ESPN QBR v. FO DYAR

[ctpatsfan77]

Which list of top 10 QBs from Week 2 seems more accurate to you?

ESPN:
1 Romo DAL
2 Campbell OAK
3 Hasselbeck TEN
4 BRADY
5 Roethlisberger *IT
6 Fitzpatrick BUF
7 Freeman TB
8 Schaub HOU
9 Sanchez JEST
10 Rivers SD

Football Outsiders:
1 BRADY
2 Campbell OAK
3 Roethlisberger *IT
4 Fitzpatrick BUF
5 Romo DAL
6 Dalton CIN
7 Hasselbek SEA
8 Rivers SD
9 Rodgers GB
10 Stafford DET

 

[DarrylS]

I like the traditional NFL passer rating best... perhaps it is familiarity..

NFL Stats: by Player Category

Brady
Rodgers
Brees
Stafford
Kolb
Fitzpatrick
Romo
Dalton
Schaub
Campbell

 

[The Scrizz]

Not familiar with DYAR but QBR is not trying to measure stats, it's trying to measure a QB's contribution to the game, much the same baseball stats like WAR try to measure the affect one player has on a game.

I don't like QBR because it has subjective components that they foolishly try to gloss over, but it weights a QB differently based on the situation of the game.

Stafford doesn't even show top 10 because he gets no 'clutch' points because the game is so out of hand. It is similar to high-leverage vs. low-leverage innings for pitchers in baseball.

 

[BSR]

I like football outsiders better but I don't think they put in their defensive adjustments in until week 3 so at this point it is just YAR not DYAR.

 

[Mainefan]

I have no idea why anyone is talking about QBR anymore. It was conceived for physicists, not football fans.

 

[strngplyr]

I like the traditional NFL passer rating best... perhaps it is familiarity..

NFL Stats: by Player Category

This. I'm used to it, I like it better, and under most cases it seems accurately place the top QB's in order more often than not.

The QBR is dumb and uses opinions based on what someone thinks is important or not.

 

[Rob0729]

I hate value stats like these. I think Football Outsider's is better than ESPN's screwy value added ratings, but I still have little use for either.

 

[37Harrison]

I think the QBR is a good idea however as with most 'New' systems it has bugs in it and needs to be tweaked. I think it CAN be a very good system in the future if they work out these bugs. As of now, I still like using my eyes rather than any stat sheet.

 

[BradyFTW!]

Any system that doesn't list Brady as the top QB in the NFL over the past two weeks is inherently broken.

 

[lamafist]

Which list of top 10 QBs from Week 2 seems more accurate to you?

ESPN:
1 Romo DAL
2 Campbell OAK
3 Hasselbeck TEN
4 BRADY
5 Roethlisberger *IT
6 Fitzpatrick BUF
7 Freeman TB
8 Schaub HOU
9 Sanchez JEST
10 Rivers SD

Football Outsiders:
1 BRADY
2 Campbell OAK
3 Roethlisberger *IT
4 Fitzpatrick BUF
5 Romo DAL
6 Dalton CIN
7 Hasselbek SEA
8 Rivers SD
9 Rodgers GB
10 Stafford DET

DYAR isn't the right FO metric to compare QBR to. They're measuring different things.

QBR is a rate stat, which attempts to measure how efficient the QB was on a per play basis, while DYAR is a hybrid rate/counting stat, that measures value accrued by the quantity of performance at the efficiency rate. DVOA, FO's pure efficiency rate stat, is the more natural comparison.

So QBR and DYAR tell you different things. Take a QB who gets knocked out of a game halfway through due to injury. He's unlikely to rank high in DYAR, because he had less of an opportunity to contribute to game's outcome, whereas he'd be as likely as any QB to rank at the top in QBR, which is trying to tell you how well he contributed during however much opportunity he had.

 

[lamafist]

Not familiar with DYAR but QBR is not trying to measure stats, it's trying to measure a QB's contribution to the game, much the same baseball stats like WAR try to measure the affect one player has on a game.

I don't like QBR because it has subjective components that they foolishly try to gloss over, but it weights a QB differently based on the situation of the game.

Stafford doesn't even show top 10 because he gets no 'clutch' points because the game is so out of hand. It is similar to high-leverage vs. low-leverage innings for pitchers in baseball.

I'm not seeing all this subjectivity. I think it's a matter of misleading terminology. When you hear that they're counting "dropped passes," one imagines them deciding on a case by case basis whether the WR "should have had it" or not. But this isn't how it works. Rather, the stat-keepers are simply keeping track of whether the pass was nowhere near an intended receiver, in the vicinity of the receiver, or was touched by an intended receiver. This way, the negative change in expected points can be weighted by the percent of times a pass might have been caught -- 0% for a ball nowhere near the intended target, and, for balls that contact a receiver, a different percent depending on how often receivers catch balls they contact thrown that distance to that portion of the field.

 

[BradyFTW!]

I'm not seeing all this subjectivity. I think it's a matter of misleading terminology. When you hear that they're counting "dropped passes," one imagines them deciding on a case by case basis whether the WR "should have had it" or not. But this isn't how it works. Rather, the stat-keepers are simply keeping track of whether the pass was nowhere near an intended receiver, in the vicinity of the receiver, or was touched by an intended receiver. This way, the negative change in expected points can be weighted by the percent of times a pass might have been caught -- 0% for a ball nowhere near the intended target, and, for balls that contact a receiver, a different percent depending on how often receivers catch balls they contact thrown that distance to that portion of the field.

I think the point is that it's foolish to 'punish' Stafford for the fact that his offense has been putting games out of hand in the second/third quarter. Would he be a better QB if he sucked early in the game so that he could lead a memorable comeback later?

 

[The Scrizz]

I think the point is that it's foolish to 'punish' Stafford for the fact that his offense has been putting games out of hand in the second/third quarter. Would he be a better QB if he sucked early in the game so that he could lead a memorable comeback later?

It's not punishing Stafford, it's just not measuring QB performance which is what everyone thinks it's measuring. It's measuring contribution.

They were stupid to even mention it in the same context as passer rating because they aren't even attempting to measure the same thing.

 

[The Scrizz]

I'm not seeing all this subjectivity. I think it's a matter of misleading terminology. When you hear that they're counting "dropped passes," one imagines them deciding on a case by case basis whether the WR "should have had it" or not. But this isn't how it works. Rather, the stat-keepers are simply keeping track of whether the pass was nowhere near an intended receiver, in the vicinity of the receiver, or was touched by an intended receiver. This way, the negative change in expected points can be weighted by the percent of times a pass might have been caught -- 0% for a ball nowhere near the intended target, and, for balls that contact a receiver, a different percent depending on how often receivers catch balls they contact thrown that distance to that portion of the field.

Then it's a case of them explaining themselves piss poorly. They make it sound like they judge on a case by case basis.

"The part of QBR that could be cynically called "subjective" is that there are judgment calls with regard to what are dropped passes vs overthrows or underthrows or defended passes. ESPN's video trackers have strict guidelines on how to chart these items so that they are consistent across the different people doing charting. If you as a fan go out and chart these yourself for a game or two, you will see how several calls are easy, but some are quite hard to judge. We have standards that make things more uniform and every game is done twice to reconcile inconsistencies. Despite the standards, the gray areas will still exist and, because they exist, the division of credit quantitative analysis described below is important. That analysis is what says that a "drop" isn't necessarily all about a receiver because there are gray areas in drops."

Judgement calls are not "the receiver touched the ball but didn't catch it, let's compare that to the statistical norm of a receiver touching the ball and not catching it."

They admit themselves that there are gray areas and judgement calls. So yes it is subjective. They say nothing about "a different percent depending on how often receivers catch balls they contact thrown that distance to that portion of the field" which would be a statistically viable way to handle it.

So either you are incorrect or whoever wrote their FAQ is an imbecile. Because what you write and what they write are mutually exclusive.

 

[lamafist]

Then it's a case of them explaining themselves piss poorly. They make it sound like they judge on a case by case basis.

"The part of QBR that could be cynically called "subjective" is that there are judgment calls with regard to what are dropped passes vs overthrows or underthrows or defended passes. ESPN's video trackers have strict guidelines on how to chart these items so that they are consistent across the different people doing charting. If you as a fan go out and chart these yourself for a game or two, you will see how several calls are easy, but some are quite hard to judge. We have standards that make things more uniform and every game is done twice to reconcile inconsistencies. Despite the standards, the gray areas will still exist and, because they exist, the division of credit quantitative analysis described below is important. That analysis is what says that a "drop" isn't necessarily all about a receiver because there are gray areas in drops."

Judgement calls are not "the receiver touched the ball but didn't catch it, let's compare that to the statistical norm of a receiver touching the ball and not catching it."

They admit themselves that there are gray areas and judgement calls. So yes it is subjective. They say nothing about "a different percent depending on how often receivers catch balls they contact thrown that distance to that portion of the field" which would be a statistically viable way to handle it.

So either you are incorrect or whoever wrote their FAQ is an imbecile. Because what you write and what they write are mutually exclusive.

The guy who wrote the FAQ is the statistician Dean Oliver. He's like the Bill James of basketball. Decidedly not an imbecile. Probably could be better at communicating what he's doing in laymen's terms. His explanation is more clear and detailed in the main writeup of his methodology than in the FAQ.

 

[slam]

Romo did contribute a lot to the Jets' week 1 victory, so I guess he does deserve the #1 slot.

 

[The Scrizz]

The guy who wrote the FAQ is the statistician Dean Oliver. He's like the Bill James of basketball. Decidedly not an imbecile. Probably could be better at communicating what he's doing in laymen's terms. His explanation is more clear and detailed in the main writeup of his methodology than in the FAQ.

I'd like to read it. Layman's terms suck. I'd like to know what he's actually doing. Do you have a good link?

 

[BradyFTW!]

It's not punishing Stafford, it's just not measuring QB performance which is what everyone thinks it's measuring. It's measuring contribution.

They were stupid to even mention it in the same context as passer rating because they aren't even attempting to measure the same thing.

Which in no way invalidates my point. You're splitting hairs to make a distinction that is 100% irrelevant to what I said.

Is it true that Stafford does not rank as highly on the QBR as he otherwise would simply because the offense that he pilots is putting games out of reach early on? If this is true, then QBR is either measuring wrongly or it's measuring something dumb that doesn't warrant much thought.

Personally, I'd rather have a QB that can bury the opposition in the third quarter than one who can scrap his way to a fourth-quarter loss.

 

[Deus Irae]

QBR is a lousy stat. I thought week 1 had already made that obvious.

 

[TheSolderKing]

BOTH SUCK!!! Why???

where is CAM NEWTON? 400 plus yards and no love



clearly they made a mistake