Offseason Checkup: Ball Screen Offense

Dylan Burkhardt

The ball screen has become one of our most frequent topics here at UM Hoops.

Six or seven years ago when John Beilein arrived in Ann Arbor, the high ball screen was barely on the radar. Michigan ran Beilein’s classic two guard offense and quite frankly didn’t have the dynamic playmakers required to run an offense that featured the pick and roll.

Darius Morris, Trey Burke, Tim Hardaway Jr., Nik Stauskas and LaVall Jordan changed all of that. Michigan had the talent to run the pick and roll and had an assistant coach that specialized in teaching the high ball screen game. Fast forward a couple of years and the Wolverines had the country’s best offense in consecutive seasons while featuring as many high ball screens as any team in the country.

While Beilein has always been respected as an offensive mastermind, the addition of the pick and roll game added another element to Michigan’s already talented offense. Last year Michigan’s offense hit a road block in almost every facet, including the pick and roll. The ball screen stats from last season aren’t pretty, but in many ways they might also be reassuring.

Michigan’s offense scored .911 points per possession on pick and rolls last season according to Synergy Sports – its worst pick and roll output in five years. The next graph shows Michigan’s ball screen offense over the years with the size of the bubble representing ball screen volume.


While it was disappointing to see Michigan’s offense regress – a phenomenon that was not limited just to the ball screen – we can also see that it didn’t regress that much. Michigan still ran a high volume of ball screens and was ranked in the 83rd percentile nationally.

The encouraging element of Michigan’s perceived ball screen struggles is the fact that it wasn’t the usual suspects who provided the production. Michigan’s go-to pick and roll players entering the season were Caris LeVert and Derrick Walton – both of whom regressed in limited action – but it was Spike Albrecht, Zak Irvin and Muhammad-Ali Abdur-Rahkman that provided the bulk of Michigan’s productive ball screens.

Don’t believe me? Here’s the visual proof. The following graph shows Michigan’s current guards and their level of pick and roll production. Once again, bubble size represents volume (in this case raw number of pick and roll possessions).


The first thing that stands out is Caris LeVert and Derrick Walton’s regression. Both players used roughly the same number of possessions as they did the year before, but they were less efficient. (Given that they played roughly half as many games, that’s still almost double the volume with reduced efficiency.) LeVert and Walton both struggled to shoot the ball off the bounce and didn’t pass it as well as Nik Stauskas the previous season.

LeVert also continues to be a far better player in isolation situations than with a screen while Walton’s struggles finishing around the rim killed his production across the board.

Spike Albrecht was able to bounce back from a below-average pick and roll season in 2014 and eventually led Michigan in ball screen possessions. Nobody would have expected that heading into the season, but Albrecht has proven his offensive worth. He can run the offense, find people, knock down jumpers and even come up with a few crafty finishes at the rim.

Zak Irvin’s progression is most eye opening. He was a guy that basically hadn’t seen a ball screen as a freshman, but by the end of the year he emerged as a legitimate threat. We’ve seen a lot of impressive player development in recent years, but Irvin’s jump from spot-up shooter to legitimate creator was perhaps the most surprising. If he can continue to produce at that level and expand his game, he should be on track to surpass most expectations.

And then there’s the freshmen. Aubrey Dawkins finds himself in a similar spot as Zak Irvin after his freshman season while Muhammad-Ali Abdur-Rahkman was the most efficient pick and roll scorer for the Wolverines, notching 1.16 points per possession off of pick and rolls.

Finding the roll man

Point guards get all of the headlines in the pick and roll game, but a reliable big man can make the whole offense run more efficiently. The chemistry in the ball screen is critical and we saw some terrific examples of when everything works with Darius Morris to Jordan Morgan and Trey Burke to Mitch McGary.

Michigan’s 2014-15 offense lacked some of that punch at the rim that McGary and Morgan were able to provide. Here’s a look at Michigan’s top scoring roll men over the past few seasons.


The first thing to note is that the 2011, 2012, 2013 and 2014 (x2) teams all had at least one roll player with more scoring volume than last year’s group.

Even that statistic doesn’t tell the whole story. Max Bielfeldt was Michigan’s leading roll man last season, but of his 33 roll possessions 12 were via the pick and pop. He scored just 23 points on 21 possessions rolling or slipping to the basket.

Ricky Doyle shows the promise — he was nearly as efficient as Jordan Morgan was as a senior — but was unable to shoulder the volume. Conditioning slowed Doyle throughout the season whether it was during a game or battling a longer term illness. Doyle scored 69% of his chances as the roll man last season and clearly looks to be the long term answer as Michigan’s go-to screener. On the other hand, he also needs to continue to learn to set better screens and help his guards.

Mark Donnal was lethally efficient (14 points on 9 possessions rolling to the hoop) and despite his reputation was only credited with one shot off of the pick and pop. Zak Irvin’s name also pops up on the chart as he’s an effective pick and pop option against defenses that won’t switch at the four position.


But the more important question is what does this mean going forward? Michigan needs to improve on the offensive side of the ball next season and the two guards that it adds back to the mix are the ones that struggled in the pick and roll game offensively.

On paper, that seems like a negative. But the reality is that LeVert and Walton didn’t have much help over the first two months of the season. The majority of Zak Irvin’s ball screen production came in the last two months of the season and Muhammad-Ali Abdur-Rahkman had yet to crack the rotation.

Assuming they stay healthy, LeVert and Walton should still lead the team in pick and roll possessions this coming season. They should also be much more efficient players because of the threats they have around them. Go back and watch the tape of Michigan’s game at Arizona and watch the players off the ball.

There was very little movement in Michigan’s offense and LeVert was essentially trying to beat the Wildcats 1-on-5. Dawkins and Abdur-Rahkman combined for 16 minutes, three missed shots and two turnovers in that loss. So even while Irvin or Abdur-Rahkman won’t be called upon to run every ball screen, Michigan’s offense should finally rediscover some balance with a healthy roster of experienced players.

The additional options of DJ Wilson and Moritz Wagner as roll/pop hybrids and Duncan Robinson as a shooter or even pick and pop player should only continue to help, but it remains to be seen if Caris LeVert and Ricky Doyle can develop the kind of connection we’ve seen in the past.

  • gobluemd16

    This article is absolutely superb. Ball screen offense is obviously now a staple under Beilein. Hopefully Caris and Derrick (and Zak) improve in this regard with so much more help around them!

  • UMHoopsFan

    Nice write up. There’s a typo,I think: the first Spike Albrecht in the second paragraph under the first graph should be Walton.

  • mikey_mac

    Very creative data visualizations, and lots of great insight, Dylan.
    MAAR’s PNR effectiveness plus his defensive reputation will make it hard to keep him out of the rotation. It’s interesting that UM’s expected most-prolific PNR combo for next season (LeVert/Doyle) has so little experience playing together.

  • Juniorhemingwayjunior

    Great article.

  • eddieben

    One thing that I feel is often overlooked in the ball screen game, is the actual screen. Some are better at it than others. McGary set monster screens, Donnal and Doyle have set rather weak screens. As the young bigs mature, I suspect that this is one element of their games that will improve drastically and will really help the ball handler and his ability to make the correct read immediately post screen.

    • Definitely important. Morgan set very good screens too.

      • Yes he did, but that thing that made Morgan real special on the ball screen had to do with the fact that he never gave up on the roll, He’d patiently take it all the way to the rim regardless of how the play looked to be developing. Then he’d finish when the ball finally showed up.

    • Corperryale

      Good point. Donnal especially seems to set weak, non-committal, poorly timed and poorly angled screens. I think the quality of Doyle’s screens improved over the season and expect he will turn into a very good screener, especially with that big body and improved conditioning.

    • Chris De Sana

      Need to keep in mind that the point guard has a lot to do with the quality of screen as well; if gets lazy and makes too big a loop the defender will room between he and the screener. The point guard or handler also has to make sure to wait long enough for the pick setter to be established which can also cause problems and fouls for moving picks.
      More experience by all plus playing more minutes together will be helpful.

  • UMTigersWingsLions

    Cant wait to see Duncan light it up. We have by far the best shooters in college hoops. Caris, AD, Duncan, Zak, DW, Spike