An irregular series about Michigan's 2019-20 basketball outlook.

ROSTER

Zavier Simpson (Sr.): Captain Hook pushed his ORTG up a few points while ticking upward in usage, went from 206th to 17th in assist rate. Went from 52% to 67% at the line as FT rate plummeted with hack-an-X no longer viable. Three point shooting static and extremely streaky. Defense was excellent as always.

David DeJulius (So.): high school skillset did not immediately translate, averaged about 3 MPG. Had the occasional impressive take but 5/15 from two and 1/15 from three on the year.

Eli Brooks (Jr.): promising start, horrendous middle, redemptive finish. Brooks was Michigan's only bench guard but a 94 ORTG on 13% usage in year two is a tough starting point to get to "good bench player" from.

I HAVE SOME QUESTIONS

What's the X project this offseason?

[Patrick Barron]

One of the joys of the recently completed season was Zavier Simpson, of all people, bringing back a 100% bonafide Kareem Abdul-Jabbar skyhook. It dropped at a ~60% clip despite steadily ramping up in degree of difficulty once opponents understood that it was a viable option. And every time it went down the announce crew reacted like they'd seen God.

Simpson spent last offseason in the gym with his dad resurrecting a classic. He also pushed his free throw shooting up to a level where it's a non-issue. He barely gets to the line; nobody's going to hack him at the end of games. As long as he maintains ~67% that project can be done.

[After THE JUMP: exploring X's final form, and his backups]

So what's next? The obvious thing is probably not happening. Simpson is not getting a jump shot between his junior and senior years of college. His existing jumper is more of a set shot.

It takes forever to get out; it's only viable when Simpson has a switch; even in those situations he hit 31%. He could boost that up to 35% or 38% and it wouldn't have much impact since the situations in which he gets the shot off are limited. Simpson's bump from 29% last year to 31% this year is negligible—literally two makes over the course of the season—and improvements next year are likely to be similarly modest.

More achievable than "have an entirely different shot" is Simpson diversifying his driving game. He goes right about 90% of the time. The hook is always a right-handed shot. This allowed teams to overplay it and ramp up the difficulty level on those until finally he could not crack the code against Texas Tech. Now that the hook is a legitimate threat opponents have to account for, unlocking the ability to go left should be significantly easier. It's a counterpunch.

If Simpson can be a legitimate two-way driver he can push his usage up another few points and maybe add back some of the efficiency from two he lost when Moe Wagner and the wide open lanes he created went to the NBA.

Other than that Simpson seems close to topped out. His defense is unlikely to get better—what would that even look like? He's already got a top 20 assist rate. The one thing that does stand out is his TO rate, which is 19. That is in part because so much of the offensive creation is on him. Senior God-slaying version of Derrick Walton had a 13 TO rate but also had an assist rate 9 points lower than Simpson.

It is extremely difficult to create as many shots as Simpson does without a correspondingly high rate of failed attempts. High major players in the A-rate ballpark this year:

• Cassius Winston, MSU: 45 assist rate, 17 TO rate
• Alex Robinson, TCU: 38 assist rate, 24 TO rate
• Jaylen Hands, UCLA: 37 assist rate, 22 TO rate
• Ethan Happ(!), UW: 36 assist rate, 17 TO rate
• Tremont Waters, LSU: 34 assist rate, 23 TO rate
• Justin Robinson, VT: 34 assist rate, 22 TO rate
• Jared Harper, Auburn: 33 assist rate, 17 TO rate

It may not be possible for Simpson to reduce his TO rate more than a couple points unless someone else significantly increases their assist rate.

The most likely outcome here is that Simpson is close to the same player he was this year. As a group, juniors becoming seniors don't improve a ton. Simpson in particular has limitations that seem unlikely to change. But dude did just perform necromancy on the skyhook. If any sub-six-foot guy who can't shoot can significantly improve going into his senior season, it's him.

Can either backup emerge into a viable alternative?

[Marc-Gregor Campredon]

It was not a pretty picture when Simpson was not on the floor.

Without Simpson the offense devolved into a lot of terrible threes and mid-range pull-ups. This isn't all on the guys called upon to play in his stead—if Michigan had a reasonable secondary creator or some decent iso players that gap would have been smaller even if the backup PGs played at exactly the same level. It's still alarming.

David DeJulius didn't play enough to have any meaningful stats—most of his shots were late jacks after Michigan dribbled the clock out in walk-on time. Eli Brooks, the nominal backup point guard for most of the year, played about a third of Michigan's minutes. He racked up a 94 ORTG on 13% usage, shooting 46/29 with a total of four FTAs.

Brooks did have a minor renaissance to finish the year, going 7/13 on twos and 4/11 on threes in the aftermath of the Weird Guys game against MSU. This broke him out of a slump that lasted most of the Big Ten season, and only got him to ehhhhh in any case. The sophomore Beilein PG-type substance leap did not materialize for him. When Lester Quinones announced that Michigan was still recruiting him and that he'd be taking a visit, a lot of fans mentally struck through Brooks's name on the hypothetical 2019-20 roster. Even with the late push that still seems like it might happen as Brooks's shot at playing time dims with the addition of NBA-sized shooting guards like Cole Bajema and potential Mystery Recruit Who's Hopefully Franz Wagner.

Things are brighter for DeJulius because he has yet to have his shot at the sophomore leap. He certainly felt more dangerous and capable during his brief run as Simpson's backup, but this did not translate on the statsheet and eventually his playing time regressed to zero meaningful minutes. The assassin who would have single-game highlight reels with 5 off-the-dribble threes as a matter of course…

…did not materialize. DDJ was frequently hesitant and unable to get the requisite separation. His occasional moments—most prominently a tough take to the bucket against Maryland—were too occasional.

Writing off freshman Beilein guards is a fool's undertaking but it is fair to say that DDJ's struggles are concerning and that his path is murkier than it seemed when he enrolled.

Can anyone feed the post? Or can Michigan run offense based on something other than ball screens?

[Campredon]

1-5 switching has gone from a Purdue stunt in which Isaac Haas flails helplessly against Derrick Walton to a commonplace response to ball-screen offense nationwide. It was particularly deadly against Michigan because of Simpson's inability to get good pull-up threes against it and the team's general ineptness at post entries.

The former isn't going to change. The latter… well, we'll see. Michigan has been the most post-averse team in the country since Beilein's arrival, and for a lot of reasons that approach made a ton of sense. Post-ups are generally inefficient possessions, and Michigan has almost never had the kind of center who would defy that larger trend. They may or may not have that guy in Jon Teske, but it seems imperative that Michigan get more comfortable dumping it down low even if that results in some goofy, inefficient possessions.

Michigan started down this path late in the year with a series of tosses over the top that were erratic at best. An offseason in which combatting switches has to be the top priority will hopefully refine the rough outline of a plan Michigan implemented in a panic late this year.

The other route Michigan might take would be to go the Purdue route and run a lot more offense for their shooters based on a variety of off-ball screens. There were hints of this late in the year, as well, with Poole coming off curl screens at the top of the key. Either way Michigan must have more ways to score than ball screens and iso.

OUTLOOK

Worst case scenario is that Michigan gets the same level of performance as this year, which is a second-team all-conference player with no viable backup. Not too bad except the rare occasions when Simpson eats autobench.

The best case scenario is that Simpson's able to add some more sexy garbage and cuts down on turnovers somewhat and gets to a 115 ORTG or so. It's already remarkable that he's at 108 on 19% usage. Our additional best-case scenario involves a DDJ emergence into a bench microwave who plays 15 MPG.

As per usual something in the middle is the most likely outcome, with Simpson increasing his usage a little and either maintaining or adding to his efficiency by being that little bit sharper on passes and more careful with the ball. He'll probably add a hook counter that allows him to score a bit more and be a little more efficient from two. His outside shoot will probably be exactly what it's been.

Meanwhile the bet here is that DDJ looks like a different player next year after his year of hazing is complete, and is the main backup at both the 1 and 2 because he adds a shooting dimension that others don't.