The Distribution of the Virginia Cavaliers' Precious Possessions

It's difficult to discuss Tony Bennett's basketball system without encountering rampant pacism.¹ Mostly re-treaded pacist jokes return to the apparently hilarious theme that slow games are boring. But with the boredom of some comes the success of the Bennetts.² Most Virginia fans have come to embrace the pace; it keeps the score close, emphasizing the Hoos frequent defensive advantage. Judging from the Twitter reactions during the big Wisconsin win, the final defensive possession was the highlight of the night³, even among a Mike Scott-like offensive performance from Joe Harris.

With any successful system or strategy there are inherent trade-offs. When the Hoos play at a rate of 60.3 possessions per game, compared to an NCAA average of about 67 possessions⁴, lost scoring opportunities and wasted possessions loom large on both ends of the floor. Bennett's slow pace places a premium on efficiently allocating offensive responsibility.

Last season, "maximizing efficiency in possession usage" meant "give the ball to Mike Scott." The departure of Scott, absence of Malcolm Brogdon, and limited availability of Jontel Evans have created a situation where a host of talented freshmen are forced to assume disproportionate offensive responsibility. In this post, I'll examine the variation in possession usage between the freshmen and other players as the season has progressed. The post concludes with a brief look at how the offensive responsibilities should change in future games.

[Measuring Possession Usage]

The Possession Percentage statistic estimates the percentage of offensive possessions that an individual player "used" while he was on the floor. "Usage" can take many forms. A turnover ends a possession, and the offending player⁵ is considered to have "used" that possession by preventing any scoring opportunities. A made shot also concludes, and therefore uses, a possession. The metric gets a little tricky when determining usage based on missed shots, since not all are rebounded by the opposing team,⁶ but still provides a serviceable and interesting measurement. The following chart shows the possession percentage for each Virginia player in each of the first eight games of the season.

[Total Possession Percentage Comparison]

We'll first take a look at total possession percentage⁷ used by freshmen versus other players. A graph of this appears below.

To examine the distribution of offensive possessions, let's split the games into blowouts (Seattle-Lamar-North Texas) and close games (Mason-Fairfield-Delaware-Wisconsin), leaving the Green Bay game out in its own separate category.⁸ As expected, the freshmen used their three highest percentages of possessions in the blowouts. The close games offer a more interesting comparison; in these, the total freshmen possession percentage has steadily declined.

It's important to note, however, that this doesn't necessarily mean Bennett is losing faith in our talented first years. The slow and steady decline of freshmen possession percentage in close games might simply suggest that the class is learning to incorporate themselves into the offensive system without over-stepping their intended roles. I would expect the freshmen percentage in close games to remain roughly at the level of Wisconsin before increasing later in the season.

[Individual Possession Percentage Comparison]

I also compared possession percentage changes among individual freshmen. Just for good measure, I threw in Joe Harris' percentage as a sort of "offensive gold standard" in the graph below.

Mike Tobey and Justin Anderson have the two greatest variations in possession percentage. Both had their lowest percentage against the most difficult opponent of the year⁹, while Tobey, probably unintentionally, actually rivaled Harris in the close games against Fairfield and Delaware. Nolte's consistency could suggest that he has the best incorporated himself into the flow of the offensive gameplan.

[Proposed Changes in Possession Percentage]

The following chart¹⁰ compares the overall possession percentage for Virginia players to their overall Offensive Rating. Since the ORating measures offensive efficiency, this chart can be used to suggest which players should use a greater percentage of their on-floor possessions, and which should look to pass to improve the offense.

Akil Mitchell and Harris are clearly the stars of the offense.¹¹ The most glaring aspect of this chart, though, is Evan Nolte's second ranked offensive rating. Perhaps it's time for Bennett to give Nolte a little more leeway to look for his own shot.¹²Paul Jesperson has been decidedly adequate. Tobey has been the biggest possession usage offender, and could have trouble earning minutes and offensive responsibility as the cupcake games wind down, while Jontel Evans' low rating and relatively high possession percentage will likely improve as he continues to acclimate to the pace of the game.¹³

Bennett's precious possessions have been efficiently distributed in the current five game winning streak. It will be interesting to see how they break down on Wednesday in the big game vs. Tennessee.

_______________________________________________

1 post-Wisconsin, Hoos ranked 345 / 347 in Ken Pom Adjusted Tempo (60.3 possessions per game)

2 Tony and his father Dick, the original Pack Liner, have combined for 9 NCAA Tournament trips while leading Wisconsin-Green Bay (trips in 3/9 seasons), Wisconsin (3/6), Washington State (2/6), and the Hoos (1/3)

3 Even Ben Brust's desperate attempt to draw a foul by launching himself into Joe at the end of the possession couldn't save Wisconsin's offense. After one game I can imagine how Brust could get under the skin of opposing teams and fans. For me, it started with his first wild attempt to draw a foul, when he collapsed under the rim and threw up about a .005% shot that somehow went in. His subsequent demonstrations of how to drill open threes didn't help either.

4 in case you were wondering, the fastest pace belongs to Seattle at 77.8 possessions per game. They are 2.7 possessions faster than the second fastest team.

5 tried to work "turnovor" and "turnovee" into this sentence, but it just didn't sound right

6 the possession might continue, hence the description of the metric as an "estimate" of possession percentage

7 I decided to use the sum instead of the average because there were more 2nd to 4th years with 0% possession percentages

8 since it was close at the half, but a big second half lead resulted in a 14 point win.

9 while Harris absolutely carried the team

10 sorted by Offensive Rating

11 this just underscores the fact that Mitchell's growth this season has been incredible

12 we need to start a Twitter hashtag #FeedEvanNolte (as in, for the open 3)

13 this could also be a function of the fact that he doesn't shoot often, but is entrusted with handling the ball; turnovers are a disproportionate number of his used possessions