Normalizing the QB Rating statistic by accounting for sacks (1 Viewer)

[Saint Kamara]

So all that know the formula for the standard QB rating knows that it does not take into account anything to do with how often a QB takes sacks.

Obviously taking a sack is fairly detrimental to sustaining a drive and should be reflected in your QB rating.

So I've recalculated the QB Rating of the top 6 QBs of all time that penalizes a sack as both an incomplete pass and the negative yardage being subtracted from passing yards.

Here are the results (standard formula QB rating in parenthesis):

1. Rodgers........ 93.6 (102.4)
2. Brees............. 93.5 (98.4)
3. Manning........ 92.7 (96.5)
4. Brady............. 91.2 (97.0)
5. Wilson........... 90.6 (101.2)
6. Romo............ 90.5 (97.1)

Those would be the only QBs above the 90 mark.

With Rodgers and Brees going in quite opposite directions the last 2-3 years, one would expect Brees to take over #1 over Rodgers after next season.

 

[Pocket Hercules]

Really interesting. Do you have the ability to calculate with a sack counted as a completed pass for negative yardage?

 

[guidomerkinsrules]

Really interesting. Do you have the ability to calculate with a sack counted as a completed pass for negative yardage?

What’s the thinking behind this?

 

[chameleonknight]

What’s the thinking behind this?

because it’s physically impossible to have both an incompletion and a loss of yardage (absent a penalty) so instead of “punishing” the rating twice, counting it as a “completion” would skew the numbers a different way

that being said I like the OPs original thought process

 

[guidomerkinsrules]

because it’s physically impossible to have both an incompletion and a loss of yardage (absent a penalty) so instead of “punishing” the rating twice, counting it as a “completion” would skew the numbers a different way

that being said I like the OPs original thought process

It would be interesting to see if sacks are bigger druve killers than penalties
If they’re in between say illegal procedure and holding then, no, there shouldn’t be a dbl penalty
- if they’re worse than holding then counting loss if yardage AND incompletion makes some sense

 

[Saint Kamara]

because it’s physically impossible to have both an incompletion and a loss of yardage (absent a penalty) so instead of “punishing” the rating twice, counting it as a “completion” would skew the numbers a different way

that being said I like the OPs original thought process

An incompletion is a loss down.

A sack is a loss down.

So this is an incomplete pass with loss yardage on the same play.

I ran the numbers several months ago with only counting for negative passing yards, thus affecting only the yards per attempt aspect of the formula, but I felt it didn't really fully account for the negativity of this play. So then I came up with the idea of calling it an incomplete pass as well.

A sack is usually a drive killer. It needs to be accounted for appropriately.

 

[Pocket Hercules]

What’s the thinking behind this?

Because an incomplete pass never results in negative yardage. However, a completed pass in which the tackle is made behind the LOS is the same exact outcome and has the same exact effect as a sack.

Having said that, there is no skill or achievement involved in taking a sack, whereas there is always some skill involved in completing a pass even if it results in a TFL, which would bump the QB rating a hair. So the actual rating should probably lie somewhere between Saint Kamara's example (which attributes two negative and mutually exclusive outcomes to the same play). Simply taking the sack yardage away would not be enough of a penalty because it wouldn't really account for the loss of down that sacks create.

 

[onanygivensunday]

Obviously taking a sack is fairly detrimental to sustaining a drive and should be reflected in your QB rating.

The shortcoming with your approach is the QB's play isn't the only factor that results in a sack.

The sack could be a coverage sack. Why should the QB's passer rating be reduced by a coverage sack? It doesn't make sense to me. Nobody's open so he eats the ball... good decision in this case.

And what about the sack that occurs so quickly that the QB doesn't stand a chance to get rid of the ball... like an o-lineman whiffing on his block... or by a full-house blitz. Why should the QB's passer rating be reduced by these types of sacks? Again, that doesn't make sense to me.

I understand your approach but certain sacks shouldn't be counted if the effect is to degrade a QB's passer rating.

 

[Stalwart385]

I'm not a huge fan of accounting for lost yardage due to sacks. It's a very random stat imo. Playcall can have more to do with that than any measure of skill. I get the idea of making it worse than an incomplete but just too random imo. I do like counting sacks as incompletes though.

I've seen studies that say the vast majority of sacks are on the QB and not the o-line anyway. Either he didn't get rid of the ball when he should have (at least throw it away) or he didn't shift the pocket to a clear opening. Not always of course but the majority of the time. That's why Brees is so good at avoiding sacks. He always finds an outlet, usually gets rid of it at least and has great pocket awareness.

QB who take lots of sacks are basically being credited in QB rating for a drive-killing play when it could have just been a less costly incomplete but it looks worse on their rating.

 

[Saint Kamara]

Because an incomplete pass never results in negative yardage. However, a completed pass in which the tackle is made behind the LOS is the same exact outcome and has the same exact effect as a sack.

Having said that, there is no skill or achievement involved in taking a sack, whereas there is always some skill involved in completing a pass even if it results in a TFL, which would bump the QB rating a hair. So the actual rating should probably lie somewhere between Saint Kamara's example (which attributes two negative and mutually exclusive outcomes to the same play). Simply taking the sack yardage away would not be enough of a penalty because it wouldn't really account for the loss of down that sacks create.

And that's why I think calling it a incomplete pass perfectly penalizes the QB on his passer rating. A sack is worse than an incomplete pass. That's the simplest way I can explain why I think this should be the way we proceed in accounting for a sack.

 

[Saint Kamara]

The shortcoming with your approach is the QB's play isn't the only factor that results in a sack.

The sack could be a coverage sack. Why should the QB's passer rating be reduced by a coverage sack? It doesn't make sense to me. Nobody's open so he eats the ball... good decision in this case.

And what about the sack that occurs so quickly that the QB doesn't stand a chance to get rid of the ball... like an o-lineman whiffing on his block... or by a full-house blitz. Why should the QB's passer rating be reduced by these types of sacks? Again, that doesn't make sense to me.

I understand your approach but certain sacks shouldn't be counted if the effect is to degrade a QB's passer rating.

Well we can go down that rabbit hole and take away every interception that was due to the intended receiver having the ball go through his hands only to be caught by a defender, or count an interception on a pass that should have been intercepted that was just dropped by the defender.