On Team Building for winning

[skibrett15]

 

Some highlights:

Quote

 

"Loosely speaking, the very best running game of the last decade is worth as many wins as the sixth best starting QB. 11/20"

"Passing is more important than everything else, turnovers are more important than rushing, while penalties are the least important. And offense always matters more than defense. By the way: Rushing has a positive effect on winning."

"Being one standard deviation better than the mean in terms of Passing Yards/Attempt will give you approximately 1.3 wins more. If your team is completely average but has a Passing Game that is one standard deviation better than average, it is expected to have 9.3 wins. In order to gain these approximately 1.3 wins, a team has to have a Passing Game that ranks in the 16th (highest) percentile."

 

 

So, when building your team.  Pass the ball efficiently, defend the pass, don't turn the ball over.  In that order.  Then you can worry about running it.

[Norm]

And cut your kickers

[squire12]

37 minutes ago, skibrett15 said:

 

Some highlights:

 

So, when building your team.  Pass the ball efficiently, defend the pass, don't turn the ball over.  In that order.  Then you can worry about running it.

Is this predictive?   Wondering if this is causation or more correlation of variables after the fact?

[AlexGreen#20]

5 minutes ago, squire12 said:

Is this predictive?   Wondering if this is causation or more correlation of variables after the fact?

It's not a direct measurement, but there's also an absolutely massive sample size to pull from. I might run it through Minitab when I get home tonight to see if I can get the predictive coefficients and run a seasons worth of data.

[skibrett15]

13 minutes ago, squire12 said:

Is this predictive?   Wondering if this is causation or more correlation of variables after the fact?

no, not predictive.

 

A better thread title (not mine btw, just pulled from twitter) might be "what happens to NFL teams when they win" rather than what makes them win.

[HyponGrey]

It's a passing league. It helps a lot to have a passable run game, though. Don't want to blow anybody's shoulder.

[skibrett15]

27 minutes ago, HyponGrey said:

It's a passing league. It helps a lot to have a passable run game, though. Don't want to blow anybody's shoulder.

actually having a passable run game doesn't really matter.  Having a poor run game hurts a little, and having a very efficient run game helps a little.  That's kind of the point of the analysis.

[Ragnar Danneskjold]

I don't see it, but does he have anything in the analysis looking at interaction of the pass game and the run game?  Seems like the analysis is for how each variable affects win total assuming independence from the other variables?

[squire12]

@ny92mike

This might be something you would be interested in for you sim league.

[boratt]

Passing more important than running, but running still important. Yeah. I think we could have eyeballed that one.

[skibrett15]

4 minutes ago, Ragnar Danneskjold said:

I don't see it, but does he have anything in the analysis looking at interaction of the pass game and the run game?  Seems like the analysis is for how each variable affects win total assuming independence from the other variables?

O_R_YDS : O_P_YDS correlate positively but slightly, looks to be like 0.2-0.3 or so.

[skibrett15]

Just now, boratt said:

Passing more important than running, but running still important. Yeah. I think we could have eyeballed that one.

• Passing more important than running
• Passing D more important than running O
• Turnovers surrendered more important than running O
• Turnovers created more important than running O
• Offensive penalties more important than Running D
• Defensive penalties least important of everything
• Better passing efficiency leads to more points, and almost the same in point differential.  Scoring more points by "throwing the ball all around" doesn't create a shootout, it actually just means you win by more.  Keeping the other team's offense off the field is BS.
• Passing is 4-5x more effective than running on offense and 3x more effective on defense.

but yeah, go ahead and eyeball.

[boratt]

I would have said turnover/passing/running/penalty.  Close enough. 

[skibrett15]

1 minute ago, boratt said:

I would have said turnover/passing/running/penalty.  Close enough. 

but you didn't? and it isn't?  and the degree of difference is what's interesting to me.

 

If passing efficiency is 5x more imporant to winning, why are OL evaluated whatsoever on their run blocking?

 

On a separate conversation, keep in mind these are all Y/A measures, not volume numbers.  So they don't suggest you should run it 80% of the time.  They do suggest that your best way to improve the win total is to improve the passing game efficiency.

On a related separate note, why not run it 20% of the time?  Would it be worth bringing out your 2nd qb if arm tiring was really that big of a deal from throwing it 60+ times?

[boratt]

27 minutes ago, skibrett15 said:

but you didn't? and it isn't?  and the degree of difference is what's interesting to me.

 

If passing efficiency is 5x more imporant to winning, why are OL evaluated whatsoever on their run blocking?

 

< 4x  you mean? But still, I get the point. Passing is significantly more important than running in relation to wins.