How May We Assist You Today?

Welcome to the second full day of the NCAA Tournament — and we sincerely hope your bracket isn’t busted yet. Yesterday’s analysis of Davidson v. Kentucky that focused on the three ball fell a bit short on our estimate due Kentucky’s 2nd half slow down as they protected their lead — maddening.

Our focus today will be on teamwork — namely assists — how they are related to field goals made, and how teams go about spreading the wealth. Everybody likes touching the rock, right?

We picked the (14) Bucknell vs (3) Michigan State matchup for our case study — you’ll see why.

Background

Michigan State possesses one of the most effective offenses in the country. They rank #1 in assists-per-game (19.3), are hitting 41.3% from 3-point range (#4 in the NCAA), shoot 50.4% overall as a team (#5 in NCAA) and average 1.154 points-per-possession (#6 in NCAA).

Bucknell is a prolific scoring team in their own right, averaging 81.1 points per game (#27 in NCAA). They rank #7 in the NCAA in free throw attempts per game (24.9) and #12 in free throws made per game (17.8). They shoot 47.2% as a team (#41 in NCAA) and their 2-point efficiency is a solid 55.6% which ranks #23 in the NCAA (out of 351 teams).

With two quality offensive teams playing one another, it’s important to look at team balance: how the baskets are created for and made by each squad.

Analysis

For the exercise, we looked at the player contribution per team, per game, for field goals and assists. The goal was to see if we could quantify how much each player contributes to the team’s overall offensive performance.

Using the `bigquery-public-data.ncaa_basketball.mbb_players_games_sr` and `bigquery-public-data.ncaa_basketball.mbb_teams_games_sr` tables from the BigQuery public dataset we were able to create a view that joins box score and player aggregates.

This view requires that each player had at least 1 assist in the game. From there, we can create some nice violin graphs to look at the density and distribution for each team.

The first graph looks at the percentage of field goals made by a player per game.

You can see that Michigan State has more players that contribute to their made field goals. And Michigan State’s top two player clusters have a lower field goal contribution than Bucknell’s, which reinforces the idea that they are more balanced. (To Bucknell’s credit, they are also well-balanced.)

Sidenote: What’s even more amazing about Michigan State is that all of their starters are averaging 11 or more points per game, so not only are they pretty balanced but they are also scoring lots of points.

Next, let’s take a look at how these field goals are created — specifically, how are they being assisted (or not).

This graph looks at a player’s assists relative to the team’s assists. From this, we can interpret that both teams are pretty balanced. Michigan State’s top cluster produces, on average, 40% of the team’s assists.

Contribution balance is all about getting as many players as possible to make lots of points AND create lots of points for their teammates. Pass the rock and score, pass the rock and score…

With this in mind — here is our prediction for Bucknell vs. Michigan State: We expect to see a total of 29 assists combined.

Pass it on.