Game-related statistics that discriminate between winning and losing teams in the National Basketball Association

Abstract

Possessing knowledge about the most crucial elements that distinguish between winning and losing teams in basketball is of great importance to coaches, particularly when the game score is close and the difference in performance between the two teams is small. The main purpose of the present study was to identify those basketball game-related statistics, commonly found in a box-score, as well as adapted game-related statistics, specific to the four factors, which best discriminate between winning and losing teams in the National Basketball Association (NBA). The data from a sample of 500 games from the 2012-2013 NBA regular season were gathered from the official website of the NBA. Games were divided into balanced games (≤ 8 points difference) and unbalanced games (> 8 points difference). An independent t-test was used to compare the means of 21 gamerelated statistics across all games. Mann-Whitney U tests were used within balanced and unbalanced games. Eighteen significant game-related statistics, when all games were considered, were reduced to 17 in unbalanced games, and 15 in balanced games. The analysis of balanced games revealed statistically significant differences between winning and losing teams in the game-related statistics of; field goals attempted (P < 0.05), field goals made (P < 0.001), field goal percentage (P < 0.001), effective field goal percentage (P < 0.001), 3-point field goals attempted (P < 0.001), 3-point field goal percentage (P < 0.05), free throws attempted (P < 0.001), free throw made (P < 0.001), free throw rate (P < 0.001), defensive rebounds (P < 0.001), total rebounds (P < 0.001), assists (P < 0.001), blocks (P < 0.01), steals (P < 0.05), and personal fouls (P < 0.01). These results can be used by coaches and players to inform training sessions, game strategies, and tactics, as well as by scouts in the player recruitment process