2014 University of Michigan Basketball Analysis (SQL / Power BI)

Background and Purpose

In 2014, the University of Michigan Men’s basketball team won the regular season Big Ten Championship (outright) since the 1985-86 season! They also had the most conference wins (15) since the 1992-93 season. As a lifelong Michigan fan being born in 1992, I had never witnessed a Big Ten championship season, so naturally I was very excited. That excitement led me to take a closer look at some of the season statistics to try and get a better appreciation of the team performance.

I decided to analyze the shooting performance (field goals, three pointers, and free throws) as that is usually a good performance indicator for the overall team performance. There are, of course, so many other factors that impact the results (defense, turnovers, injuries, etc.) but for the purposes of this study I wanted to only focus on shooting statistics.

Data Collection

The data I used was accessed via a public database in Google Big Query. The table I pulled data from consists of team-level box scores from every men’s basketball game from 2013-2018, provided by Sportradar LLC.

Dataset ID: bigquery-piblic-data.ncaa_basketball

Table ID: bigquery-piblic-data.ncaa_basketball.mbb_games_sr

I wrote a SQL query to filter the table to include only the data I was interested in – 2014 Big Ten conference games played by Michigan. I’ve included the query in its entirety and a screenshot of the results below. The resulting table included the data, teams involved, points, whether Michigan won or loss, and shooting statistics.

Data Visualization

I saved the resulting table from my SQL query as a .csv and loaded the file into Microsoft Power BI. At this point, I viewed the data in Power Query to make sure it was clean and formatted the way I wanted. I also wrote a few DAX expressions to aggregate shooting stats over the course of the season.

I created a dashboard that lets the user filter based on opponent, whether Michigan was home or away, and whether they won or lost. It also displays the season average shooting percentages at the top for reference. There are scatter plots for each of average field goal percentage, three-point percentage, and free throw percentage by opponent, that also clearly shows the user the number of attempts (via marker size) and the result of the game (via color). Also included in these plots is the target line, for field goals and three pointers, and an average line for free throws. At the bottom of the dashboard is a table which shows the team’s overall record.

Key Insight

The average field goal and three-point percentage over the course of the season was 48.94% and 40.64% respectively, and the average free throw percentage was 76.84%. Michigan outperformed all other Big Ten teams in these categories.

During my analysis, I found that Michigan won every game in which their field goal percentage averaged over 43% except for one game. I found a similar target for three-point percentage; however, this was less clear. Michigan won every game in which they shot over 38% from the three-point line, but also won a few when they shot below that mark. Free throw percentage did not seem to have an obvious impact on performance. Michigan lost games in which their free throw percentage was higher and won games when the percentage was lower. The average per game did seem to be lower in games where the attempts were higher. The clear takeaway from this analysis is that field goal percentage has a great impact on overall performance, and when the 2014 Wolverines shot over 43% on any given day, they were heavily favored to win the game.