What's with Larry Sanders' awful offensive lineup splits?

By all accounts, Larry Sanders has been the best story out of Milwaukee this season. His improved rebounding and team defense has made him a crucial cog in the Bucks' top-10 defense, and give him a legitimate case as one of the top individual defenders in all the NBA. But it's his offensive improvement that has been most striking, as Sanders has transformed himself from a gangly, unskilled shot-blocker to an athletic rim-rocker. No longer do passes bounce right off his hands; these days, even balls high above the rim are subject to dunking. Sanders' dedication to staying near the rim and attacking the basket in the pick and roll has paid major dividends. The percentage of his shot attempts coming from outside 9 feet has dropped from over 25% last season to 20% this season.

All this has made Sanders a far more effective player on the offensive end which, combined with a slashed foul rate, has led to a twofold increase in minutes per game. That's great for Milwaukee's defense, which has been revitalized thanks in large part to Sanders and fellow Tube Man Ekpe Udoh, but it's having a surprising effect on Milwaukee's ability to score efficiently as a team. When Larry Sanders is on the court, the Bucks score 96.5 points per 100 possessions, just slightly better than the Wizards' 30th-ranked attack. With Sanders on the bench, their offensive rating spikes to 106.7, a top-10 mark. Basically, Larry Sanders presence on the court seems to be morphing the Bucks from a top-10 offense into an absolute disaster. What gives?

Is he stifling ball movement? At first glance, that seems unlikely: the percentage of Milwaukee's shots coming off assists is 4-5% higher when Sanders is on the court. Of course, Larry himself is being assisted on over 70% of his own makes, so maybe that number is swayed by his offensive style. While his 6.1 assist rate is tiny, his 15.6 usage rate doesn't paint the picture of an offensive black hole. Realistically, if Larry is touching the ball, it's likely a swing pass around the perimeter or a dump-off at the basket--part of the reason he's been so good this season is the decision-making process has been drastically simplified. It's not like he's getting the ball in the high post and keying possessions himself.

Is he throwing the ball away too much? Larry's turnover rate is high among centers with at least moderate playing time, and the Bucks do turn it over almost 2 additional times per 100 possessions when he's on the floor. Some of those could come in the pick and roll with Sanders, though the guards are probably just as much to blame in that sense. But while interior passes are more likely to be stolen or knocked away, they also lead to better shot attempts, so it's a stretch to think that trend would tank Milwaukee's efficiency. The turnover split is definitely worth paying attention to, but it doesn't really account for the full split.

Are the Bucks shooting better shots when he's on the bench? Hard to say. According to the NBA's advanced stats page, the Bucks shoot slightly better in the restricted area when Sanders is on the pine, but that seems like a red herring: Sanders is hitting almost 70% of his shots at the rim, and most of his attempts come from point-blank. Another odd wrinkle: with Sanders on the court, Milwaukee is shooting just 28% on above-the-break 3s with a considerable number of attempts. When Sanders sits, that percentage shoots all the way up to 39%. Strange stuff, and I can't claim to have a great explanation. Brandon Jennings and Mike Dunleavy are both hitting at least 37% of their above-the-break 3s, and Sanders might have fewer minutes than expected alongside those two players. The Bucks might also emphasize screening for three-point shooters more when Sanders isn't playing. The on/off split in this case is so severe that it seems likely to regress as the season wears on.

Do the Bucks play a different style without Sanders? Surprisingly enough, the Bucks play at a slightly higher pace when Sanders is on the bench, but it's just about a 1-point difference. It's very possible Sanders is contributing to that trend by lengthening opponent possessions with his shot-blocking. Most teams aren't great at grabbing a high percentage of shots blocked by their teammates (you may remember this being a particular strength with Andrew Bogut), resulting in slightly longer possessions for their opponents if the offense resets. Perhaps more notable is the foul-shooting split: Milwaukee takes and makes about 2 more free throws per 48 minutes when Sanders is on the bench versus on the court. That may not seem like much, but free throws are very high-value shots, so adding even a few can provide a big boost.

There's no doubt Larry Sanders has been one of Milwaukee's best players this season, but this surprising data trend is hard to overlook. Some of the more extreme splits will likely move back toward average as the season progresses, but for now it's a curious revelation, and something we'll likely revisit in the future.

Additional statistical support provided by Basketball-Reference and HoopData.