Rushing statistics and their correlation to sucess

[waldoputty]

If that's all you got from my post then you missed some of it and certainly the spirit of it. I would hope that everyone can debate this and most topics without mocking or derision.

Again, I don't think anyone including Adam is saying what some of you have locked onto and are quoting erroneously. I certainly hope that is not by design but by misunderstanding.

You'll never get an argument from me that running is unimportant. Of course running affects winning just not as much as passing or even turnovers. Teams run by design to force unfavorable match ups and schemes. You just don't have a current working formula that correlates running with winning.

Sorry, but I simply do not recall your post at this point.

I have gone out of my way to post sections from statistics websites to help move the argument along by providing the proof that some said they seek. Furthermore, I even posted definitions and 'translations' to help those without the background to read those reference themselves. Though those who disagree appear unwilling to read those references despite my best efforts.

You said "I don't think anyone including Adam is saying what some of you have locked onto and are quoting erroneously... You'll never get an argument from me that running is unimportant. Of course running affects winning just not as much as passing or even turnovers."

That statement about "running affects winning just not as much as passing ..." seems fair to me.
Though your characterization of Adam is incorrect.
This is copied directly from Adam in post # 208: "Nobody has ever argued that running is not important or useful. What is not very important is how WELL you run it overall, or how well you stop the run. There are many advantages to running the ball, even if you run it poorly."

He really means that running well is not important to winning, and that stopping the other team's run is not important.
Why, I do not know.

 

[Idgit]

At least Adam owns his statement......he says running the ball efficiently has no correlation to winning

You always try to come behind him and massage his words into something else...'that it is important but still has no effect on winning'.......which makes no sense

Take a stance and own it.....don't mealy-mouth some middle of the road nonsense

Adam can speak for himself.

What *I've* written in response to your questions already is perfectly clear. I can't help it if you didn't understand it and keep asking the same thing over and over.

 

[waldoputty]

You don't always want to limit possessions and any outliers will likely be turnovers which generally trumps most else.

This entire conversation is primarily statistical and does not take into account the best input/output, ie the human brain that is informed and able to react quickly to changing dynamics which you could call variables. There is no all encompassing equation that defines football just as there is none for a theory of everything.

And if you dismiss input as flawed you cannot use circular logic to address that issue. One wants affirmation they are flawed besides I told you they were. One can have an opinion and state it, civily, you disagree with this or that but don't confuse stats with fact.

I went back and found your original post.

Regarding civility, it should go both ways.

Regarding why the input was flawed, my post #1 actually discussed this:

The primary issue is the lack of correlation between rushing efficiency and points scored:
1. The problem here is that rushing efficiency is a flawed stat. If the defense plays 8 in a box, they take away from rushing efficiency. However, while efficiency is poor, you may score more because the passing game is opened up (increasing scoring potentially). Thereby, any 'correlation' is reduced.
2) The ability of an offense to score when the defense is play 8 in the box is dependent on the OL, QB and WRs. Can they handle the pass rush? Can the passing game execute under a run blitz with less reaction time? If they can handle these well, then they will score more points. If not, the offense will score less points. Again, these wash out 'correlation'. The effect of 8 in the box - reducing rushing efficiency while points scored may go up or down depending on the rest of the team's ability to cope.
3) Beyond how the defense plays, rushing efficiency is affected obviously by the offense play calling as well as the score. If you are ahead late in the game (e.g. 4th quarter), then the offense will rush to take time off the clock. During those rushes, the defense is expecting the run, thereby lowering the rushing efficiency. Again this takes away any 'correlation'.

 

[Idgit]

How many of the stockpiled choices matter?

That is the question at the end of the day.

I think this was snark, but I can't tell what you're snarking, so I'll take it seriously. Here's what they've added in addition to what they had already:

2017
Philadelphia's first-round pick
Tennessee's second-round pick
Indianapolis' seventh-round pick

2018
Philadelphia's second-round pick

They conditionally surrendered their 2017 fourth-round pick to the Eagles and 2017 sixth- or seventh-round pick to the 49ers. This doesn't include the pick they just gave the Pats.

I think that was smarter than taking Wentz and trying to build around him after, assuming they didn't destroy him in the process.

 

[Nightman]

Adam can speak for himself.

What *I've* written in response to your questions already is perfectly clear. I can't help it if you didn't understand it and keep asking the same thing over and over.

Just plead the Fifth next time and I will grant you immunity......let you save face

 

[Toruk_Makto]

How many of the stockpiled choices matter?

That is the question at the end of the day.

Have you even followed what the Browns are doing over the last year or did you just want to bash analytics even at the risk of looking foolish?

 

[T-RO]

I think this was snark, but I can't tell what you're snarking, so I'll take it seriously. Here's what they've added in addition to what they had already:

2017
Philadelphia's first-round pick
Tennessee's second-round pick
Indianapolis' seventh-round pick

2018
Philadelphia's second-round pick

They conditionally surrendered their 2017 fourth-round pick to the Eagles and 2017 sixth- or seventh-round pick to the 49ers. This doesn't include the pick they just gave the Pats.

I think that was smarter than taking Wentz and trying to build around him after, assuming they didn't destroy him in the process.

Yup. They will have the #1 pick overall, and four picks in the top 48. And they understand the draft. They didn't see a quarterback they had a high grade on...so despite needing one...they deferred to 2017. They'll have their pick of the litter.

 

[jobberone]

Sorry, but I simply do not recall your post at this point.

I have gone out of my way to post sections from statistics websites to help move the argument along by providing the proof that some said they seek. Furthermore, I even posted definitions and 'translations' to help those without the background to read those reference themselves. Though those who disagree appear unwilling to read those references despite my best efforts.

You said "I don't think anyone including Adam is saying what some of you have locked onto and are quoting erroneously... You'll never get an argument from me that running is unimportant. Of course running affects winning just not as much as passing or even turnovers."

That statement about "running affects winning just not as much as passing ..." seems fair to me.
Though your characterization of Adam is incorrect.
This is copied directly from Adam in post # 208: "Nobody has ever argued that running is not important or useful. What is not very important is how WELL you run it overall, or how well you stop the run. There are many advantages to running the ball, even if you run it poorly."

He really means that running well is not important to winning, and that stopping the other team's run is not important.
Why, I do not know.

You'll both have to define what you mean if you want to argue the point more. My guess is Adam is thru with the debate.

For me it means at least well enough to create an honest defense.

If the vast majority of posts are too personal then I'd say the thread is reaching its maximum benefit to the site and members.

 

[Idgit]

Just plead the Fifth next time and I will grant you immunity......let you save face

"But this one goes up to 11!"

Really, dude, if you can't grasp the simple distinction, I don't know what to tell you. It's not anything that hasn't been said a hundred times already in dozens of other threads and understood by posters on both sides of the debate each time it comes up.

 

[jobberone]

I went back and found your original post.

Regarding civility, it should go both ways.

Regarding why the input was flawed, my post #1 actually discussed this:

The primary issue is the lack of correlation between rushing efficiency and points scored:
1. The problem here is that rushing efficiency is a flawed stat. If the defense plays 8 in a box, they take away from rushing efficiency. However, while efficiency is poor, you may score more because the passing game is opened up (increasing scoring potentially). Thereby, any 'correlation' is reduced.
2) The ability of an offense to score when the defense is play 8 in the box is dependent on the OL, QB and WRs. Can they handle the pass rush? Can the passing game execute under a run blitz with less reaction time? If they can handle these well, then they will score more points. If not, the offense will score less points. Again, these wash out 'correlation'. The effect of 8 in the box - reducing rushing efficiency while points scored may go up or down depending on the rest of the team's ability to cope.
3) Beyond how the defense plays, rushing efficiency is affected obviously by the offense play calling as well as the score. If you are ahead late in the game (e.g. 4th quarter), then the offense will rush to take time off the clock. During those rushes, the defense is expecting the run, thereby lowering the rushing efficiency. Again this takes away any 'correlation'.

One can only worry about one's posts in regards to civility.

And I'm aware of your initial post and I understood it then.

 

[waldoputty]

You'll both have to define what you mean if you want to argue the point more. My guess is Adam is thru with the debate.

For me it means at least well enough to create an honest defense.

If the vast majority of posts are too personal then I'd say the thread is reaching its maximum benefit to the site and members.

I actually think both Adam and I have defined our terms.
So it is not clear to me what you mean.

It would be nice if the posts are less personal.
When I go through the trouble to find a reference to help others understand and get their 'proof', it would be nice if the response is 'thank you' instead of throwing it back in my face saying it is just bunch of greek...

 

[CCBoy]

Here is another article on errors in statistical analysis - it is from Wikipedia
I will provide definitions again for convenience and even a so-so attempt in translation in english.

Definitions:
variable - the data. in our case they are 'rush efficiency' and either won-loss or points scored
dependent variables - the end result - in our case, either won-loss or points scored
independent variables - the cause - in our case, rush efficiency
error in variables - in our case, the way we are defining external effects such as defense formation (e.g. 8-in-the-box), offense playcalling, etc.
regression - the statistical technique that is getting you the correlation that Adams uses
non-linear models -fancy modeling that probably would be most relevant for this type of complex scenario

what the 2 paragraphs are saying:
1. ordinary regression cannot account for errors in the dependent variables (regressors). this means ordinary regression analysis cannot handle complications like external factors such as defense formation and offense playcalling when it is simply trying to look for correlation between rush efficiency and win-loss/points scored.
2. fancy modeling would be needed, but the situation is complex and unpredictable when there are external factors.

translation:
ordinary statistical analysis that Adam is using is incapable of handling the rush efficiency vs. win-loss/points scored analysis. you need to do a complex analysis and it is very hard and predictable.

https://en.wikipedia.org/wiki/Errors-in-variables_models
Errors-in-variables models
From Wikipedia, the free encyclopedia

In statistics, errors-in-variables models or measurement error models[1][2] are regression models that account for measurement errors in the independent variables. In contrast, standard regression models assume that those regressors have been measured exactly, or observed without error; as such, those models account only for errors in the dependent variables, or responses.

In the case when some regressors have been measured with errors, estimation based on the standard assumption leads to inconsistent estimates, meaning that the parameter estimates do not tend to the true values even in very large samples. For simple linear regression the effect is an underestimate of the coefficient, known as the attenuation bias. In non-linear models the direction of the bias is likely to be more complicated.[3][4]

Taking a statistical relevant view of progressions based upon valid start points in analysis....

But then always turn to a crap load start point based in arbitrary points chosen to further an individual's agenda for support of his own invalid arguments...

poor standard applied to poor standard, still doesn't improve...even if one takes some Brasso to try to polish it up.

Listen to the talk of football...and quite claiming glory when two more victories, against Cleveland and then Pittsburgh, and the Cowboys WILL be in the playoffs. Now argue if one game won with Dak Prescott and one won with Tony Romo makes a difference with that run perspective as well.

 

[waldoputty]

One can only worry about one's posts in regards to civility.

And I'm aware of your initial post and I understood it then.

Since you understood my initial post, you understand that this is a philosophy of statistical analysis discussion, not a discussion with dueling regression correlations.

Since a strict statistical proof is not viable, I pasted sections from statistical websites regarding the common errors in statistical analysis. I went through the trouble to define the statistical vocabulary and translated the 'technical writing' to enable others without the background to understand. It seems the individuals interested in the proof would not accept or read these references, but go back to the circular argument demanding for statistical proof.

I am not the only one who tried to explain the errors in Adam's analysis. Another poster, probably much more knowledgable than me, also tried to explain the problem in the analysis. It is unfortunate that these individuals also apparently ignored his explanation

 

[jobberone]

This thread is no longer productive and is being closed.