Predicting the Winner of Tied National Football League Games

Abstract

We construct a data set of all 429 tied at the half regular season National Football League (NFL) games between 1994 and 2012. We then examine whether or not the path taken to reach the tie (e.g., rushing yards, turnovers, etc.) has any ability to predict the eventual winner. Our main finding is that only the point spread is significantly predictive, although there is weak evidence to suggest that allowing more sacks reduces the chances of winning. Surprisingly, we find that the team receiving the first possession of the second half does not enjoy a statistically significant advantage. Teams should thus simply try to maximize their first half lead without expecting that first half strategies such as “establishing the run” will pay dividends in the second half.

Keywords

football sports strategy forecasting

Get full access to this article

View all access options for this article.

References

Alomar

(2006). The passing premium puzzle. Journal of Quantitative Analysis in Sports, 2, 1–10.

Alomar

(2010). Measuring risk in NFL playcalling. Journal of Quantitative Analysis in Sports, 6, 1559–0410.

Kovash

Levitt

(2009). Professionals do not play minimax: Evidence from major league baseball and the national football league. Mozilla University and University of Chicago. (NBER Working paper no. 15347).

Levitt

(2004). Why are gambling markets organised so differently from financial markets? Economic Journal, 114, 223–246.

Rockerbie

(2008). The passing premium puzzle revisited. Journal of Quantitative Analysis in Sports, 4, 1–3.

Romer

(2006). Do firms maximize? Evidence From professional football. Journal of Political Economy, 114, 340–365.

Sandford

Shea

(2013). Optimal setting of point spreads. Economica, 80, 149–170.

Schatz

(2003). The establishment clause. Football Outsiders. http://www.footballoutsiders.com/stat-analysis/2003/establishment-clause