Here is why I don't put much stock in pressure stats

[DallasInDC]

PFF and ESPN obviously define a pressure differently.

Based on my recollection, PFF basically means the defender made the QB move. That’s a ridiculous broad definition and doesn’t even look at play outcomes. Perfect example is from about 12 years ago when Anthony Spencer got a pressure on the game-ending over TD catch and run by Stevens-Howling against Arizona.

This is my biggest pet peeve on pressure stats...if you're going to track it, only count those that result in less than 0 yds gained on the play (the same as you do with sacks and TFLs/STF).

 

[Spottswoode]

This is my biggest pet peeve on pressure stats...if you're going to track it, only count those that result in less than 0 yds gained on the play (the same as you do with sacks and TFLs/STF).

So you don’t find making the QB throw the ball away or throwing before they are ready a useful metric?

I would think pressure to make a QB uncomfortable in the pocket would be a good thing. We witnessed that last night. I would also guess that a decent percentage of interceptions are caused by hurries.

 

[DallasInDC]

So you don’t find making the QB throw the ball away or throwing before they are ready a useful metric?

I would think pressure to make a QB uncomfortable in the pocket would be a good thing. We witnessed that last night. I would also guess that a decent percentage of interceptions are caused by hurries.

I probably should have said less than or equal to zero yards...but that was a lot more to type. Also the point was the result should be a negative play...includes INTs...I dont think you should be awarded for something that results in a positive play.

 

[Spottswoode]

I probably should have said less than or equal to zero yards...but that was a lot more to type. Also the point was the result should be a negative play...includes INTs...I dont think you should be awarded for something that results in a positive play.

I understand your point but they already track negative plays with other metrics.

Ultimately it’s just 1 of hundreds of metrics captured and likely dumbed down for public consumption, while teams drill down to situations, personnel, etc., etc. Some NFL teams hire the Wall Street equivalent of a quant to analyze data.

I think everyone can agree that a QB under duress is a positive thing, regardless of the outcome of individual plays. IMO, all pressure, when viewed alongside other metrics, can be a useful measure of that goal. But what we look at is a fraction of what NFL teams consider.

 

[thunderpimp91]

I probably should have said less than or equal to zero yards...but that was a lot more to type. Also the point was the result should be a negative play...includes INTs...I dont think you should be awarded for something that results in a positive play.

We already have those types of stats though that tell us what was accomplished on a particular play....sacks, ints, yards, TDs, catches, FFs, even tackles. The issue is that these are largely team stats. For example QB drops back and throws the ball into the end zone, that QB is dependent upon 1) the OL to hold protection and 2) for the WR to catch the ball. A DE that gets pressure is still dependent upon the secondary to force that QB to hold onto the football an extra second as well as the rest of the DL to contain the QB in the pocket to give the DE time to finish off the sack.

A huge part of the advanced analytics is an effort to isolate individual player contributions to a plays regardless of the outcome. That DE that gets quick pressure, but has a garbage secondary will still get recognized by the advanced stats where they wont be credited with nearly as many sacks as they could have if they were on a stronger team. That WR who was wide open for two TDs isnt penalized because of poor throws by the QB.

Neither side is going to be a perfect story teller here, but when you take into account both the traditional counting stats and the advanced stats you will typically get a more accurate picture of players value.

Really only tracking stats that result in a direct positive or negative plays is like says that 40 yard run doesnt mean anything because we still ended up punting on the drive. You missed the ultimate goal of scoring points, but that run still flipped field position, converted a new set of downs, and helped keep the other teams offense off the field for at least 3 more plays. Same with a quick pressure that still results in a completed pass....yeah you didn't accomplish the goal of creating a negative or neutral play there, but you're still helping speed up the QBs mental clock, making the opposing play caller wonder if they need to keep a TE in to chip, etc. In that regard you're still impacting the game without impacting the stat sheet.

 

[J12B]

Pressuring the QB means something. Pressure stats don't. They are too subjective. You and I can watch the same game and come up with completely different pressure statistics.

In order for a "pressure" to be recorded as a stat, a player needs to be within 2 feet of the quarterback before the ball is thrown.

It's completely objective.

 

[gtb1943]

What is pressure to Danny Dimes is not pressure to Mahomes

 

[tm1119]

Yeah the advanced stats in football are pretty awful. This clip below was making the rounds earlier in the year that really proves it

 

[MarionBarberThe4th]

I feel like broadcasts should do a better job of showcasing los play

They could utilize a Birds Eye view cam and let the replays speak for themselves and perhaps get new graphics to help the viewer see where the movement is and what holes are developing.


So we wouldn’t have to dork out on a stat but by halftime the viewer would have a clear understanding of the top offensive and defensive linemen bc it would’ve been presented and discussed



I think the way they do it now is they only focus on edge rushers and they just kind of go “oh boy look out !”

 

[Risen Star]

Regardless of which you prefer, they both use their own defined criteria for pressures and all players are judged on that criteria.

Comparing a players ESPN pressure to another players PFF is useless. But if you compare 2 players using the same criteria (ESPN or PFF), you get a fair comparison, regardless of whether you agree with the pressure definition or not.

Does one person do all the analysis? Because even if we're supposed to be using the same criteria, whether they met that criteria is subjective. It's an opinion and if you have multiple people involved you have multiple opinions that render the exercise useless.

 

[Spottswoode]

Does one person do all the analysis? Because even if we're supposed to be using the same criteria, whether they met that criteria is subjective. It's an opinion and if you have multiple people involved you have multiple opinions that render the exercise useless.

I have no idea how many people do the analysis. And I agree there is some subjectiveness. There is likely a degree of subjectiveness in most football stats.

But it doesn’t render the exercise useless. It just needs to be understood that there is a margin for error. It doesn’t need to be a double blind, randomized, placebo controlled study with p-values assigned to provide value.

There is a criteria put in place for each group tracking pressure (ESPN, PFF, etc.) that I assume is understood by everyone involved. But there will likely be plays interpreted differently by certain individuals. Most of the time that bias will likely apply to all the questionable plays that person evaluates. And if they were smart, they would have an evaluation process to periodically review results.

There is no doubt that pressure matters and having a metric to track would provide value when developing a bigger picture. The goal isn’t perfection. And the stat isn’t intended to tell a story in isolation.

The depth of football analytics has exploded because people understand that individual metrics in isolation can be poor measures of outcomes and that more in-depth metrics can create a clearer, more accurate picture. But if the criteria for inclusion is zero subjectivity for all metrics, we can probably scrap 95% of football stats, along with referees, instant replay, and a lot of other things.

 

[Coogiguy03]

Pressuring the QB means something. Pressure stats don't. They are too subjective. You and I can watch the same game and come up with completely different pressure statistics.

Pressure to some is he got passed his blocker, but the QB still made play for positive yards!!!


Pressure to others means he actually affected the play of the QB to make him make a bad read, bad throw which caused no gain in yards!!

 

[Pass2Run]

Regardless of which you prefer, they both use their own defined criteria for pressures and all players are judged on that criteria.

Comparing a players ESPN pressure to another players PFF is useless. But if you compare 2 players using the same criteria (ESPN or PFF), you get a fair comparison, regardless of whether you agree with the pressure definition or not.

It's really not a fair comparison because you have different individuals doing the evaluations and they all have different biases.

 

[Spottswoode]

I gave my opinion on post #31.

 

[Hardline]

This is why I laugh at people that say pressuring the QB is as good as sacks.

 

[Spottswoode]

This is why I laugh at people that say pressuring the QB is as good as sacks.

Who says that?

That’s like saying missing a putt by a foot is the same as making the putt.

 

[kskboys]

Pressuring the QB means something. Pressure stats don't. They are too subjective. You and I can watch the same game and come up with completely different pressure statistics.

Exactly what I was going to say.

 

[kskboys]

This is why I laugh at people that say pressuring the QB is as good as sacks.

Well, it's silly because they cannot be separated. A sack is a part of pressure.

Constant pressure is necessary to defeat any good QB.

 

[kskboys]

Who says that?

That’s like saying missing a putt by a foot is the same as making the putt.

I don't think anyone is saying. Well, maybe the homers trying to support their player love.

 

[nyc-cowboy]

PFF and ESPN obviously define a pressure differently.

Based on my recollection, PFF basically means the defender made the QB move. That’s a ridiculous broad definition and doesn’t even look at play outcomes. Perfect example is from about 12 years ago when Anthony Spencer got a pressure on the game-ending over TD catch and run by Stevens-Howling against Arizona.

You mean “almost Anthony”?