Journal of Sports Analytics

This paper uses data from the UEFA Champions League (2003–2012) to study the impact of diversity on team performance. Results indicate that more heterogeneous teams outperform less diverse sides; a one-standard deviation increase in cultural diversity (measured by linguistic distance) can double a team’s goal differential over the course of the tournament. One threat to our conclusions is that certain teams have greater resources to search the world for talent. We address this issue by controlling for players’ transfer values, quality ratings, and exploiting exogenous variation in diversity generated by differences in the non-European player quotas of national soccer leagues.

We develop a model for pitch sequencing in baseball that is defined by pitch-to-pitch correlation in location, velocity, and movement. The correlations quantify the average similarity of consecutive pitches and provide a measure of the batter’s ability to predict the properties of the upcoming pitch. We examine the characteristics of the model for a set of major league pitchers using PITCHf/x data for nearly three million pitches thrown over seven major league seasons. After partitioning the data according to batter handedness, we show that a pitcher’s correlations for velocity and movement are persistent from year-to-year. We also show that pitch-to-pitch correlations are significant in a model for pitcher strikeout rate and that a higher correlation, other factors being equal, is predictive of fewer strikeouts. This finding is consistent with experiments showing that swing errors by experienced batters tend to increase as the differences between the properties of consecutive pitches increase. We provide examples that demonstrate the role of pitch-to-pitch correlation in the strikeout rate model.

A sports game is about competition. The competitiveness of a game is important in terms of fan interest. Teams ahead or behind in a non-competitive game may also be more likely to substitute reserve players to reduce risk of injury to key players or gain experience for lesser-used players. Changes in how a team plays in a non-competitive game also impact secondary competitions such as betting or fantasy sports due to player behavior changes in a non-competitive game.

This work examines the competitiveness of games in six professional sports leagues using a variety of metrics and finds that Major League Baseball (MLB) games are clearly the least competitive compared with games from each of the other leagues in our study. MLB games have the highest percentage of game segments that are played in less competitive situations. Major League Soccer and Barclay’s Premier League games tend to be the most competitive in general; largely because about half of game time for these leagues is spent with the game tied. However, if a team does take the lead in one of these leagues then they, along with MLB, have the highest chance that this team will not relinquish the lead.

To evaluate player and position importance on the BYU football team, we used the coaches’ play-by-play grades of each player as explanatory variables, with the response of expected points gained or lost on each play. Expected points were determined using an analysis of NCAA Football Bowl Subdivision (FBS) teams play-by-play data from 2005–2013 implementing the tiered polychotomous regression model of White and Berry (2002). We used a Bayesian hierarchical linear model with first-level parameters of player and second-level parameters of position to estimate the effect or “impact” each player had on the expected points gained or lost each play. We then used this model to identify the relative importance of each player and each position on the team.

National Football League (NFL) kickers have displayed improvement in both range and accuracy in recent years. NFL management in turn has displayed a rather low tolerance for missed field goals, particularly in game-deciding situations. However, these actions may be a consequence of a perceived appreciable variability in NFL kicker ability. In this paper, we consider shrinkage estimation of NFL kicker field goal success probabilities. The idea derives from the literature on estimating batting averages in baseball, though the classic James-Stein shrinkage approaches there do not apply to independent binary field goal attempt trials. We study a variety of weighting schemes for shrinking model-based kicker-specific field goal success probability estimates to a league-wide estimate, as a function of distance. As part of the development, we briefly detail collecting NFL play-by-play data with the R statistical software package, identify the complementary log-log link function as preferable to the more commonly applied logit link function in a generalized linear model for field goal success, and demonstrate the desired variance-reduction, both in and out of sample, enjoyed by our proposed shrinkage estimators. We illustrate our methods by ranking NFL kickers from 1998 to 2014, analyzing individual kicker success at mid-range and long-range field goal attempts, and studying kicker ability over the last decade. Stadium effects are added to the model and found to be highly significant and to have an impact on the kicker rankings.