Part 1 of The Fischer Effect can be found here.
In the first entry of the Data from Deep series, I identified the differing approaches between former coach Tony Shaver and current coach Dane Fischer. Fischer’s philosophy stressed defensive cohesion and a slower, more methodical offense relative to Shaver’s teams, as evidenced by an extreme decline in team assists. Just as the names on the team sheet changed around Nathan Knight, Knight’s role changed within the team. How did this new scheme and Knight’s adapted role affect overall team play? Did Fischer create a more balanced team whose foundations are built for future success even without Knight?
As a spectator, it seemed to me that the offense would go through cold stretches marked by passing along the perimeter until an expiring shot clock forced a bad shot. What was missing from the previous year’s team was that incisive pass cutting through the heart of the opposing defense. Figures 1 and 2 show the assist network for the 2018-19 and 2019-20 seasons, respectively. The assist network maps show the number of assists for each player and who they assisted on each pass.
First, looking at Nathan Knight, I noticed that his involvement in assisted plays drastically declined. This is likely because of his lack of credited assists — the red-shaded assists become much less numerous. Knight went from recording 23 percent of the team’s assists to only 13 percent. However, the share of Knight’s made baskets that were assisted by another player (what I will define as “converted assists”) increased to 24 percent.
While Knight was dominant in the 2019-20 season, one could have hoped that the players around him could have provided more scoring production to add other offensive options. The network map and assist frequency statistics succinctly show how Fischer changed Knight’s role by taking the ball out of his hands – or, in other words, playing less through Knight as a facilitator and more as the final destination.
The addition of Bryce Barnes was incredibly crucial to the success of the Tribe offense. As a transfer, he led the team in assists with 3.4 per game, accounting for more than 25 percent of the team’s assists. Another surprising contributor was Thornton Scott, who accounted for 18 percent of assists despite being injured or coming off the bench in the majority of games this season.
While Barnes was an effective distributor of the ball, he failed to make a large impact in converting other people’s passes into made baskets. Barnes only accounted for 5.8 percent of converted assists, behind Quinn Blair (7%) and Miguel Ayesa (10%). Scott accounted for even less production in that fashion, converting only 4.5 percent of the team’s assists. These two players, Barnes and Scott, were both simultaneously the best passers on the Tribe team and some of the most unlikely players to link up with their teammates for a basket, albeit for different reasons: Scott was more likely to shoot off the dribble, while Barnes was just more hesitant to pull the trigger.
As shown in Table 1, the two best passers as measured by assists in the 2018-19 season were Knight and Justin Pierce. Unlike Barnes and Scott, these two players were also in the top three for converted assists as well. Knight and Pierce being able to distribute the ball, run the offense, and score within the flow of Shaver’s patented “turbocharged Princeton” brought a greater versatility to the Tribe, since scoring production was not just limited to individual shot creation on the perimeter, to which the 2019-20 offense was often forced. Such outcomes are seen by the relative importance of Ayesa in the process of converting assists to made baskets. Ayesa provided much needed accuracy from beyond the arc and was used primarily as a catch-and-shoot player. If the Tribe offense was often camped out on the outside, it would be expected that Ayesa would have a significant share of converted assists. Ayesa finished the season with the fourth most converted assists on the team, the vast majority from the three-point line.
A more rigorous analysis of a player’s offensive impact is with authority and hub scores. An authority score seeks to summarize the value of the relative variety of assisters a shooter has and the volume of converted assists he makes. A more “authoritative” player, one who is assisted a lot by many of his teammates, will score closer to 1 than 0. Given two players with the same number of converted shots, the player with more made three-pointers will score higher. The authority score can grant insight into which players were the most dominant offensively.
A hub score, inversely, seeks to summarize the value of the relative variety of shooters a passer assists and the volume of assists. A bad “hub” player will similarly have a low volume of assists to only one teammate, scoring closer to 0 than 1. Given two players with the same number of assists, the player with more three-pointers assisted will score higher. The hub score can help identify the most effective distributors of the ball.
Surprisingly, Andy Van Vliet was the most “authoritative” player this season, most likely due to his higher volume of three-pointers weighting his point production higher. Looking at the ratio of authority score to shot frequency, the difference between Van Vliet’s ratio (2.81) and Knight’s (2.48) indicates that, controlling for shot volume, Van Vliet converted in a more varied way from a more varied subset of players. This ratio or “multiplier” identifies added value through offensive versatility and efficiency. Knight is surely the most important player on the team, but Van Vliet’s multiplier of 2.81 shows that he impacted the game more with the assisted passes that came his way. Scott also has a high authority score relative to his shot frequency, also indicating his importance to the Tribe offense despite his injuries.
Looking at hub scores, Barnes was the most effective passer for the Tribe over the past two seasons. Scott also scored very well on this metric. One surprise is that Tyler Hamilton passes Knight and Van Vliet on the hub score rankings despite racking up less assists than either of them. Despite not having significant statistics in traditional box score categories, Hamilton was an efficient player whose solid defense and quiet distribution of the ball gave the Tribe offense more versatility. Despite not assisting often, Hamilton assisted many different players, which provided creativity to the offense.
The relative rarity of this creativity, as evidenced by low hub scores across the team, echoes the problems the 2019-20 Tribe suffered all season long. Fischer’s best distributors lacked the offensive killer instinct Shaver had instilled in his players. Likewise, Fischer’s best scorers did not distribute the ball nearly as well as they had in 2018-19.
When the Fischer offense was working, the team had no trouble scoring; the 2019-20 Tribe scored 80 or more points seven times last season, only two games less than the 2018-19 team. However, in a crucial five-game stretch that ultimately decided the CAA regular season title, the Tribe failed to score even 65 points in any game. The team also mustered only 63 points in their frustrating CAA Tournament quarterfinals exit.
It is worth reiterating that Fischer did an amazing, coach-of-the-year-worthy job guiding this team to the historic success it had. But it remains that when the Tribe was stuck in an offensive slump, it lacked the creativity and adaptability that made the Shaver teams so consistently successful scoring the ball.
Stay tuned for the final post of “The Fischer Effect,” as John looks at the stats and projects the Tribe’s offensive performance in 2020-21.