Still doubt that passing is more important that rushing?

[percyhoward]

Winning teams in games are actually more efficient passing when BEHIND than when they are ahead. The same is true of losing teams in games -- they are more efficient when BEHIND than when they are ahead. Both winners and losers have a higher ANYPA and higher passer rating when behind than when they are ahead.

That's extremely misleading, because winners that were behind at some point must have performed well in order to come back and win, and losers that were ahead at some point must have performed poorly in order to end up losing.

since 2010
Winning team when trailing 106.7
Losing team when leading 68.8

Teams pass more efficiently when they're ahead than when they're behind.
when trailing 82.7
when leading 91.9

 

[Craig]

People have better ratings when behind because of prevent defense and garbage time.

 

[percyhoward]

People have better ratings when behind because of prevent defense and garbage time.

No matter how far back you look, teams pass better when they're leading.

since 2010
when trailing 82.7
when leading 91.9

since 2000
when trailing 78.5
when leading 88.9

since 1994
when trailing 77.4
when leading 87.3

 

[xwalker]

The fact that you have to quantify the difference in your "stat" to a significant degree in order to show the same correlation as ANY difference in ANYPA, and the fact that 85.5 PERCENT OF YOUR "STAT" IS BASED ON PASS PLAYS, show how laughable your position is.

ANYPA is a manipulated composite stat.

ANYPA does not appear to correlate any better than just the old axiom that the team that wins the turnover battle usually wins. Turnovers correlate to winning between 80% and 92% of the time of similar numbers depending on the exact time frame of reference.

The fact that ANYPA contains INTs as part of the formula in an effort to improve "correlation" is comical.

You continue to repeat that rushing does not correlate to winning or to improving passing, but you're generally using rushing yards in a specific game as your determining factor. I've shown multiple reason that trying to measure the impact of rushing in football using yards is silly.

Again 2 offenses can have 50 yards against the same defense but that defense can play 8 in the box against 1 offense and 7 in the box against the other offense. The available stats can't differentiate that simple change by the defense adjusting to limit rushing; however, it obviously makes passing easier when there is one less defender in coverage.

The other primary issue from a statistical viewpoint is that rushing yards tend to be around 100 per game and passing yards tend to be around 300 per games. It is obvious from the simplest mathematical viewpoint that a differential in rushing as measured in yards it not going to correlate like passing because rushing as measured in yards is a small percentage of the overall yards gained in a game.

The bottom line is that everyone knows that rushing has some impact on the passing game but that impact is currently undefined mathematically. The fact that you can't define it mathematically does not in any way suggest that it does not exist. You continue to try to use per game rushing yardage differential to define it when as above I've explained why that is a waste of time.

 

[Nightman]

I read a story that Belichick told his NYGiant players to let Thurman Thomas rush for 100 yds in Super Bowl XXV. If they did they would win. He said BUF had such a great passing attack that if they wasted time and plays running the ball it would hurt them.

It almost backfired when Thomas sprinted for a 31 yd TD in the 4th to give BUF a 19-17 lead. They allowed Thomas to get 135 yds in the game and if Scott Norwood is a little better BUF wins.

But just like this discussion it is hard to take away any lessons from that one game. NYG did what Belichick asked and won, but if Norwood makes a FG which has nothing to do with rushing or passing effeciently, BUF wins.

 

[AdamJT13]

That's extremely misleading, because winners that were behind at some point must have performed well in order to come back and win, and losers that were ahead at some point must have performed poorly in order to end up losing.

Um, that's partly the point -- it refutes the notion that teams pass more effectively once they're already ahead and automatically pass less effectively when they're behind, which is what was being argued.

Over the past 10 seasons, winning teams averaged 5.96 ANYPA when ahead, 6.06 ANYPA when tied and 6.56 ANYPA when behind. When they were already ahead, their passing actually got LESS effective. And when they were behind, it got MORE effective than when they were ahead or even tied. (Using passer rating, it's the same -- 97.0 when already ahead, 97.4 when tied and 104.7 when behind.)

Losing teams reflect the same trend -- 3.34 ANYPA when ahead, 3.82 when tied and 3.83 when behind. Again, their passing is LEAST effective when they are ahead. Their passer rating of 66.1 when ahead is also lower than their passer rating of 73.3 when tied or 72.9 when behind.

Clearly, there is NO evidence that being ahead CAUSES the winning team to be more effective passing because it's "easier" passing when ahead.

As I mentioned earlier, the reason why the overall stats make it appear that it's easier to pass when ahead (90.9 rating when ahead, 80.9 rating when behind) is because the teams that pass better are usually the ones that are ahead -- BECAUSE they pass better!

 

[AdamJT13]

ANYPA is a manipulated composite stat.

ANYPA does not appear to correlate any better than just the old axiom that the team that wins the turnover battle usually wins. Turnovers correlate to winning between 80% and 92% of the time of similar numbers depending on the exact time frame of reference.

The fact that ANYPA contains INTs as part of the formula in an effort to improve "correlation" is comical.

Only 64 percent of winning teams actually win the turnover battle. One reason for that is that in almost 20 percent of all games, NEITHER team wins the turnover battle -- each team commits the same number of turnovers. When almost 20 percent of all games must be discounted entirely, it's not the most useful correlation. Besides, simple YPA also has a very high correlation to winning in the NFL.

The bottom line is that everyone knows that rushing has some impact on the passing game but that impact is currently undefined mathematically. The fact that you can't define it mathematically does not in any way suggest that it does not exist.

Yes, we all recognize that there is minimal correlation between rushing effectiveness and winning in the NFL. What you're saying is the same as saying that nobody can prove that a 30,000-ton blue and silver platypus doesn't exist somewhere on the earth. Just because there is no proof that it does exist doesn't mean that it doesn't. Same thing. Of course, in both cases, the overwhelming evidence points to the contrary, but you can choose what you want to believe.