Virginia basketball and "The Nolte Effect"

Earlier this month, I took a look at expert predictions for Virginia basketball's 2015-2016 season.  The various media projections had a couple common threads: They all had the Hoos in their "too early" top tens, they all noted the impact of the loss of Justin Anderson, and, among all the projections, each of the rising seniors were mentioned but one...none gave a thought to fourth-year Evan Nolte.

Who can blame them? Last season, Nolte's statistics were pedestrian, to say the least.  Though he played 43% of minutes (which placed him 7th on the team), he used just 11% of possessions when in the game, the lowest rate on the roster.  When he did shoot, the results weren't pretty; though he made 61% of his 2s (on just 23 shots), his three-point shooting declined for the second consecutive year, to 28%. He got to the line less often than everyone on the team, and shot 61% when he did.

Numbers-wise, he wasn't very good.  That's what makes this analysis from the esteemed Luke Winn so wild:

Nolte is one of those odd cases of a player who has subpar individual stats—an offensive rating of just 99.6, a three-point percentage of just 27.6—yet seems to make his team better when he's on the floor. According to data from HoopLens.com, in ACC games and the postseason (a 1,307-possession sample), the Cavaliers were +0.24 PPP with Nolte in the lineup and just +0.08 PPP without him. That was the biggest positive gap of any Virginia rotation player.

Basically, no player made a bigger positive difference while in the lineup than Evan Nolte. Winn's chart, linked above, shows that when Nolte is in the lineup, the Hoos shoot 2s and 3s better (by 4% and 5%, respectively) and better defend against both as well.  With him in the lineup, UVA was an off-the-charts +.24 PPP; without him, a solid but unremarkable +.08 PPP.  And the sample size of over 1,000 possessions gives us no reason to doubt these numbers.

What does this startling revelation mean?  First of all, maybe Tony Bennett has a method to his madness. And second, maybe fans could be nicer to poor Evan.  But seriously, what is happening here?  How does using 20% of your lineup on someone whose individual contributions seem so lackluster have such unambiguously positive effects on the team?  Mostly, I am just asking you guys for your thoughts, because I am stumped; but here are some ideas on the causes of "The Nolte Effect":

• Defense: The improvement in UVA's defensive efficiency is easier to understand. Nolte now has lots of experience in the Packline system. He's tall, moves his feet well, and could defend multiple positions - Coach Bennett has consistently applauded his defensive smarts.  Finally, his foul rate declined last year to a more manageable 3.1 committed per 40 minutes. If you need something a bit more tangible, he's also done stuff like this:
• Spacing: Much of Nolte's playing time comes at the 4 spot (though he played the 3 more often in Anderson's absence). Though his three-point shooting was poor last season, he was still a threat to attempt the shot all season, especially relative to, say, Anthony Gill.  Though the shots weren't falling, the threat of a three-point attempt spreads out the defense, opening things up for drivers and entries inside.  Essentially, the effort expended by the defense to make Nolte's numbers so miserable is opening things up to make the rest of the Hoos look better.
• Matchups: UVA doesn't play in a vacuum. When the Hoos substitute out a big man and put in Evan Nolte, the opponent could choose to adjust its lineup as well to "go small," a style that may be outside its comfort zone
• Statistical phenomena: Despite the numbers from Luke Winn appearing fairly definitive, there may be some "good luck" at work, or the possibility that Nolte's presence correlated with more efficiency, but didn't cause it. My first thought was that Nolte's playing time increased late in the season, when Justin Anderson injured his hand; however, UVA's net efficiency margin was down during this time, so we can't argue that Nolte just happened to play more during a time of increased team efficiency.  Maybe Nolte happened to play more during blowouts, inflating his efficiency numbers? He averaged 16 minutes per game during the team's 3 most jarring ACC blowouts (vs. Clemson, vs. GT, @ Wake), which is actually below his season average of 17 minutes, so no dice there either. There isn't a nice and easy way to explain this one away, but there's a ton of other, non-Nolte things that Nolte's playing time may correlate with. (The parts of the game he plays? The other players who tend to play alongside him? The fact that someone who harms efficiency may not be playing?)  And Nolte's relatively low playing time could make him more prone to this type of "efficiency gap" than those who play big minutes.