Pass block efficiency-week 1

[jterrell]

That shouldn't matter. This statistic appears to be context neutral, which is something that should be prefered in sports statistics. We shouldn't be judging O-linemen on what the QBs, RBs, and WRs do. We should judge the O-linemen on how they block.

that's the largest issue.

on some plays a QB already pre-determined he'll move a certain way.

a hurry on a qb who was planning to move by play design doesn't hurt the offense and isn't a big knock on the OL. a mental error by an OL that blocks no one but air and gives up a sack is horrendous play. trying to count those things as equal in value makes everything else pointless.

this is my issue with using hurries to defend spencer. he gets blocked by lesser guys by design because he the offense believes he isn't fast enough to actually ruin the play. so they focus block ware/ratliff/hatcher or whomever while allowing spencer a free "hurry" and they complete the pass for 10 yards and a first down.

 

[jterrell]

btw, i'd love to see a "forced error" metric. plays that cause a QB to either fumble, throw in INT or miss a WR by a large margin.

 

[Shinywalrus]

I don't see a difference. If there's a difference in magnitude, there's a difference in kind, which should be adjusted for -- otherwise they're not comparable stats.

I disagree.

Let's take a baseball example, comparing a double vs. a triple. When we calculate slugging percentage, we very comfortably accumulate the two in a single calculation by assigning the obvious number of bases associated with each hit as part of the calculation. The magnitude is different but the kind is the same.

A sack, hit and hurry are all of a same kind when it comes to individual pass protector evaluation - they are a defeat of a blocker in pass protection. The key measurement problem is that the distinctions in magnitude are not deterministic. What I am saying is that for the purposes of evaluating the effectiveness of that blocker, most of what determines the distinctions between a sack, hit and hurry have more to do with exogenous factors than the effectiveness of a blocker. If we did a principal components analysis on what drove differences between a hurry and a sack, I suspect that we would find it to be (1) the likelihood of an available throw for the QB in the post-defeat period and (2) whether or not the QB is aware of the pressure. While the former would be correlated with the length/effectiveness of the block, I think it's fair to abstract from this.

Put more succinctly, I agree with you that this statistic is not especially valuable for determining the quality of a defense, the likelihood of a play to have impacted the game, etc. But for evaluating individual pass protection, I think that it is a perfectly reasonable approximation.

 

[Shinywalrus]

btw, i'd love to see a "forced error" metric. plays that cause a QB to either fumble, throw in INT or miss a WR by a large margin.

Do you mean the statistic to be for measuring pass rushers? I feel like that's something that is unlikely to be related to the actions of a pass rusher himself, and more related to the quarterback or coverage in general. (except for the fumble forcing, that is)

 

[Ring Leader]

The issue could be weighting.

In my own mind, let's say a hurry is worth 1 point.

I think a sack should be worth 3 points and a hit should be worth 2 points.

Now run that formula again and see what you get.

To me the big problem is that a hurry is too vague. With the current 1-step rule a hit truly means you got in on a QB just after they released the ball. A sack means that you took away a passing play with a net average of say 7 yards and converted it on average to a 6 yard loss which is a loss of down and 13 yard change in average field position. There is also a shot at forcing a fumble and the QB takes a hit, of course.

You can argue a bit about the weightings but I think they favor hurries way too much in that metric. A hurry is not close to a sack or a hit, IMO.

Excellent post.

 

[AbeBeta]

For example, the Washington offensive line did not allow a sack, but there were two sacks due to other players.

Let me ask this -- is you start using a metric that doesn't credit these sacks to the OL, then what do you do about plays where, say a RB gets enough of a rusher that the OL let through to prevent a sack? What do you do if the OL lets someone through and the QB makes a nice play to avoid him and then completes a pass. In both cases, the OL failed but didn't give up a sack.

 

[MichaelWinicki]

Let me ask this -- is you start using a metric that doesn't credit these sacks to the OL, then what do you do about plays where, say a RB gets enough of a rusher that the OL let through to prevent a sack? What do you do if the OL lets someone through and the QB makes a nice play to avoid him and then completes a pass. In both cases, the OL failed but didn't give up a sack.

I think it would depend upon if the OL was "chipping" the pass rusher, "handing him over" to another player or it was his man to block, totally blew his assignment and then the back had to take over.

I'm thinking the offensive lineman would have to be credited with giving up the sack but in many instances for a defensive player to beat the offensive linemen, then beat the back and then sack the QB– Boy that would take some time... At which the QB could end up taking "credit" for the sack because of the inordinate amount of time he took without releasing the football.

 

[MichaelWinicki]

But for evaluating individual pass protection, I think that it is a perfectly reasonable approximation.

Exactly.

 

[StanleySpadowski]

That must be a misprint.

Batting average does not matter?

So all esle equal, at eam full of .230 hitters is as agood as a team full of .320 hitters?

Obviously hitting for power and situational hitting is telling...but

I'm sure I missed something.

I won't hijack a decent thread but it wasn't a misprint.

 

[TNCowboy]

There is something better.

Watching the games and using your own brain.

 

[MichaelWinicki]

Nothing wrong with using your brain.

But when you're talking 32 teams, 5 offensive linemen per team and better than a 1,000 passing plays per week AND THEN your job is to somehow sort the list by best to worst... The stock answer of "I sit in my Lazy-Boy eating cheese doodles and watching a few games a week." doesn't cut it.