Baseball Home Field Advantage Without Fans in the Stands

Abstract

Home field advantage is universally accepted across most major sports and levels of competition. However, exact causes of home field advantage have been difficult to disentangle. The COVID-19 pandemic offers a unique, natural experiment to isolate elements related to home field advantage since all 2020 regular season Major League Baseball games were played without fans. Results provide no statistically significant evidence of a difference in home field advantage between the 2019 and 2020 seasons, evidence that home crowd support is not a driver of home field advantage. There does appear to be a statistical advantage by the home team batting second in the inning. Travel fatigue seems to have no impact on home field advantage, and while home field advantage seems to increase throughout the 2020 season, we chalk that up to small sample noise. Despite lacking historical precedence, betting markets seemingly respond efficiently to the new home conditions. Keywords: home field advantage, market efficiency, baseball, ghost games

False Start? An Analysis of NFL Penalties With and Without Crowds

“…The same cannot be said about Major League Baseball (MLB) however. Losak and Sabel (2021) find no evidence that home team win percentage was different in 2020 (when MLB teams played in front of no fans) than in the 2019 season, though this was not uniform across the season.…”

Separately measuring home‐field advantage for offenses and defenses: A panel‐data study of constituent channels within collegiate American football

“…In the NFL, Ehrlich et al (2021) account for distance and use dummies for full-, reduced-, and zero-crowd games, but they do not include a continuous crowd size or density measure, while Smith et al (1997) account for distance and time zones but not crowd, familiarity, or other factors. Our main specification advances the literature within college football by introducing various measures of familiarity, including team-stadium, opponent-stadium, and team-opponent familiarity, which loosely resemble some familiarity measures used in other sports (e.g., Boudreaux et al, 2017;Fischer & Haucap, 2021;Losak & Sabel, 2021;McHill & Chinoy, 2020). In the various robustness checks and model extensions we explore in Section 6, we also introduce crowd density (e.g., Böheim et al, 2019;Inan, 2020;Schwartz & Barsky, 1977), weather (utilized, at least to some extent, by Cross & Uhrig, 2020;Fischer & Haucap, 2021;Losak & Sabel, 2021), and head-coach familiarity (perhaps novel to the entire home-field advantage literature, though Fischer & Haucap, 2021, utilize a dummy variable indicating head coaches in their first season with their team) into the college football literature.…”

“…Our main specification advances the literature within college football by introducing various measures of familiarity, including team-stadium, opponent-stadium, and team-opponent familiarity, which loosely resemble some familiarity measures used in other sports (e.g., Boudreaux et al, 2017;Fischer & Haucap, 2021;Losak & Sabel, 2021;McHill & Chinoy, 2020). In the various robustness checks and model extensions we explore in Section 6, we also introduce crowd density (e.g., Böheim et al, 2019;Inan, 2020;Schwartz & Barsky, 1977), weather (utilized, at least to some extent, by Cross & Uhrig, 2020;Fischer & Haucap, 2021;Losak & Sabel, 2021), and head-coach familiarity (perhaps novel to the entire home-field advantage literature, though Fischer & Haucap, 2021, utilize a dummy variable indicating head coaches in their first season with their team) into the college football literature. 25 For example, suppose our model estimate indicates that a 10,000-person increase in crowd size leads the home team to score 2 fewer points and leads the away team to score 5 fewer points, ceteris paribus.…”

Behavioral Biases in Daily Fantasy Baseball: The Case of the Hot Hand

“…This variable should also capture the following team-specific information that would be relevant to fantasy scoring: park factors, opposing pitcher effects (starting pitcher and bullpen), weather effects, and fatigue/travel factors, among others. While this variable may also capture home team effects, we include Hom normale i , t as a separate variable to address any potential betting market price inefficiencies related to home team bias (see Gandar et al, 2002; Losak & Sabel, 2021; Paul et al, 2008 as examples where home bias is considered in baseball betting markets).…”