Statistical Analysis of 2024 Women’s College Basketball
sports
data
Published
May 22, 2023
Code
# Create dataframe for offensive and defensive team statsteam_stats = pd.merge(raw_stats, raw_opponent_stats, on="School", how="inner")team_stats.columns = team_stats.columns.str.replace('_x', '')team_stats = team_stats[['School', 'Overall_SRS', 'Overall_SOS', 'Overall_W', 'Overall_L','Team_FG', 'Team_FGA', 'Team_3P', 'Team_3PA', 'Team_FTA', 'Team_ORB', 'Team_TRB', 'Team_TOV','Opponent_FG', 'Opponent_FGA', 'Opponent_3P', 'Opponent_3PA', 'Opponent_FTA', 'Opponent_ORB', 'Opponent_TRB', 'Opponent_TOV']].copy()# Set datatypeteam_stats = team_stats.astype({'Overall_SRS': 'float','Overall_SOS': 'float','Overall_W': 'int','Overall_L': 'int'})# Cast all other columns to floatcast_to_float = team_stats.columns[5:]team_stats[cast_to_float] = team_stats[cast_to_float].astype(float)team_stats.dtypes# Divide teams into SRS percentilebins = [0, 0.25, 0.5, 0.8, 1]labels = ['Bottom 25%', '25th-50th %', '50th-80th %', 'Top 20%']team_stats['Pct_Group'] = pd.qcut(team_stats['Overall_SRS'], q=bins, labels=labels)
Statistical Analysis of 2024 Women’s College Basketball
Unlike Men’s College Basketball, statistical analysis on the Women’s game is relatively sparse. Since “Everybody watches Women’s Sports”, I decided to dig into the data
Analysis for the Men’s game is widely available and well documented, so I applied some of the most popular methodologies to basic team stats for the Women’s game. In this article, I’ll focus on calculating:
Four Factors
Offensive and Defensive ratings
These stats, are most commonly used to better understanding the strengths of teams, predict the win probability of matchups, and identify the games that might be most exciting.
The data used here comes from Sports Reference. While some of the data we be calculating could be pulled directly from Sports Reference’s Advance Stats, we still calculate each stats since some of the method may deviate slightly.
To build a winning March Madness Bracket, we need to know when two teams match up against each other, which team is more likely to win? A popular way to measure this statiscally has been the “Four Factors of Basketball Success” from Dean Oliver in his paper “Basketball on Paper”. Similar to the moneyball of Basketball.
Four Factors of Basketball
From a data analysis perspective, the four factors of success in basketball boils down to:
In order for us to measure the offensive and defensive efficiency of each team, we’ll first calculate these stats for each team, this is their offense. For defense, we’ll calculate how well their opponents do in each of these stats when playing against them.
Effective Field Goal % (eFG%) {#sec-eFG%}
Code
# Calculate both offensive and defensive effective field goal percentageteam_stats['Calc_Off_eFG%'] = (team_stats['Team_FG'] +0.5* team_stats['Team_3P']) / team_stats['Team_FGA']team_stats['Calc_Def_eFG%'] = (team_stats['Opponent_FG'] +0.5* team_stats['Opponent_3P']) / team_stats['Opponent_FGA']
The effective field goal % captures the teams ability to shoot the ball. Because at the end of the game, the team with the most points wins. If you’re not scoring, you can’t win. This is calculated as Field Goals Made + 0.5 * 3-pointers Made) / Field Goals Attempted. Unlike FG%, this calculation adds 50% more credit for 3-pointers made, since they are worth more points.
Once each of these metrics has been calculated, let’s evaulate the impact each metric has on the overall simple rating system of the team.
Code
# Set up the figure with four subplotsfig, axes = plt.subplots(2, 2, figsize=(12, 10)) # 2x2 gridoff_four_factors = team_stats[['School', 'Overall_SRS', 'Calc_Off_eFG%', 'Calc_ORB%', 'Calc_Off_TOV%', 'Calc_Off_FTR']]# List of offensive factorsoffensive_factors = ['Calc_Off_eFG%', 'Calc_ORB%', 'Calc_Off_TOV%', 'Calc_Off_FTR']# Loop through the factors and plot each one against 'Overall_SRS'for ax, factor inzip(axes.flatten(), offensive_factors): sns.regplot( data=off_four_factors, x=factor, y='Overall_SRS', ax=ax, scatter_kws={'s': 10}, line_kws={'color': 'red', 'linewidth': 0.8} ) ax.set_title(f'Relationship between {factor} and Overall_SRS')plt.suptitle('Impact of Offensive Four Factors on Overall SRS', y=1.02)plt.tight_layout()plt.show()
Code
# Set up the figure with four subplotsfig, axes = plt.subplots(2, 2, figsize=(12, 10)) # 2x2 griddef_four_factors = team_stats[['School', 'Overall_SRS', 'Calc_Def_eFG%', 'Calc_Opp_ORB%', 'Calc_Def_TOV%', 'Calc_Def_FTR']]# List of defensive factorsdefensive_factors = ['Calc_Def_eFG%', 'Calc_Opp_ORB%', 'Calc_Def_TOV%', 'Calc_Def_FTR']# Loop through the factors and plot each one against 'Overall_SRS'for ax, factor inzip(axes.flatten(), defensive_factors): sns.regplot( data=def_four_factors, x=factor, y='Overall_SRS', ax=ax, scatter_kws={'s': 10}, line_kws={'color': 'red', 'linewidth': 0.8} ) ax.set_title(f'Relationship between {factor} and Overall_SRS')plt.suptitle('Impact of Defensive Four Factors on Overall SRS', y=1.02)plt.tight_layout()plt.show()
In all calculations, I use per game average since as a basketball fan, it allows me to reason about the data and catch mistakes easier. For example, if free throw attemps is 78, I’m able to recognize and issue witih my data. To increase accuracy calculations could be done on a per game basis before averaged out.
Once we have all four factors calculated, both for offense and defense, we can see how each factor impacts their SRS rating.
In many of the analysis, we’ll also compare Sports Reference’s Simple Rating System (SRS) with the stats that we calculate. The Overall SRS, “takes into account average point differential and strength of schedule”, where zero is average and a high positive number signals a strong team.
