Editor’s note: Please issue a warm welcome to John Kearns — William and Mary Class of 2020, like Cameron and I. We’re incredibly excited to have John here to write about analytics and advanced stats. So, without further ado, here is John’s first post from his new series, Data From Deep.
March 13, 2019, Samantha Huge shocked the world (or, at the very least, the small yet fervent Tribe sports community) by firing long-time men’s basketball coach Tony Shaver. Shaver was the most successful leader in Tribe basketball history. His teams were known for their high-intensity, fast-paced offense. But a relative weakness on defense, especially in the last few years before his departure, plagued Shaver’s teams, ultimately preventing him from converting any of his four CAA championship game appearances into a NCAA Tournament berth. First-time head coach Dane Fischer was brought in from George Mason to right the ship, charged by Huge to create a more balanced team capable of finally making it to the Big Dance.
Having graduated William and Mary this May with an economics degree, I immediately want to look at the numbers to evaluate the job Fischer has done. This article will be the first of three going on a deeper dive on the statistics behind the Tribe’s historical 2019-2020 season and what that means for next year’s team.
Looking at the box score stats, the impact of Fischer’s new scheme was obvious. The Tribe only allowed an average of 69 points per game scored by opponents, down almost eight points from the year before and the lowest for the team since the 2014-15 season. The average points scored also decreased three points, from 74.5 to 71.5 points per game. The job Coach Fischer did in balancing defensive strength with offensive production was admirable. Though Fischer ran some different options in his motion offense than Shaver, the general approach to the game was the same. When the Tribe offense worked well, ball movement was fluid and scoring was generated from multiple positions.
Under the surface, however, some differences in philosophy appear between Fischer and Shaver. While team steals and blocks declined significantly from Shaver to Fischer, rebounds increased significantly. Marked improvements in defensive rebounding explain most of the change; while offensive rebounding only increased by .06 rebounds per game, defensive rebounding increased by almost two rebounds per game. The Tribe went from the 222nd-ranked defensive rebounding team in the nation to 36th the following year.
Some analysts have found that defensive rebounds have a significant influence on the success of a team, since it limits the opportunities for the opposing team to score. In addition, a team’s level of defensive rebounds is consistent across seasons, meaning that it is connected directly to the skill of individual players and the scheme a coach has implemented. Blocks, meanwhile, seem to be uncorrelated with team success and the consistency of steals across seasons is very low. Indeed, a high number of blocks necessarily means the opposing team is able to launch a high number of shots. A team defensive scheme focusing on risky plays like steals and blocks, which Shaver encouraged in his time at W&M, may therefore not be optimized to give a team its best chance to win.
Team assists also declined from 17.1 per game to 13.6. The change in the number of assists between seasons (-3.5) is the largest of any singular box score statistic. The drastic shift is could be due to having five players leave unceremoniously, or it could be Fischer implementing a new scheme. On the first point, the five players that left the team (Pierce, Audige, Rowley, Milon, and Owens) accounted for 10.1 assists per game. Their replacements (Barnes, Van Vliet, Hamilton, Hermanovskis, and Ayesa) combined for just 7.1. For as good a job as Fischer did recruiting players, Hermanovskis and Ayesa especially cannot compare to the overall offensive contribution of their predecessors.
However, scheme has something to do with it. Nathan Knight saw his assists per game decline by 1.7, indicating his usage had changed within the offense. The numbers seem to show that Fischer implemented an offense less focused on quick passes and scoring and more focused on solid team defense and individual scoring opportunities.
This new scheme meant big changes for our star player, Nathan Knight. For a team with a player of his caliber — an all-defensive, all-conference, all-nation star — William and Mary should have been dominant in the paint. It is worth noting that Knight essentially doubled his double-double tally from the previous season and finished with the second most double-doubles of all players in D-I basketball. This season, while Knight attempted fewer two-point shots, he almost tripled his three-point attempts, nearly equaling a career -high season-3P% of 30.6 percent on a much higher volume. Knight’s volume of assists also fell by half. Fischer implemented a system that took the ball out of Nathan’s hands, putting him in positions to score both in the paint and from the perimeter, which gave space for his teammates to work around him.
Stay tuned for part 2 next week — taking a closer look at the distribution of assists for the 2018-19 and 2019-20 Tribe men’s teams and what that meant for their success.