game stretch during which the Ravens went 15-1, with Trent Dilfer at the helm. Those 15 wins included playoff wins and a Super Bowl championship. On the other hand, there were plenty of times when the Broncos went 7-9 or 6-10 with John Elway under center. I do not believe that Trent Dilfer was a better quarterback than Elway. Nor do I believe that he played at an above-Elway level during that 16 game stretch. What I do believe is that the Ravens were much better at non-QB positions than most of the Broncos teams for which Elway played. Citing a QB's "record as a starter" ignores the fact that football is a team sport, and that non-QBs can also affect the outcome of games

 

So we cannot use record as a starter to assess whether or not the QB is playing well enough to win, despite the fact that the team's record is, quite literally, the one stat that tells us that?

 

However, we CAN use QBR, the same stat that tells me that Derek Anderson is the best QB in the NFL, and that Austin Davis is better than Aaron Rodgers?

 

Gotcha.

 

Seriously folks, the total abandonment of reason for the sake of picking a side and hoping to be right is higher with regard to EJ than with any other player I've ever seen discussed on this board.

 

It's crazy.

 

I won't be distracted from the facts:

 

He's started 13 games, and is 6-7 in those games

He's had a few good showings, a few bad ones, and several mixed bags

He has followed up his worst outings with solid games each time

This season, EJ has played well enough for the team to win 2 of 3 games--this isn't up for debate

 

Let's see if the trends continue.

[San-O]

Serious question: why is that half of football a good representation in comparison to the other 25 halves he's played in the NFL?

 

Well, first let me say after posting and reading and watching the Bills and E.J. play, most people are going to see whatever they want to see.

 

I believe it shows what his game is about, or how the Bills are using him anyway.

 

o Mostly very short or behind the LOS passes.

o Accuracy issues.

o Doesn't see the field well. Commentators point this out several times, not that they're always right.

o Bad reads.

o Has good running ability: that one play where he was almost certainly sacked showed some great athleticism. That play in particular was interesting to me as it was his only completion of the 1st half over 2 or 3 yards. All others completed passes were at most 2 yards past the LOS or behind the LOS.

 

 

Also, there were almost no mid/intermediate passes with the exception of the one over the middle which was at the turf/feet of Watkins I believe.

 

He did throw 1 nice long ball to Goodwin on the right side deep, which was not caught, and another left side deep which was almost picked off

had Watkins again I believe played a little DB.

 

This is what I am seeing anyhow.

Edited September 28, 2014 by San-O

[Orton's Arm]

So we cannot use record as a starter to assess whether or not the QB is playing well enough to win, despite the fact that the team's record is, quite literally, the one stat that tells us that?

 

However, we CAN use QBR, the same stat that tells me that Derek Anderson is the best QB in the NFL, and that Austin Davis is better than Aaron Rodgers?

 

Gotcha.

 

Seriously folks, the total abandonment of reason for the sake of picking a side and hoping to be right is higher with regard to EJ than with any other player I've ever seen discussed on this board.

 

It's crazy.

 

I won't be move

 

 

 

So we cannot use record as a starter to assess whether or not the QB is playing well enough to win, despite the fact that the team's record is, quite literally, the one stat that tells us that?

 

However, we CAN use QBR, the same stat that tells me that Derek Anderson is the best QB in the NFL, and that Austin Davis is better than Aaron Rodgers?

 

Gotcha.

 

Seriously folks, the total abandonment of reason for the sake of picking a side and hoping to be right is higher with regard to EJ than with any other player I've ever seen discussed on this board.

 

It's crazy.

 

I won't be distracted from the facts:

 

He's started 13 games, and is 6-7 in those games

He's had a few good showings, a few bad ones, and several mixed bags

He has followed up his worst outings with solid games each time

This season, EJ has played well enough for the team to win 2 of 3 games--this isn't up for debate

 

Let's see if the trends continue.

 

> So we cannot use record as a starter to assess whether or not the QB is playing well enough to win, despite the fact that the team's record is, quite literally, the one stat that tells us that?

 

I've been posting here since 2005. You can go back through any of the posts I've written over these years, to see if I've ever, under any circumstances, used "record as a starter" to evaluate a QB's play. Not once did I do that.

 

If a FG kicker goes one for four, and his team wins, nobody pats him on the back afterward and tells him he "played well enough to win." If a RB averages 1.5 yards per carry over the course of the game, and his team wins, nobody celebrates that RB for having played well enough to win. If a DE gets five sacks during a game, as well as numerous tackles and even a forced fumble, but his team loses, nobody is going to accuse him of having played like chopped liver. Players at all these other positions are evaluated on their individual stats--their individual contributions. Why should quarterbacks be the exception? Why should a QB's individual stats be ignored?

 

If you're using wins as a measurement of QB play, what conclusions are you drawing about Trent Dilfer's 15-1 stretch with the Ravens? The Ravens of 2000 had one of the three best defenses in NFL history. Is that fact relevant to the discussion of how those 15 wins were achieved? Or does that 15-1 record represent the beginning, middle, and end of any discussion of how well Dilfer played?

 

> However, we CAN use QBR, the same stat that tells me that Derek Anderson is the best QB in the NFL, and that Austin Davis is better than Aaron Rodgers?

 

We are only three games into the season. There are going to be statistical anomalies in almost any statistical measure you use. Austin Davis has faced three below-average defenses. That's going to make his stats look better than they should. Over the course of an entire season, statistical anomalies like that tend to get smoothed out. By the end of the year, the ranking of QBs by QBR should look more reasonable than it does at the moment.

 

> Seriously folks, the total abandonment of reason for the sake of picking a side and hoping to be right . . .

 

It's interesting you should say that. After last season, I made the argument that Thad Lewis had outplayed Manuel. Thad Lewis had a significantly higher yards per attempt. At the time, that was my favorite stat. But then someone countered by pointing out that Manuel's QBR was higher than Lewis's. I hadn't previously heard of QBR. Because it was a (previously unknown) stat, it would have been easy to dismiss it with a wave of the hand. Had my only goal been to make Manuel look bad, that's exactly what I would have done.

 

But doing so would not have been accurate or intellectually honest. I looked into QBR, and decided it had potential. I retracted my earlier comments about Thad Lewis having played better than Manuel. With yards per attempt making Lewis look better than Manuel, and QBR making Manuel look better than Lewis, it was no longer clear to me which QB had had the statistically better season. (As an aside, Lewis's yards per attempt made him look like a solid backup/decent starter. His QBR made him look like chopped liver. Without knowing which statistical indicator was more valid, it was hard to form a firm opinion of the guy. This year's preseason, and the end of his time in Buffalo, pretty much resolved that dilemma.)

 

But if Lewis's low QBR was a more accurate indicator of his play than his shiny yards per attempt; then what does that say about Manuel's low QBR for the first three games of this season? Granted, we're just three games in, so I don't want to draw hard and fast conclusions from any statistical indicators. But Manuel's supporters are claiming he's significantly progressed since last season. The statistics aren't there to support that. Not QBR, and not air yards per attempt.

[K-9]

... Not QBR, and not air yards per attempt.

 

Perhaps the most ridiculous statistic ever compiled. Why do you suppose that there are 28 passers rated higher than Russel Wilson in that category? Or 29 better "air yard" QBs than Dalton? This stat is the height of laziness and ignorance of the teams in the league and who their players are.

 

GO BILLS!!!

[FireChan]

> So we cannot use record as a starter to assess whether or not the QB is playing well enough to win, despite the fact that the team's record is, quite literally, the one stat that tells us that?

 

I've been posting here since 2005. You can go back through any of the posts I've written over these years, to see if I've ever, under any circumstances, used "record as a starter" to evaluate a QB's play. Not once did I do that.

 

If a FG kicker goes one for four, and his team wins, nobody pats him on the back afterward and tells him he "played well enough to win." If a RB averages 1.5 yards per carry over the course of the game, and his team wins, nobody celebrates that RB for having played well enough to win. If a DE gets five sacks during a game, as well as numerous tackles and even a forced fumble, but his team loses, nobody is going to accuse him of having played like chopped liver. Players at all these other positions are evaluated on their individual stats--their individual contributions. Why should quarterbacks be the exception? Why should a QB's individual stats be ignored?

 

If you're using wins as a measurement of QB play, what conclusions are you drawing about Trent Dilfer's 15-1 stretch with the Ravens? The Ravens of 2000 had one of the three best defenses in NFL history. Is that fact relevant to the discussion of how those 15 wins were achieved? Or does that 15-1 record represent the beginning, middle, and end of any discussion of how well Dilfer played?

 

> However, we CAN use QBR, the same stat that tells me that Derek Anderson is the best QB in the NFL, and that Austin Davis is better than Aaron Rodgers?

 

We are only three games into the season. There are going to be statistical anomalies in almost any statistical measure you use. Austin Davis has faced three below-average defenses. That's going to make his stats look better than they should. Over the course of an entire season, statistical anomalies like that tend to get smoothed out. By the end of the year, the ranking of QBs by QBR should look more reasonable than it does at the moment.

 

> Seriously folks, the total abandonment of reason for the sake of picking a side and hoping to be right . . .

 

It's interesting you should say that. After last season, I made the argument that Thad Lewis had outplayed Manuel. Thad Lewis had a significantly higher yards per attempt. At the time, that was my favorite stat. But then someone countered by pointing out that Manuel's QBR was higher than Lewis's. I hadn't previously heard of QBR. Because it was a (previously unknown) stat, it would have been easy to dismiss it with a wave of the hand. Had my only goal been to make Manuel look bad, that's exactly what I would have done.

 

But doing so would not have been accurate or intellectually honest. I looked into QBR, and decided it had potential. I retracted my earlier comments about Thad Lewis having played better than Manuel. With yards per attempt making Lewis look better than Manuel, and QBR making Manuel look better than Lewis, it was no longer clear to me which QB had had the statistically better season. (As an aside, Lewis's yards per attempt made him look like a solid backup/decent starter. His QBR made him look like chopped liver. Without knowing which statistical indicator was more valid, it was hard to form a firm opinion of the guy. This year's preseason, and the end of his time in Buffalo, pretty much resolved that dilemma.)

 

But if Lewis's low QBR was a more accurate indicator of his play than his shiny yards per attempt; then what does that say about Manuel's low QBR for the first three games of this season? Granted, we're just three games in, so I don't want to draw hard and fast conclusions from any statistical indicators. But Manuel's supporters are claiming he's significantly progressed since last season. The statistics aren't there to support that. Not QBR, and not air yards per attempt.

 

The disconnect in this argument is that some turn a blind eye towards the statistical anomalies in their chosen stat, and disregard other stats based on their statistical anomalies.

[GG]

 

 

The disconnect in this argument is that some turn a blind eye towards the statistical anomalies in their chosen stat, and disregard other stats based on their statistical anomalies.

 

I think it happens 3.5 times in a thread.

[thebandit27]

 

 

The disconnect in this argument is that some turn a blind eye towards the statistical anomalies in their chosen stat, and disregard other stats based on their statistical anomalies.

 

Confirmation bias at its finest.

[The Wiz]

 

 

I think it happens 3.5 times in a thread.

More like every 3.5 posts in this thread.