COVID-19 countermeasures, Major League Baseball, and the home field advantage: Simulating the 2020 season using logit regression and a neural network

Background: In the wake of COVID-19, almost all major league sports have been either cancelled or postponed. The sports industry suffered a major blow with the uncertainty of sporting events being held in the near future. Various scenarios of how and when sports might recommence have been discussed. This paper examines various scenarios of how Major League Baseball team performance is going to be impacted by the presence of fans, or the lack thereof, in the context of physical distancing and other COVID-19 countermeasures Methods: The paper simulates, using a neural network and a logit regression model, the win-loss probabilities for various scenarios under consideration and also estimates the home effect for each team using data for the 2017-2019 seasons. Results: The model demonstrates that individual team home effect is symmetric between home and away and teams will not necessarily have a win or loss of any additional games in neutral stadiums, as teams with a high home field effect will lose more neutral games that would have been at home but will win more neutral games that would have been away. However, the result of individual games will be different since home effect is asymmetric between teams. Our simulation demonstrates that these individual game differences may lead to a slight difference in Play-Off Berths between a full season, a half season, or a full season without fans. Conclusions: Without fans, any advantage (or disadvantage) from home field advantage is removed. Our models and simulation demonstrate that this will reduce the variance. This stabilizes the outcome based upon true team talent, which we estimate will cause a larger divide between the best and worst teams. This estimation helps decision makers understand how individual team performance will be impacted as they prepare for the 2020 season under the new circumstances.