Not all players feel at home: exploring individual differences in the home advantage phenomenon across 23 NBA seasons

This study examined inter-individual variability in player performance between home and away games, identifying distinct performance profiles and proposing strategies to mitigate away-game declines. Data from 23 NBA seasons (1997–1998 to 2022–2023) were analysed, excluding the 1998–1999, 2011–2012, and 2020–2021 seasons due to lockouts and pandemic restrictions. The dataset comprised 243,067 games and included season, game location, games played, points, rebounds (total, offensive and defensive), shooting percentages, assists, steals, blocks, turnovers, personal fouls, and efficiency. All variables were normalised per 48 minutes. A Principal Component Analysis off standardised home-away difference z-scores explained 58.66% of total variance. The first component (38.53%) was primarily associated with rebounds, blocks, and two-point percentage. The second component (20.13%) was associated with points, efficiency, turnovers, and assists. A k-means clustering (k = 3) revealed three player profiles: those who perform better at home, those who perform better away and those with a stable performance regardless of location. These profiles provide actionable information for coaches and team managers to optimise line-ups, rotations, and minute allocation according to location.