Thursday, January 21, 2010

More on passing, running, and winning in the NFL

++Addition++Steve offers more on 2009, deeming it the year only passing mattered. He wonders whether or not the NFL's popularity will decrease if a successful passing offense and a winning record continue to become increasingly synonymous with one another.

It's my impression that football votaries like smashmouth, low-scoring field position games a lot more than casual fans do. I'd guess more than a few of those casual fans would judge a typical 48-31 game more thrilling to watch than a 9-6 game. Assuming I'm in it to see the most competitive sixty minutes possible, I wouldn't--I'd take the second game for sure. But I'll still follow the NFL as high-scoring, pass-happy games become more common, whereas casual fans (especially women) will be unlikely to stick around if this year's Jets became the NFL archtype. So I'd guess the current trend will be good for business.

To the extent that passing is more important in the regular season than it is in the post-season, it's unfortunate as far as the league's level of popularity is concerned--the other way around would be more desirable. More indoor arenas and closed stadiums might help.


Steve wondered if the pattern in 2009 of passing's dominance over rushing as a determinant of success was carried over from the season before. It could have been a fluke year, or the beginning of a secular trend. The following table shows the correlations between winning and both passing and running yards per attempt by regular season going back to 2002:


* not statistically significant at 90% confidence

In 2008, the pass-run gap was less pronounced than in 2009, but still significant. However, 2007 was similar to the just-concluded regular season. As we move back in time, passing maintains the upperhand, but not as overwhelmingly so. In 2004, when the playoff-bound Falcons (Vick alone ran for nearly 1,000 yards that year), Steelers (Staley-Bettis one-two punch), and Jets (Curtis Martin in his best season before complications with a knee injury forced him into retirement) built their offenses around the running game, rushing came within striking distance of passing in importance.

As Steve has discussed, the nimble, impatient quarterback who's comfortable being flushed out of the pocket to run is not the future of football, it is the past. The nearly total domination of passing relative to running today suggests that a new paradigm is emerging, where the running game is clearly superceded in importance by the ability to move the ball through the air.

If a football votary is up for investing a little time in it, it'd be interesting to see how the pass-run dynamics correlate with success going back several decades.'s stats on per-play averages extend back over 60 years, so the data are there.


Chuck said...

This may be a nit-picking, but I wonder if sacks or busted up pass plays are part of yards per rushing attempt or yards per pass attempt.

I think that sacks are counted against rushing statistics, but really, since a sack is a failed passing play, that should be factored in for passing yards per attempt.

I doubt that would change the significance very much; who knows.

There are also trends in the aquirement of resources i.e. defensive lineman versus secondary. It seems to me that the 1990s heavily favored the run; the 2000s could have been less run oriented as defenses stacked up to combat the run. It's possible to account for teams' draft picks to find out the complete shift. Either way, this stuff is interesting.

It's really a purely economic problem.

OneSTDV said...

Interesting. Still have the ESPN talking heads yapping about "Running and defense wins in the playoffs". I wonder how the increasing importance of passing scales with the increase of domes. I'm not sure they're enough domes to make a difference, but it could have a small effect.

Oh and wasn't Staley injured most of 2004? 2004 was one of the best years in recent memory: Pats (14-2), Eagles (13-1 with TO), Steelers (15-1), and Colts (Peyton's 49 TD passes) all would have won championships in almost any other year this decade.

As for judging this trend versus history, you have to take into consideration rule changes. Back in the 70's, they implemented a ton of current rules that opened up the passing game tremendously. Probably the most known is the 5 yard rule (receivers get way better breaks, Peyton's timing offense and Rams in the Warner years never would have worked prior to those changes).

Anonymous said...

Sacks are counted against passing yards.

Steve Sailer said...


I'm guessing

"* not statistically significant at 90% confidence"

should read

"* statistically significant at 90% confidence"


Audacious Epigone said...


Anon's correct, sacks are counted against pass yardage.

In looking at draft picks, I wonder to what extent it becomes a chicken and egg question. Rule changes have steadily benefitted the passing offense. The only change I'm able to recall off the top of my head that has gone in the favor of defense is the abolition of the force out rule on catches.


Yeah, you're right about Staley. I'd forgotten. I was remembering Staley in on first and second and Bettis in on short yardage and inside the red zone in 2004, and then not hearing much else about Staley from that point on. But his career basically ended during the '04 season.


No, it indicates statistical insignificance. It's a little confusing because all the * are for runs, though it might appear as though it applies to both runs and passes during the season in question.

Steve Sailer said...

Okay, sorry, I was thinking the asterisk meant a lack of statistical significance between passing and rushing, but it means that rushing success wasn't significantly correlated with wins in most years.

Chuck said...


You're right about my question. But now I'm thinking about the nature of a pass play.

Does a sack count as a pass attempt? Also, does a scramble that goes for ca. 0 yards count as an attmept as well?

These might be counted against pass yardage, but do they factor in to pass attempts as well. The increase in attempts would have a larger effect on significance than the yards lost since the yards lost/gained is relatively small compared to the total number of passing yards a QB has per game.

Audacious Epigone said...


Good question. My guess is that it wouldn't make too much difference, but that's just speculation. If yardage is positive, it is counted in run statistics. I'm not certain about no gains, but strongly presume it comes down to where the ball is spotted. Past the line of scrimmage it's a run (since it's not a sack), behind the line it's a pass (because it is a sack, even if it is officially recorded as "no gain").

OneSTDV said...

Hmmm...9-6 vs 48-31.

They both suck, but I'd probably also go with the 9-6 game. The 4th quarter would generally be very tense. Not so much for the first 3 though.

My favorite team used to have these defensive battles vs their archrival from the mid 90's to the mid 2000's. The games would always end something like 13-10 or 17-9. They were almost unbearable to watch.