Track of the Week: On + Off

I’m back for another track of the week. “On + Off” is a new single by NYU’s Maggie Rogers.

Hailey Nuthals of the Washington Square News—NYU’s independent school newspaper—wrote an excellent article on Rogers and her newest single that will appear on her upcoming debut EP “Now That The Light Is Fading.” The release is on Feb. 17.

Given the title of this single, we’re going to look at some on/off numbers for the Wolves.

On vs. off data is useful in assessing the impact a player has on the team's overall performance. For instance, the Wolves are currently scoring 110.4 points per 100 possessions with third-year wing Andrew Wiggins on the floor. When he’s sitting on the bench, the team’s offensive rating (ORtg) dips to 104.5 points.

Thus, Wiggins is +5.9 when it comes to ORtg in his 1626 minutes this season.

On the other hand, opponents are scoring 110.9 points per 100 possessions with Wiggins on the court and 108.0 with him off. That means the team’s defensive rating is +2.9 points better without him.

Now for some basic math: 5.9 (boost in ORtg) - 2.9 (shift in DRtg) = 3.0!

When Wiggins is on the floor, the Wolves score 3.0 more points per 100 possessions relative to their opponents than when he's not on the floor. As you will see below, only four players on the team currently have a better on/off through 44 games.

From best to worst (data via Basketball Reference):

Of course, keep in mind there are multiple factors that influence on/off numbers including role, minutes played, when they play those minutes, skill level of the competition—do they only play against bench players in garbage time?—lineup combinations, and more.

The point is, while on/off statistics can help us understand how the team functions with certain players on the court, it’s only one statistic to glean information from with many variables at play. Tyus Jones clearly deserves more playing time based on his success in a small sample this season, but his team best +7.4 on/off also doesn’t mean he should be the starting point guard. It simply means he’s done an excellent job when called upon and the team is performing at a high level when he’s in the game. Thibodeau also deserves credit for putting him in the right situations to succeed.

Karl-Anthony Towns is +0.4 and he’s clearly the best player in Minneapolis, but Gorgui Dieng is the true on/off star on the roster at +6.5 in 1413 minutes.

The difference between the two starting bigs is their defense. The Wolves are +7.0 points better defensively per 100 possessions with Towns on the bench (112 DRtg with him, 105 without him). Dieng is +3.5 in ORtg and -2.9 in opponents ORtg, meaning teams score almost 3 points less per 100 possessions against lineups with G.

Zach LaVine’s on/off numbers are terrible, though his improvement this season is massive. So, what’s up? Well, to the surprise of very few people who follow the team closely, the defense is wildly better when he’s not on the floor. His -8.2 in 1524 minutes is an obvious red flag and most of that is coming on the defensive end where his struggles are well documented.

With LaVine in the game, opponents are scoring 113.1 points per 100. Without him, the Wolves allow only 103.4 points. No other player on the roster has such an alarming gap in opponents ORtg (+9.7). Now does his awful on/off numbers mean he’s a bad player? Of course not; every team in the league would kill to have an elite pace and space 21-year-old wing that fits perfectly into the modern era. But his on/off does suggest that the defense is at its worst when he’s involved and that should be a major concern moving forward. Towns is extremely guilty (as noted above) as well.

Here’s another alarming number if you’ve been following the Wolves for years: Shabazz Muhammad is -8.8 in opponent ORtg. Yes, that means the DEFENSE (you are in fact reading that word correctly) is almost 9 points better with Muhammad than without him. I never thought Mr. Buckets could be apart of lineups posting a 104.6 DRtg.

Thibs dust...could it be real?

While on/off numbers aren’t everything, plenty of the lineup debates in Wolvesdom over the course of this season can be validated at least partially by the numbers above. At the very least, it’s interesting to see the data and whether or not it matches what other statistics are telling us, and what our eyes and hearts want to believe.