Question

I wanted to look at NBA shooting stats.

The question is:

Do teams with better shooting percentages usually win more games?

This uses 2023-24 NBA regular season team data.

Data

The data is from Basketball-Reference team stats and standings.

I used three variables:

• wins
• field goal percentage
• three-point percentage

##            team wins   fg three     pred     resid
## 1        Boston   64 48.7  38.8 47.90252 16.097477
## 2        Denver   57 49.6  37.4 52.79408  4.205925
## 3 Oklahoma City   57 49.9  38.9 54.42459  2.575407
## 4     Minnesota   56 48.5  38.7 46.81551  9.184488
## 5   LA Clippers   51 48.9  38.1 48.98953  2.010465
## 6        Dallas   50 48.1  36.9 44.64149  5.358511

Correlation

Correlation is used to measure how two variables move together.

\[ r = \frac{\sum (x_i - \bar{x})(y_i - \bar{y})}{\sqrt{\sum (x_i - \bar{x})^2\sum (y_i - \bar{y})^2}} \]

For this project, \(x\) is field goal percentage and \(y\) is wins.

Plot 1

## `geom_smooth()` using formula = 'y ~ x'

Linear Regression

I also used a simple linear regression model.

\[ \widehat{wins} = b_0 + b_1(fg) \]

This means the model predicts wins from field goal percentage.

Model Result

## (Intercept)          fg 
## -216.784792    5.435058

The slope is positive, so higher field goal percentage is connected with more wins in this data.

This does not mean shooting is the only reason teams win.

Residuals

A residual is how far off the prediction was.

\[ e_i = y_i - \widehat{y_i} \]

A positive residual means the team won more games than the model predicted.

Plot 2

Plotly Plot

## `geom_smooth()` using formula = 'y ~ x'

R Code Example

ggplot(nba, aes(fg, wins)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE) +
  labs(title = "Field Goal Percentage and Wins",
       x = "Field Goal %",
       y = "Wins")

Conclusion

The teams with better field goal percentage usually had more wins.

There are also some teams that do not fit perfectly. That makes sense because wins also depend on defense, injuries, turnovers, coaching, and schedule.