Is Running a High Number of Plays Good? Opportunity Cost in College Football
The Wall Street Journal published an article yesterday that argued that the reason Syracuse beat Clemson in 2017 was because Syracuse ran a lot of plays during the game. [Paywall]
No mention of Clemson losing their starting quarterback Kelly Bryant just before the end of the first half was mentioned, because according to the article, the reason Syracuse defeated one of the best teams in all of FBS was because they ran a lot of plays. Yes, Syracuse ran a lot of plays, 89 in the 2017 game against Clemson.
As you might have guessed, I disagree; because running a play is actually a cost (in economic terms we call it an opportunity cost), because if the team does not do enough positive things to overcome using the scarce resource (play); this reduces a teams ability to score. Just like "outs" are an opportunity cost in baseball (think Moneyball "when your enemy is making mistakes, don't interrupt them"), or shot attempts in basketball.
Plays in football are an opportunity cost. We see this in pro football in research done by David Berri, as well as referred to in our book, The Wages of Wins. Running a play has a statistically significant negative impact on the ability of NFL teams to score points. Likewise, the same is true in the FBS. For the 2017 season, I estimate that each play results in (-0.444) points scored. Likewise the more plays your opponent runs results in your opponent reducing their ability to score points by an estimated (-0.359) points scored.
To overcome this opportunity cost, offenses need to do things that are positive on the field to increase their ability to score points (gain yards, make first downs, and most importantly - not turn the ball over) and defenses need to do things that stop their opponent from doing things that increase their ability to score points.
So, let's go back to Syracuse. Syracuse defeated what ended up being the #10 ranked team overall using my Complex Invasion College Football Production Model for the Football Bowl Subdivision. So was it more likely due to Syracuse playing a lot of plays or that Clemson's starting quarterback was missing for half the game? If running lots of plays, then Syracuse should also be successful against other teams as well; but Syracuse finished up at 4 wins and 8 losses. In terms of the my Complex Invasion College Football Production Model for the Football Bowl Subdivision Syracuse finished as the #104 ranked team in total productivity, with the #87 ranked team in offensive production and the #109 ranked team in defensive production. That Syracuse team was below average on both sides of the ball, and as such, running plays doesn't seem to be the deciding factor.