Premium
Analysis of monday night football viewership
Author(s) -
Shetty Bhupesh,
Ohlmann Jeffrey W.,
Gaeth Gary J.
Publication year - 2016
Publication title -
statistical analysis and data mining: the asa data science journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.381
H-Index - 33
eISSN - 1932-1872
pISSN - 1932-1864
DOI - 10.1002/sam.11317
Subject(s) - audience measurement , football , league , schedule , descriptive statistics , regression analysis , computer science , statistics , econometrics , advertising , geography , mathematics , business , astronomy , operating system , physics , archaeology
We conduct a three‐pronged analysis of one of the most watched television series in history: Monday Night Football (MNF). First, we identify factors that affect the viewership of MNF from 1993 to 2014. This descriptive model explains over 90% of the variability in viewership, but includes factors known only ex post facto and the week of the game. As the Monday Night Football schedule must be set prior to the beginning of the season, we construct an alternative regression model to predict the full season's viewership of potential games. This predictive model relies only on factors that are known before the release of the National Football League schedule in April preceding the season. Using the predictive regression model to estimate the objective function coefficients for potential games, we then recommend MNF schedules that maximize the total season viewership using an integer programming formulation. We conduct simulations to determine the impact of forecast error on the structure of our optimal MNF schedule. © 2016 Wiley Periodicals, Inc. Statistical Analysis and Data Mining: The ASA Data Science Journal, 2016