Revisiting a very interesting analysis

[Matts4313]

The case of running vs passing when it comes to winning. What is the biggest indicator of who will win a game? This is a statistical argument by a former member here AdamJT13. He is a cap and draft comp pick guru. 

Quote

Let me break this down for you.

I've said all along that how well you run the ball -- your per-play efficiency, such as YPC -- doesn't really affect whether you win or lose the game. It doesn't really matter if you average 2.0 yards per carry or 4.0 yards per carry or 6.0 yards per carry. It barely increases your chance of winning by having a higher YPC. It doesn't really matter if you hold the opponent to 2.0 YPC or allow 6.0 YPC. What DOES matter, and what DOES affect your chance of winning to a much greater degree, is your per-play efficiency when you pass and your pass defense. If you pass more efficiently than your opponent does, you will almost always win the game, no matter how poorly you ran the ball or how poorly you stopped the run. If you don't pass better than your opponent, you will almost always lose, no matter how well you ran or stopped the run.

I noted that we went 20-1 in our 21 games with our LOWEST YPC from 1992-95. That is an indisputable fact. Averaging less than 3.0 yards per carry did not stop us from winning. And our opponents' ability to stuff our running game did not help them win.

In other words, exactly what I have been saying is exactly what happened. That's not surprising, though, because it's almost always what happens in the NFL -- every week, every season, for every team, for the past 30-plus years.

Quote

Under 200 yards at 6.2 to 7.0 YPA and 0 interceptions gives you a 94-8 record (.922 winning percentage) since 2005. Against winning teams, it's 38-4 (.905). Against teams that made the playoffs, it's 32-2 (.941).

Passing efficiency (offense and defense) wins games, not total passing yards.

Quote

Under 200 yards, under 30 attempts and no interceptions = 1032-253-3 since 2005 (.803 win pct). 

Quote

In our past 30 games, the team that has passed more effectively has won.

 

Quote

You almost always have to be more effective than your opponent in the passing game to win in the NFL. If your pass defense can make the opponent less effective than your pass offense, then you'll usually win, no matter how productive your running game is. But if your opponent passes more effectively than you, then even a great running game is almost never good enough to win.

 

Quote

Quote

In the early 1990s, we almost always won games when we passed the ball better than our opponent, regardless of whether we ran it better than our opponent. And when we didn't pass it better than the opponent, we usually lost, regardless of how well we (or Emmitt) ran it. 

In fact, from 1992-95, we had a higher winning percentage when we ran the ball WORSE than our opponent (21-5, .808) than when we ran the ball BETTER than our opponent (38-11, .776). 

The team that passed better in our games from 1992-95 went 62-13 (.827). The team that ran better went 43-32 (.573). 

And we went 20-1 in the 21 games with our LOWEST rushing averages during those seasons. 

Clearly, our games were almost always decided by whether we passed better than our opponent, not whether we ran it better.

 

 

Quote

The team that has passed the ball more efficiently has won 44 of our past 45 games. The team that has run the ball more efficiently is 24-21.

That's not league-wide. That's this team. YOUR team.

 

Quote

Apparently, you don't understand what "passing better" and "defending the pass better" means.

We averaged more adjusted net yards per pass play (the main stat behind the "theory"). We also averaged more yards per pass play and had a higher passer rating. All of those have a very high correlation to winning, and we won.

That "theory" is as true as it ever was. It's been true in 11 of our 12 games this season and in 48 of our past 50 games overall.

 

 @DaBoys Apparently he is still active on the other site. As you can see, I remembered incorrectly some of this.

[Matts4313]

Quote

Yes, opponents committed to stopping the run in the early 1990s (and yes, I watched and charted every game back then, just like I do now). Sometimes they did shut down our running game -- and they still LOST, usually because we passed better than they did. Remember -- how well you run or stop the run has almost no effect on whether you win the game. That's exactly what happened -- we ran poorly and won, and they stopped the run and lost. And what about games when the opponent didn't stop the run at all, or was only mediocre at stopping the run? Did that make our pass offense any better or worse, and did it change our chances of winning? Nope and nope. Whether we won or lost still almost always depended on whether we passed better than the opponent, and our passing efficiency didn't correlate at all with our rushing efficiency. 

As far as this "dramatic impact" that you assume happened on the passing game because we were good at running the ball, it doesn't happen nearly to the extent that most people think it does. It does on some plays, but not enough to have a significant effect on overall efficiency. That's why there is a low correlation between how well you rush and how well you pass. That's not as assumption or a theory, it's just a fact. One reason is because many pass plays -- and probably the most important pass plays -- are times when the opponent KNOWS you're going to pass. And THOSE are the plays that make a difference between winning and losing. Can you convert on third-and-long? Can you pass your way down the field in the two-minute offense? And JUST AS IMPORTANTLY, can your defense stop the opponent in those same situations? If you succeed in THOSE situations, when you have to pass or your defense has to defend the pass, that's often what makes the difference between winning and losing.

 

[DaBoys]

None of this means anything in regards to whether we should call a run or a pass. In fact it supports me in my argument that we should lesson Daks attempts and make him more efficient. 

[DaBoys]

No where does it say that the team who passed the ball more has a better record. 

"Under 200 yards, under 30 attempts and no interceptions = 1032-253-3 since 2005 (.803 win pct). "

 

you don't say?

[DaBoys]

And if you are so convinced that running the ball is not a weapon but merely a set up for the passing game THEN YOU SHOULD WANT MORE THAN 16 CARRIES TOO.

[Matts4313]

Quote

The stats are what they are because the facts are the facts. All I do are state the facts -- which have been proved over and over, every week, every season, for every team, for several decades. There are dozens and dozens of studies that have been done since at least the early 1980s (none of which have had anything to do with me), and they all reveal the same thing -- passing efficiency is extremely important, and rushing efficiency is almost insignificant. To argue otherwise is to ignore the facts

Quote

In technical jargon, rush efficiency explains only 4.4% of the variance in wins. You might as well guess randomly.

Pass efficiency explains 62% of the variance in wins in the NFL.
 

 

Quote

Regarding passing, running and winning, all of this has been thoroughly studied. It is not merely coincidental -- by and large, teams get a lead by passing more efficiently than their opponent, they maintain the lead by preventing the opponent from passing more efficiently than they do, and they win when they pass more efficiently than their opponent over the course of the game. On the other hand, running well and stopping the run does not correlate to winning on the scoreboard at any point in the game. For example, if there was cause-and-effect, the relationship between passing well and winning -- and the lack thereof between rushing well and winning -- should exist if one looked only at the halftime stats and score, before teams change their strategy and get either desperate or overly conservative, depending on the score. Those studies have been done, and the same results have been found -- winning the pass efficiency battle in the first half correlates highly with being ahead at halftime, while losing the rushing efficiency battle in the first half doesn't correlate with being ahead at halftime.

Quote

Romo averaged more passes in the first half in 2014 than he did in the first halves of games from 2011-2013. He averaged more passes in the first three quarters in 2014 than he did in the first three quarters of games from 2011-2013.

Quote

I have never once stated or even implied that all you need to do is pass. I've repeatedly stated that it doesn't matter how often you run or pass -- you can run more or pass more by preference. You can run more or pass more depending on the game situation. It's not about how much you run or pass. It's about how WELL you pass (and stop the pass).That's it. Not how often -- how well.

Let me say it again. It's NOT ABOUT frequency of running or passing. It's not even about the percentage of runs vs. percentage of passes. It has NOTHING to do with how often you run or how often you pass. 

 

[Matts4313]

4 minutes ago, DaBoys said:

None of this means anything in regards to whether we should call a run or a pass. In fact it supports me in my argument that we should lesson Daks attempts and make him more efficient. 

Follow me through this:

You said: Zeke hasn't hit 20 carrys all year, and we won one of those games. How do you suppose that happens? Do you think it's from calling run plays?

^^^ Attempts arguement

I said: Its been statistically proven for the last 15 years the correlation to winning is 100% directly tied an efficient passing game. Wins almost have no correlation to a the run games numbers. 

^^^^ Stating that rushing attempts has no correlation to winning. Passing efficiency does.

You said: We are 17 and 5 when Zeke has 20 carries or more. If Dak is throwing a bunch we lose.

^^^^^ Attempting to make it an attempt argument again. Ignoring what my argument was completely.

I said: The team with the more efficient QB almost always wins. The numbers are staggering.  The team with more efficient RB win about 50% of the time. IE - it has no statistical barring if you look at it from a macro view.

^^^^^ Again, attempting to explain to you that attempts dont matter. Passing efficiency does. 

You said:  2014 and 2016 were all about the run game. Fact. We are 17 - 5 when Elliott gets 20 carries. Fact. We are 4 - 7(over the last 2 seasons) when Dak throws over 30 attempts.

^^^^^ Once again ignoring my entire point

I said:  You mean in 2014 when Romo had one of the most statistically efficient seasons ever... Or 2016 when Dak had the GOAT rookie season and was one of the most statistically efficient QBs in the league? 

^^^^^ Once again attempting to point out the same point. Again. That point being that all teams across the entire history of the NFL win more often due to their PASSING EFFICIENCY REGARDLESS IF THEIR RUN GAME IS TRASH OR GODMODE. REGARDLESS OF ATTEMPTS.

:Skipping ahead:

You said: 17-5 when Zeke has over 20 carries. 10-9 when Dak throws for at least 30 attempts.

AdamJT said: (talking about our Cowboys) The team that has passed the ball more efficiently has won 44 of our past 45 games. The team that has run the ball more efficiently is 24-21.

----------------------------------------------------------------------------------------------------------

My point still stands. Also a very interesting tidbit to counter your Romo 2014 stat. That was one of his lowest throwing *ATTEMPTS* years in a healthy year. That was also one of his *HIGHEST ATTEMPTS* in the first 3 quarters. He was PASSING MORE OFTEN than all his gun slinger years in the first 3 quarters. He was just doing it efficiently. So when it came to the 4th Q, he didnt have to throw it another 15 times.

 

 

[Matts4313]

21 minutes ago, DaBoys said:

And if you are so convinced that running the ball is not a weapon but merely a set up for the passing game THEN YOU SHOULD WANT MORE THAN 16 CARRIES TOO.

You are still missing the point. Its not the quantity of carries that matters. It could be 10 carries or 45 carries. 

The passing offense efficiency and passing defense efficiency are the numbers we should be focused on. How do become a better passing team? Thats the ultimate question. 

As I just stated - even in 2014 - Romo was passing the ball MORE at the beginning of games. I know you contend we were all about the run, but literally Romo was throwing more often. 

[Matts4313]

31 minutes ago, DaBoys said:

None of this means anything in regards to whether we should call a run or a pass. In fact it supports me in my argument that we should lesson Daks attempts and make him more efficient. 

From AdamJT:

Quote

 There is very little correlation between:

Rushing well and passing well
Rushing well and winning
Stopping the opponent from rushing well and stopping the opponent from passing well
Stopping the opponent from rushing well and winning

That's why being able to run well or stop the opponent from running well is of little importance.

 

And, BTW, that was the reason I have been harping about Zeke. I know you went gaga over his rush stats. I was pssed off because he was crap in blitz pick ups and screwing up our pass game. If Zeke does his job in the pass game, we probably win. Or at least put ourselves in a much better position. 

[DaBoys]

What are Zekes numbers his first 10 carries of the game vs. his next 10? I've already looked. They are not as good. He gets stronger as the game goes on. 

 

We are simply not going to agree here. 16 carries is not enough. It's poor play calling relative to how the roster is built.

[resilient part 2]

1 hour ago, Matts4313 said:

edit

 

[resilient part 2]

don't know why I can't copy entire first post. but Matt4313

wow, what a post. I remember those years pretty well. I also remember how Dallas/Jimmy Johnson/Aikman would pass more in the first half to loosen up the defense and then pound Emmitt in the second half more. I don't know how to find the stats but Dallas had first half leads is so many of their games. 

great post overall. might be my vote for POY.

[Matts4313]

1 hour ago, DaBoys said:

What are Zekes numbers his first 10 carries of the game vs. his next 10? I've already looked. They are not as good. He gets stronger as the game goes on. 

 

We are simply not going to agree here. 16 carries is not enough. It's poor play calling relative to how the roster is built.

I just told you that you can run him 140 times a game, im not arguing against it, it still is not very likely to effect the outcome of the game. There is historically a 4.4% variance based on the run game. I really dont care if we run him 25 times a game. Great. Good. Yay. And btw, whether he is running at 2.1 YPC or 9.5 YPC, thats still not a mjor factor if we win. So "getting stronger" also doesnt equate to winning.

Im focused on the bigger picture. How do we get Dak more efficient? Because when our QB is more efficient than the other QB we have won 44 of the last 45 games. When Zeke dominates in the run game, we only have a 50% chance of winning. Which means we have a 50% chance of losing. Or did you not see Sunday? Zeke killing it on the ground and we lose because he sucked in the pass game. 

[DaBoys]

17 - 5 when Zeke has 20 plus carries.

10 -9(4-7 over the last two seasons) when Dak throws over 30 times 

 

You can keep typing there is no correlation until there is a blister on your thumb, but it doesn't change anything. Because I'm giving you the actual results. 

[DaBoys]

I want Zeke to get 20 carries. Period.

But overall I do agree that we have been losing because Dak sucks.