Recap: Troy Aikman Show, Ticket FM 9/21 - Coach McCarthy, Running the Ball, Analytics

[wick]

But no, no one is leading the league in rushing attempts if they're only getting 1 yard/carry.

This just proves the point that rushing does not cause winning. If it did, then we would maximize rushing attempts to cause maximum winning. And that would be easy enough because we'd simply call a rushing play every offensive snap. If you get 12 drives a game, we'd expect somewhere in the neighborhood of 36 rushes per game as a floor, which would have been the most attempts in the league in 2022.

It would cause the most wins, right?

 

[Rayman70]

Kellen Might get fired along with Staley in SD actually. Things arent heading in the right direction. Winless with KM there.

 

[pansophy]

The rushing attempts stat seems to be one that is skewed. Teams with a lead, especially trying to protect that lead and shorten the length of the game, will trend very heavily in favor of the run. They will almost sacrifice an attempt to even get a first down so that they can burn clock or force the other team to use their timeouts. Teams with a big lead will "call off the dogs" and run the ball. Teams trailing will trend heavily towards a near 100% pass ratio in the final 5 minutes.

It would be interesting to break the rushing stat down and look at halftime leads or even third quarter leads. See how many times the team leading at the half has more rushing attempts. That might be a better objective measure of the value instead of the "we're running the ball BECAUSE we are winning" vs. "we're running the ball TO win."

This is what most people don’t really get. They may say correlation is not causation but then say “hey we should have run more, the team with more rushes wins 80% of the time“.

I’d be curious to know what data analytics people have that makes them think momentum doesn’t exist. Every researcher in the world spends a lot of time trying to eliminate the placebo effect from their data. Often times the placebo effect has as much impact on results as the intervention they are testing. “Momentum” is another way to articulate the placebo effect because it’s the belief they have the upper hand.

For all they know “momentum” accounts for some of the failure rates on plays based on situation.

 

[vlad]

https://post.futurimedia.com/ktck/playlist/21/listen-5639.html?cb=1695302340.549203

Troy Aikman Recap:

• Cowboys are built in a method that they can beat you in many ways.
• When I watched the Jets game, I thought Mike McCarthy did a really good job calling the game. Knowing what the Jets had on defense, and playing an offensive style that gave them a chance. They were very successful against a strong, tough defensive team. To run the ball as many times as they did, it doesn't matter what the success rate was in yards/carry.
• I remember talking with Mike when he coached the Packers. It always stuck with me about him. A lot of coaches would tell me what they didn't average on carries. He always looked at it as the amount of rushes is what determined the success. And a game like the Jets, it's highly important to be physical and aggressive. (FYI, Seen the stats and Troy is right. Amount of rushes is higher determinant on winning.)
• I'm not an analytics expert. My understanding is the coach or analytics person on the team is not allowed a computer. They have a binder that maps it out. They communicate it down based on down and situation. I think most coaches, including Mike, look at the analytics but it's not a hard and fast rule. There's a lot that goes into it. Most coaches I talk to are cautious about it. These are still human beings playing the gamea nd have to be mindful. Analytics say there is no so thing as momentum but most coaches and myself disagree with that. Most coaches wanted to go for it on 4th down, but they were afraid they would lose their job. Now it's part of the game. Maybe it's the reverse now though because the analytics say they should have gone and didn't.

Here are the analytics and topic Aikman is talking about. Stats with a stronger correlation to winning.

Stat	Record	Win Pct
More Rush Atts	3954-969 (141 tied)	0.803
Fewer Turnovers	3174-849 (1041 tied)	0.789
Fewer Pass INTs	2806-808 (1450 tied)	0.776
More Rush TDs	2626-805 (1633 tied)	0.765
More Pass TDs	2645-988 (1431 tied)	0.728
Fewer Sacks Taken	3001-1198 (865 tied)	0.715
More Rush Yds	3581-1458 (25 tied)	0.711
Fewer Sack Yds Lost	3316-1522 (226 tied)	0.685
Fewer Fumbles Lost	2055-985 (2024 tied)	0.676
More Total Yds	3404-1642 (18 tied)	0.675
More First Downs	3150-1664 (250 tied)	0.654
Fewer Fumbles	2405-1415 (1244 tied)	0.630
Fewer Penalties	2531-1986 (547 tied)	0.560
More Net Pass Yds	2815-2233 (16 tied)	0.558
Fewer Penalty Yds	2718-2242 (104 tied)	0.548
More Pass Yds	2643-2403 (18 tied)	0.524
More Pass Cmp	2092-2695 (277 tied)	0.437
More Pass Att	1545-3365 (154 tied)	0.315

This Cowboys Zone Thread covers most of these and where our team currently ranks.

https://cowboyszone.com/threads/cowboys-team-stat-rankings-parcells-would-be-proud.514112/

Great post thx

 

[vlad]

It’s funny I’m some current thread there is one guy hammering another guy like he’s an idiot but this data supports the nail not the hammer.

Humans are certainly full of themselves lol. Good for the nail

 

[RonnieT24]

If you're arguing over the validity of the stats you have missed the point entirely. Running the ball is about attitude and about imposing your will on the other team. If you succeed in doing so you usually win the game. Yes teams that are winning will run more to secure the win.. but in most cases they ran it to get the lead in the first place.