Using Julius for college basketball analytics

I used the free demo version of Julius to answer a few questions about data from my college basketball team! Will share the results below.


I have a bunch of data from my college basketball team that shows the team’s scoring efficiencies, turnover percentage, points per possession, percent time, etc. for EACH game scenario. For those who may not be familiar, example scenarios are ball screen (when one offensive player uses a “screen” from another offensive player), spot up (catching and shooting the ball), and transition (shooting before all five defenders are back on defense). Different teams play different styles of basketball, and Julius was able to make me a pie chart showing the distributions for my old team, Whitman College.


I also wanted to look at individual player data. I picked shooting efficiency as the metric to visualize, which is a pretty broad metric with many interpretations. I wasn’t able to take a screenshot of my prompts for these, and I can’t go back now because I was using the free demo, but I pretty much asked Julius "Draw me a scatterplot showing individual player shooting efficiencies. It decided on its own to plot 2 point shooting percentage compared to 3-point shooting percentage for all players on the Whitman College roster. It was super nice not to have to specify which axes I wanted, and the result was pretty cool. Check out this screenshot of the scatter plot! I also asked Julius for a bar graph representing my own personal shooting efficiencies, just out of curiosity.

There are way more things I’d do with this data if I had more free questions to ask Julius. I would look for distributions of certain key metrics (normal, right-skew, left-skew, etc.) One thing I wasn’t able to test, but would like to, is Julius’ ability to detect strategic insights from basketball data on its own. For example, if my basketball team has a high turnover% in one particular scenario, Julius could identify this and then come up with strategies on how to avoid getting into this scenario in a game. Or even say, “I recommend Player X have the ball in this scenario because his individual turnover% is lowest in this scenario and there’s a large enough sample size to infer that he’s comfortable with the ball in this situation.” I’m going to play around more with Julius + college basketball data and might even post more screenshots later. Would love to hear people’s thoughts on this!