Mike McCarthy's Analytics Fraud

Toruk_Makto

Well-Known Member
Messages
14,231
Reaction score
17,331
Common sense tells you if you have a productive running game your play action should benefit from it as teams are keying on shutting it down.
Care to prove that with actual, you know statistics?

It seems obvious...but in fact it's not true.

That's kind of the point.
 

FuzzyLumpkins

The Boognish
Messages
35,685
Reaction score
27,237
Rushing attempts hardly matters. Rushing effectiveness doesn't matter.

Taking a small Dak sample size or a few outliers doesn't disprove that correlation. Often times it confirms it.

In any case McCarthy is saying something demonstrably untrue.

True for the NFL average but the "average NFL team" is not real but instead a statistical construct. There is no analog.
 

SHAMSzy

Well-Known Member
Messages
1,317
Reaction score
3,203
Call me stupid but IDC what analytics say

running the ball effectively opens up play action, I’ve watched & played enough football to know this
 

Toruk_Makto

Well-Known Member
Messages
14,231
Reaction score
17,331
That is not necessarily true. If you are a child or someone has no fear of being punched then they will not flinch.

In this case front 7 players are taught to react to run keys by taking gap assignments. If the DC is not afraid of the run then they will change the keys so they stay back and/or pass rush for longer. IE instead of when you see the OL start to run block you wait until after the exchange to verify it happens.

Now with us, Zeke being on the field likely makes it such that it doesn't matter because defenses sell out to stop him no matter what. Pollard is a different animal though.

PFF stats are for the whole league and as your article pointed out the conclusions are not universal but instead break down when you look at specific offenses, RB, and the like.
Another false statement.

Defense did not sell out to stop Zeke anymore than they did for our past running backs. I haven't looked at the 2019-2020 season. But this was true for Zeke's first 3 years in the league.

Sometimes actually counting something helps.
 

Toruk_Makto

Well-Known Member
Messages
14,231
Reaction score
17,331
True for the NFL average but the "average NFL team" is not real but instead a statistical construct. There is no analog.
This is the what's true for everyone isn't true for me fallacy that leads people to consistently make bad decisions. It's precisely why analytics are important.
 

Runwildboys

Confused about stuff
Messages
50,240
Reaction score
94,116
CowboysZone DIEHARD Fan
This is not true. Like at all.

If i've never punched you in the face....but I go through the motions and almost deck you....even if I pull up first you'll flinch.

Same concept applies.
After the 5th or 6th time, a normal person would stop flinching, unless you actually hit him in the face one of those times.
 

Brooksey

Well-Known Member
Messages
5,852
Reaction score
7,141
Look. I'm an engineer by training and now work at a big bank in investments. So yeah I like math and "analytics." I think they are instructive especially when careful study shows where what you believe to be true is in fact false. People who understand these counter-intuitive truisms claim a consistent edge.

There was much talk about how much Mike has learned in his year hiatus and how analytics was a big part of the thing he leaned on and has incorporated. And good lord he is not off to a great start. In fact it seems he is a fraud.

See below:



This is an awful quote. What he says here is the exact opposite finding from the analytics community.

In fact this is something I looked into when studying Zeke's pedestrian efficiency numbers last year (As a reminder we greatly overpaid Zeke after greatly over drafting him). But I digress....I posted this in August of last year...



What this shows is that play action passing is the most effective passes a QB can throw. And it hardly matters if you run the ball a lot. Or if you run the ball a little bit. And it does not matter if you run the ball well. Or if you run the ball poorly.

The point? If McCarthy is getting this basic thing wrong what confidence do we have that he actually learned anything in the last year, analytics or otherwise? And just like the Garrett era we can expect to give away the small edges that smart coaches understand and benefit from.

And that's a shame.


The only frauds are the idiots who believe running the ball is irrelevant and all you have to do is pass and stop the pass. Engineering and investment banking doesn't qualify you for anything you're talking about.
 

FuzzyLumpkins

The Boognish
Messages
35,685
Reaction score
27,237
Another false statement.

Defense did not sell out to stop Zeke anymore than they did for our past running backs. I haven't looked at the 2019-2020 season. But this was true for Zeke's first 3 years in the league.

Sometimes actually counting something helps.

So you counted the plays where the front 7 keyed run/pass on PA for Zeke?
 

FuzzyLumpkins

The Boognish
Messages
35,685
Reaction score
27,237
This is the what's true for everyone isn't true for me fallacy that leads people to consistently make bad decisions. It's precisely why analytics are important.

An average is not a universal truth that applies to all samples. Your article even points out how it is not universal. I even addressed that in my post how what you are describing is not universal.

It either applies to everyone or it doesn't. In your case it doesn't and its very clear there are at least some places where it doesn't work.
 

pansophy

Well-Known Member
Messages
4,002
Reaction score
4,112
The issue doesn't seem as simple as that. The same article states the following:

"Prescott also averaged 5.1 fantasy points per game (with a 133.3 passer rating) on play action passes when Elliott was on that field, with those numbers dropping to 1.9 fantasy points per game (and an 83.0 passer rating) during Elliott’s suspension weeks. Perhaps teams didn’t take the threat of play-action as seriously without Elliott on the field, and Prescott’s efficiency suffered as a result."

From the same article:
"On the other end of the spectrum we have four quarterbacks who actually posted a higher passer rating on non-play-action passes. This seemed weird to me, but perhaps the connection is all four passers’ teams also ranked bottom-10 in team rushing attempts over our sample."

Full article: https://www.pff.com/news/fantasy-football-metrics-that-matter-quarterbacks-on-play-action

So McCarthy is right about establishing the run, but it seems like running effectiveness is not as important as run attempts. Also, results vary significantly from team to team and from QB to QB. I am not sure we can call Big Mac a fraud just for this statement. We also can't expect him to spill all the beans on his research. Do you think Bellicheat would willingly tell the media stuff he has found using analytics??
I think you are on the right track at trying to get at the context of the numbers. That's the problem with all of these summaries statistics -- all the context is lost and it is the context that is essential to interpreting the possible causal relationships.

For example, it probably follows that teams that are in the bottom 10 in rushing attempts have a more pass oriented approach to offense to begin with. How much time do these teams spend on play-action passes to begin with? Do they have the personnel to even run it effectively given that they may draft with their passing philosophy in mind??

The most important thing McCarthy said in his interview with Peter King, and I why I came away feeling like he understood analytics well, is that he said the most important thing to figure out how you can actually apply the information. How do these analytics actually inform a specific decision, and most people make extraneous conclusions from data.

I would be curious to see how close teams are from having AI recommend 3 plays that are likely to be most effective given the identified opponent weaknesses, game situation, and opponent tendencies in those situations.
 

Redball Express

All Aboard!!!
Messages
16,253
Reaction score
12,758
Look. I'm an engineer by training and now work at a big bank in investments. So yeah I like math and "analytics." I think they are instructive especially when careful study shows where what you believe to be true is in fact false. People who understand these counter-intuitive truisms claim a consistent edge.

There was much talk about how much Mike has learned in his year hiatus and how analytics was a big part of the thing he leaned on and has incorporated. And good lord he is not off to a great start. In fact it seems he is a fraud.

See below:



This is an awful quote. What he says here is the exact opposite finding from the analytics community.

In fact this is something I looked into when studying Zeke's pedestrian efficiency numbers last year (As a reminder we greatly overpaid Zeke after greatly over drafting him). But I digress....I posted this in August of last year...



What this shows is that play action passing is the most effective passes a QB can throw. And it hardly matters if you run the ball a lot. Or if you run the ball a little bit. And it does not matter if you run the ball well. Or if you run the ball poorly.

The point? If McCarthy is getting this basic thing wrong what confidence do we have that he actually learned anything in the last year, analytics or otherwise? And just like the Garrett era we can expect to give away the small edges that smart coaches understand and benefit from.

And that's a shame.

Hmm..

I can see somebody questioning statements.

My experience with analytics is limited to being a psych major in college dealing in behavior modification.

What I learned was results must be repeatable to be valid and once validated..

changing a variable changes the ourcome.

So choosing to challenge MM and his use of them is impossible to judge.

We do not know his criteria or what he is looking for with his outcome data.

Back in the day for me..you wrote out a premise and did the analysis by declaring what was going to be changed in the analysis and reflected the deviations.

I do not want to go much further on this but..

this is football. Not lab rats.

I completely believe MM when he is doing this. And it's not just him. He is using a number of resources and like-minded others to analyze it all.

Did you see his setup at his personal home for this? I was blown away.

I immediately saw a coach with some elevated pursuits and looking for answers others did not know how to ask about.

MM is real. Just look at his resume.

This guy is no fake.
 

pansophy

Well-Known Member
Messages
4,002
Reaction score
4,112
This is not true. Like at all.

If i've never punched you in the face....but I go through the motions and almost deck you....even if I pull up first you'll flinch.

Same concept applies.
Fans and media focus too much on psychology and personality. If I say it's incredibly important to establish the run to do play action but I pass the ball 90% of the time, play-action isn't going to be that effective. Passing the ball at a "high" rate of 60% still means you are running it 40% of the time, and if you run effectively during those times that will keep drives moving. Match-ups are also incredibly important, because if I have an incredible defensive front 4 who can shut down the running lanes then play action is also going to be less effective.
 

Toruk_Makto

Well-Known Member
Messages
14,231
Reaction score
17,331
An average is not a universal truth that applies to all samples. Your article even points out how it is not universal. I even addressed that in my post how what you are describing is not universal.

It either applies to everyone or it doesn't. In your case it doesn't and its very clear there are at least some places where it doesn't work.
This isn't an average. It is an analysis of play action effectiveness across teams that run a lot or a little and for those who run well or poorly.

And across the board it is the play action itself and not your running that dictates the increased efficiency.

Facts are stubborn things.
 

gimmesix

Fat, drunk and stupid is no way to go through life
Messages
37,345
Reaction score
34,469
Listen I saw the whole analytics pitch as complete crap and salesmen ship. A coach is what he is, expecting one to change their stripes, well don’t hold your breath. Hopefully some culture will change here, but I knew that this analytics thing was complete ear candy to the masses.

I wouldn't say it's crap. If he embraced analytics as a way of providing more information, then it can be helpful. I would say that we shouldn't expect McCarthy to be much different than what he was as a head coach in Green Bay. He's not completely reinventing himself, and he shouldn't. He reached NFC Championship Games and won a Super Bowl as that coach.

Any person who wants to do better at their job, though, especially after being fired, is going to self-evaluate to see areas where they can improve and then set about trying to accomplish that. It doesn't mean that they will, but there's nothing wrong with trying.

For various reasons, Bill Belichick is a far better head coach with New England than he was with Cleveland. Some of that was taking lessons he learned from failing with the Browns and not making those same mistakes with the Patriots.

I think it's likely we get a better version of McCarthy as a coach than Green Bay had, but I acknowledge the possibility that we just get the same one. Even if McCarthy wants to change, it doesn't mean he will.
 

gimmesix

Fat, drunk and stupid is no way to go through life
Messages
37,345
Reaction score
34,469
What I found really interesting was in his interview with Eatman on the mothership, McCarthy stated that procedurally, whoever was calling plays had to also be the one installing the plays in practice. That he had not done that in GB and it caused issues. He was changing that too.

Right. That's part of the learning and changing process.

We have a football coach who has seen what it takes to reach the top of the game, but also what makes it all fall apart. Our best hope is that he has learned from those things and is able to apply that knowledge here.
 
Top