Does a higher field goal percentage predict more points scored by NBA players?

This analysis uses 2024-2025 NBA player statistics.

head(nbaClean)

## # A tibble: 6 × 3
##   Player        FGPercent   PPG
##   <chr>             <dbl> <dbl>
## 1 A.J. Green        0.431  7.4 
## 2 A.J. Lawson       0.425  9.1 
## 3 AJ Johnson        0.388  7.60
## 4 Aaron Gordon      0.536 14.7 
## 5 Aaron Holiday     0.442  5.5 
## 6 Aaron Nesmith     0.512 12

The scatter plot does not show a strong linear relationship between Field Goal Percentage and Points per Game, as players with a similar field goal percentage average a wide variety of points per game.

The regression line has a positive slope, which suggests that players with higher field goal percentages tend to score more points per game. However, there is a large amount of scatter which shows that field goal percentage alone does not strongly predict scoring.

model = lm(PPG ~ FGPercent, data = nbaClean)

The regression model shows a statistically significant positive relationship between field goal percentage and points per game. However, the model has an \(R^2\) of only 0.061, meaning field goal percentage explains only about 6% of the variation in scoring.

The regression model is:

\[ \hat{y} = b_0 + b_1x \]

For this dataset:

\[ \widehat{PPG} = 2.375 + 14.603(FGPercent) \]

where:

• \(b_0 = 2.375\) is the intercept
• \(b_1 = 14.603\) is the slope

The coefficient of determination is:

\[ R^2 = 0.06054 \]

or about:

\[ 6.1\% \]

This means field goal percentage explains about 6.1% of the variation in points per game.

• Field goal percentage has a positive relationship with points per game.
• The regression slope was positive and statistically significant.
• The model had an \(R^2\) of about 0.061.
• Field goal percentage explains only about 6.1% of the variation in points per game.
• Therefore, field goal percentage alone is not a strong predictor of NBA scoring.

Data: 2024-2025 NBA Player Per Game Statistics
Source: Basketball-Reference

• The analysis uses player-level season statistics.
• Traded players were combined into one observation.
• Players with missing field goal percentage or points per game values were removed.
• This analysis highlights association, not causation.