Today's random fun fact(s) - QBR edition

[kearly]

I figured I'd put this here since it's related to Russell Wilson and the Panthers game.

Cam Newton is 8th in QBR. Russell Wilson is 10th in QBR. Ergo, Cam Newton had a better QBR in the Carolina game than Wilson did.

Peyton Manning threw 7 TDs, 0 INTs, had 11 yards per attempt and completed 64% of his passes for a total of 462 passing yards in what was one of the most prolific single game performances in NFL history. Four other QBs in week one posted a higher QBR than Manning did. Two of those QBs were Andy Dalton and Jay Cutler.

Without looking, try to guess who's #1. I'm betting you'll guess right.

 

[JSeahawks]

Kaepernick?

 

[mretrade]

Luck. CK is 3rd.

 

[Scottemojo]

Let me guess, the GB D was 33rd in QBR?

 

[RolandDeschain]

Let me guess, the GB D was 33rd in QBR?

Actually, ESPN awarded its first "worst in the NCAA - NFL edition" ranking, and put the Green Bay defense as between high school and college level. True story.

 

[Fuzzman55]

Of course it's him. Only ESPN could make up a stat that they can subjectively score to fit their narrative.

I don't wan't to see QBR used for anything anymore.

 

[Blitzer88]

Wow..........

 

[HansGruber]

Well that pretty much sums up the worthlessness of QBR as a ranking system.

Heck that might almost be as worthless and arbitrary as the BCS algorithm.

 

[Smelly McUgly]

Heck, I like QBR if only because it demonstrates that it is possible to make an even more garbage stat for evaluating QBs than the traditional passer rating.

 

[kearly]

Stat nerds miss the point about passer rating. It is intended as an approximation. It is useful and largely falls in line with both the eyeball test and smarter stats like Adjusted net yards per attempt. It is not perfect, but it is useful. I'd compare it to OPS in baseball. There are better stats for measuring value, but OPS can be calculated pretty easy and generally gives a pretty good picture of effectiveness, even if it doesn't give the whole story.

I suspect Advanced NFL Stats had a hand in birthing QBR. They are obsessed with EPA and WPA- win probability as an individual stat. Sabremetricians in baseball invented win probability, but they never seriously considered using it is a meaningful individual stat (for good reason). Neither does Football Outsiders (their half-hesitant citation of QBR aside). Just Advanced NFL Stats, and QBR. It is their undoing, and I think perfectly explains the erratic and nonsensical results those stats produce. Basically, if you make a few plays very late in the 4th quarter, even if you lose the game, those plays could outweigh a full game of superior play by someone else simply because of the leverage factor in win probability. Win probability especially jumps and plummets late in close football games, and creates a house of mirrors effect if used in statistical analysis. It also rewards QBs for playing in tighter games, and punishes QBs who blow out their opponents.

 

[RolandDeschain]

Agreed, Kearly.

DVOA's my bread & butta, though. Not perfect, but the closest to it.

 

[MizzouHawkGal]

I figured I'd put this here since it's related to Russell Wilson and the Panthers game.

Cam Newton is 8th in QBR. Russell Wilson is 10th in QBR. Ergo, Cam Newton had a better QBR in the Carolina game than Wilson did.

Peyton Manning threw 7 TDs, 0 INTs, had 11 yards per attempt and completed 64% of his passes for a total of 462 passing yards in what was one of the most prolific single game performances in NFL history. Four other QBs in week one posted a higher QBR than Manning did. Two of those QBs were Andy Dalton and Jay Cutler.

Without looking, try to guess who's #1. I'm betting you'll guess right.

My guess is Kaepernick. QBR is a dumb metric that weights rushing ability to the positive far too much. (Not surprised it was Luck though, the guy can run far better than most think).