Are NFL Coaches Starting to Listen to Economists?

Are NFL coaches starting to listen to economists?

My gut feeling is that the answer to that question is almost certainly a resounding “no.” But there are three pieces of data that at least hint at the possibility that economists might be making some headway.

The first bit of evidence is the play-calling of Atlanta Falcons coach Mike Smith, as reported by Michael Salfino in a Wall Street Journal article this past Saturday.? Nearly a decade ago, I was the editor on David Romer‘s Journal of Political Economy paper that demonstrated that NFL coaches were far too cautious when it came to choosing to punt on fourth down.? This year, on fourth down and three or fewer yards in the opponent’s territory, the Falcons went for it 72.2 percent of the time.? I don’t have the data from ten years ago available to me, but I believe no team would have been close to this percentage back then.? It paid off for the Falcons: they converted 84.6 percent of the time, increasing their points scored by about 30 points and possibly adding three wins.? Bill Belichick of the New England Patriots was an early convert to Romer’s analysis – I am told Belichick read the paper carefully.? The Patriots were the fourth most likely team to go for it on fourth down this year, converting all seven times they tried on fourth and short in their opponent’s territory.

The second bit of evidence relates to my own working paper, co-authored with Kenneth Kovash, that we released in 2009.? In that paper, we pointed out another apparent inefficiency in NFL play calling: too many runs and not enough passes.? Even after taking into account the higher likelihood of a turnover when passing, according to our analysis a passing play generates on average .07 extra points in expectation.? The teams that pass the most score an extra 14 points a year relative to the teams that run the most, by my calculations.? For what it is worth, the percentage of offensive plays that were passes rose to 55.4 percent this year from 54.7 percent.? Not a big change, and no doubt pure coincidence, but at least the trend is in the right direction.

The third bit of evidence is no doubt the strongest.? After Ken Kovash and I wrote that paper on football, an NFL team hired him!? (Ken follows in the footsteps of another of my former students, Mike Zarren, who is the stats guy for the Celtics.)? I’m happy to report that his team’s ratio of passes to runs jumped far more than the league average, so maybe they are even listening to him.


With respect to passing more: are NFL coaches listening to economists, or are they just listening to college coaches?


And how did that work out for the Cowboys this year?

Michael Radosevich

NFL coaches live in a highly competitive world. The universal thought is, there are 29 other men trying to beat me. So the coaches scour high and low for information, and of course they take into account the analyses you mention.

It's just too bad that our politicians don't face the same kind of competition. If Pres. Obama had listened to Paul krugman and other eminent economists and followed their advice on the size of his 2009 "stimulus", the country - not to mention Mr. Obama - would be in a far better position today.


I agree with Erik. To a great extent, the pro game seems to be the most cautious, with innovation coming from below. I wonder how that compares with other sports.

Ian Kemmish

Why not just ask the coaches? An ounce of knowledge is always worth a pound of speculation. (Unless, of course, you're paid to fill a newspaper with a specified quantity of speculation.....)


Statistically, are teams better off punting to Devin Hester or giving up field position and punting out of bounds?

Brian S

Then again, let's take a look at Dallas' performance this year vs. that of the previous ten.


But the cowboys did much better when they started to run the ball more often.



However, even if passing leads to more points, that doesn't always make it the right thing to do. Often, the reason to run more than pass is to keep the opposing offense off the field. This is especially true if you're playing a team with a great QB, such as Brady, Manning, or Brees.

In my mind, the greatest example of this was the NY Giants Superbowl win at the end of the 1990 season. The Giants were given little chance against the high-powered Buffalo Bills. With a power running game featuring Otis Anderson, the Giants held the ball for 40 minutes. Jim Kelly, Thurman Thomas, and company were turned into spectators.

The clock stops after an incomplete pass. It keeps moving after a run.

I would be interested to see your analysis of the running and time of possession as it relates to wins and losses.


In response to too much running opposed to passing; in the Jets win over the Colts, I think running the ball successfully was crucial. The Jets used up just shy of 10 minutes of game clock on one drive, because they could run the ball, and mix in some short passes. This kept Peyton Manning on the sidelines where he was no danger. The Jets can't win a shootout against Peyton Manning. They need to control the clock, keep the score down, and run the ball.


Did you notice that the Cowboys had one of their worst seasons in history?


"The teams that pass the most score an extra 14 points a year relative to the teams that run the most, by my calculations."

Did you also calculate how many points-against the increased number of interceptions leads to?

Nick Harris

I agree with Kevin's comments. I would be interested in knowing how the passing attempts number relates to wins and season outcomes.


I would also like to see how passing attempts correlates to wins. Teams like the Jets probably run more to protect a young and inexperienced quarterback. So play-calling may be more a function of personnel (especially QB) than statistics.

Nick Harris

Also, I would be interested in knowing if you analysis accounted for weather/field conditions. My hypothesis is that outdoor games will account for a higher number of points per pass attempts in September than in December or January.

Jeff Needham

The Chiefs' Todd Haley went for it on 4th down a TON of times this year, often recklessly. Haley is of course a Belichick/Parcells understudy.

Steve Nations

"It paid off for the Falcons: they converted 84.6 percent of the time, increasing their points scored by about 30 points and possibly adding three wins." Take away three wins and the Falcons would not have had a bye the first week of the playoffs. (I'm not even sure they would have been in the playoffs.)

Joe Posnanski had an interesting post a couple days ago in which he noted that the winning percentage of playoff teams at home after a bye has gone down rather dramatically in the last 5-or-so years. Could it be that some coaches are doing the things during the regular season, like going for it on fourth down, that improve their odds over the regular season.

This works over the long run, giving them a couple extra victories and a good playoff seeding. But then during a single playoff game those small decisions aren't such game changers, and the team's ability comes up short to other teams whose coaches didn't use the most sophisticated tactics. This would make the playoff outcomes appear to be more random.



You can't assume that running and passing are independent. Most obviously, running sets up the pass - if you can't or don't run the ball, the pass is easier to defend because you can't play-fake and the defence has less to worry about.

Tom K.

1. Teams also punt based on how good their own defense is.

2. You need a good quarterback in order to efficiently to throw the ball more. If I were the quarterback for the Packers, the final score would be Packers 0, Bears 210. that is not so with Rodgers. The Packers with me as a quarterback would do much better when running the ball.

The individual data point of who is the quarterback is far and away more of a significant factor in scoring more points that merely "throwing the ball more". Assumptions of equal player talent is patently a false analysis.

Cowboys are a terrible example of a winning football formula.

John Murphy

The Saints pass more - and score more - because they have Drew Brees. Teams with a crappy QB pass less, and score less. This analysis does not take into account what players you have. If you have Adrian Peterson, Jamaal Charles, and Derek Anderson - you should run the ball, again and again and again.