With NCAA Basketball conference tournaments starting this week and the NCAA tournament starting next, I wanted to look at team stats this season to see how they relate to success. This brought me to a major part of the game: free throws.
Free throws are an important part of basketball. Obviously when a team is shooting free throws it means they have an extra opportunity to score and it hurts the other team in terms of points, fouls, and momentum. But are free throws as important as we think to a team’s overall success?
Note that ll data in this post is as of 3/9/2020, and does not include any conference tournament games.
The first graphic looks at the top 30 teams in terms of free throw attempts per game. The color of the bar represents their win percentage.
Auburn attempts 26.39 free throws a game, which is almost 1 whole free throw more than the second most team, Stephen F. Austin. It’s clear from the graphic there is a mixture of teams with a high win percent (orange) and low win percent (blue).
The second graphic compares a team’s free throw attempts per game to their win percent, essentially the same as above, but this time it’s all teams. Standard deviation bands are added for both measures as well as a linear trend line.
A way to think about the standard deviation bands are teams outside of them are outliers.
There is a positive correlation between free throw attempts per game and win percent, but it’s pretty weak. However, San Diego State has about 16 free throw attempts per game and have a high win percentage and Vanderbilt attempts about 23 free throws per game and only has a win percentage of 35.5%. Clearly more free throw attempts does not make a better team.
The third graphic compares a team’s free throws per field goal attempts to win percentage. This gives us a better idea of how a teams free throws make up their offense.
There is a positive correlation here too, while not incredibly strong, free throws per field goal attempts is statistically more significant than free throw attempts per game.
We see more of the better teams within the standard deviation bands. Teams like San Diego State, Dayton, Gonzaga, and Kansas all have a very average free throws per field goal attempt rate.
The fourth graphic compares at a team’s free throw percentage to win percentage.
Team’s free throw percent is the best predictor of their win percent we have looked at so far, but it still does not have a strong correlation. Clearly some of the better teams have a high free throw percentage, but there are definitely exceptions.
Kansas for example is outside of the standard deviation band with a 66.7% free throw percentage, yet they are currently ranked #1 in the country. Gonzaga is currently the second best team in the country, but they have a free throw percentage of 68.7%, which is also below the average.
So are free throws that important for a team to do well? Let’s compare another (albeit more common) measure to win percentage – field goal percentage.
The difference is clear. Teams are bunched around the trend line much more cleanly than any of the four previous free throw measures. The relationship between win percentage and field goal percentage is statistically significant and the correlation is much stronger.
What’s interesting are the ends of the line. Teams with well above and well below average field goal percentage have higher and lower win percentages, respectively. Teams in the bands are relatively closer to the trend line. This suggests that while field goal percentage is important on average, the best teams have other skills that make them great.
So are free throws important? While they may not be as important as one could suggest, I think it’s hard to measure any timely free throws.
Say a team is down four with 15 seconds to go and they make a free throw to bring them within three. That is clearly worth more to a team in that moment due to the emotions and momentum they may have after that. So, while free throws may not directly tie to a teams success, they definitely can play an important role.
The Commute Sports 
Pingback: Predicting College Basketball Games (Part 1) – The Commute Sports