Danielle Jarvie1. 1. Claremont Graduate University, Claremont, CA, USA. danielle.jarvie@cgu.edu.
Abstract
OBJECTIVES: To estimate the effects of team-mate shared experience on overall team performance as well as to determine whether concentration of time together among subgroups of players and/or focal players enhances team performance. METHODS: Social network analysis (SNA) was used to model 30 active Major League Baseball teams from 2006 to 2015 with years of experience together connecting players resulting in 300 individual team networks. Social network metrics of network density, network centralization, and average weighted degree were computed and analyzed with team attributes by generalized least squares regression to predict wins, and team rank. Logistic regression was used to predict binary outcomes of world series and division wins. RESULTS: Network density was negatively associated with team rank (β = - 0.115, p = .05), while average weighted degree was positively associated with team rank (β = 0.147, p = .01). On average, each extra year of shared player time per team was associated with 14.86% higher probability of winning a division title (B = 2.69, exp(B) = 14.86, p = .05). Each extra shared year of infield membership among team-mates predicted 2.4% lower odds of winning the world series (B = -0.024, exp(B) = 0.976, p = .01), and each extra shared year between outfield players predicted 2.9% lower probability of winning a team's division (B = -0.029. exp(B) = 0.972, p = .05). CONCLUSIONS: Prolonged shared time between players is beneficial when it is spread evenly among all players of the team, whereas having few focal players who have been on a team together for many years is a disadvantage to overall performance.
OBJECTIVES: To estimate the effects of team-mate shared experience on overall team performance as well as to determine whether concentration of time together among subgroups of players and/or focal players enhances team performance. METHODS: Social network analysis (SNA) was used to model 30 active Major League Baseball teams from 2006 to 2015 with years of experience together connecting players resulting in 300 individual team networks. Social network metrics of network density, network centralization, and average weighted degree were computed and analyzed with team attributes by generalized least squares regression to predict wins, and team rank. Logistic regression was used to predict binary outcomes of world series and division wins. RESULTS: Network density was negatively associated with team rank (β = - 0.115, p = .05), while average weighted degree was positively associated with team rank (β = 0.147, p = .01). On average, each extra year of shared player time per team was associated with 14.86% higher probability of winning a division title (B = 2.69, exp(B) = 14.86, p = .05). Each extra shared year of infield membership among team-mates predicted 2.4% lower odds of winning the world series (B = -0.024, exp(B) = 0.976, p = .01), and each extra shared year between outfield players predicted 2.9% lower probability of winning a team's division (B = -0.029. exp(B) = 0.972, p = .05). CONCLUSIONS: Prolonged shared time between players is beneficial when it is spread evenly among all players of the team, whereas having few focal players who have been on a team together for many years is a disadvantage to overall performance.
Authors: Pedro T Esteves; Duarte Araújo; Keith Davids; Luís Vilar; Bruno Travassos; Carlos Esteves Journal: J Sports Sci Date: 2012-08-01 Impact factor: 3.337
Authors: João Ribeiro; Keith Davids; Duarte Araújo; Pedro Silva; João Ramos; Rui Lopes; Júlio Garganta Journal: Sports Med Date: 2019-09 Impact factor: 11.136