[Marc-Gregor Campredon]

Basketbullets: Survive And Advance, The Season Comment Count

Brian January 10th, 2020 at 12:22 PM

1/9/2019 – Michigan 84, Purdue 78 (2 OT) – 11-4, 2-2 Big Ten

"Survive and advance" is the kind of thing people say in March after a 13 seed gives their team the business for 35 minutes and then skids to a halt just short of the finish line. As bits of conventional wisdom go, it's actually not bad. You're locked in a 68-team tournament with a bunch of opponents who are also pretty good. There's no such thing as a bad tourney win. Yes, even that one against Montana.

Michigan just played a 2 OT game in which six guys got more than eight minutes. We are in survive and advance mode. Every time Michigan poked their nose in front late your author had a thought bubble appear above his head. In it, his friend Jerry popped up, said "survive and advance!," and then muttered something about Huddersfield Town FC your author tried to ignore.

This is a comedown from annihilating Gonzaga to win Poseidon's Trident but I've mostly made my peace with it. Sugarplum dreams of this team being elite have given way to some harsh realities. Michigan is pretty short and not particularly athletic and the preseason expectations were closer to right than bouncing up to the #4 team in the country. The Big Ten is a ball of knives. We are barreling towards the first-ever NCAA tournament with Rutgers but not North Carolina.

Every game is a swing game, except one. Kenpom gives Michigan a 92% shot at beating Nebraska at home just before the season's end. They do not have an 80%+ shot in any other game. Nor do they have an insurmountable task. Games at OSU and Maryland sandwiching the Nebraska game are 26% and 29% shots. Everything else is in coinflip land. There are no bad wins from here on out, especially when Michigan is down its most efficient offensive player and one of three defensive cornerstones.

Did we win? Good. No further questions.

[After THE JUMP: Further questions!]

The finishing lineup. After falling down by five, Michigan threw out a lineup of their three guards with Teske and Wagner at the 5 and 4, respectively. This made the offense feel a million times better; Michigan did not substitute for approximately the last 20 minutes of the game. A large part of this was putting three bonafide shooters on the floor, which eliminated the issues Ace posted about after the MSU game:

Better spacing allowed Simpson to go to work against Trevion Williams in a two-man game with Teske. Williams was dominant on the offensive end of the floor but remains a plodding defender who does not block shots; Simpson went from a guy who didn't take a shot in the first 16 minutes of the game to Michigan's leading scorer because he was 8/8 from two.

Inescapable conclusion: Michigan was very fortunate that Matt Haarms exited this game with a hip injury. It is unlikely they would have been able to get away with that lineup otherwise. With Haarms out, Purdue's second-biggest player is either Aaron Wheeler or Nojel Eastern, depending on whether you go with height or weight. Those guys were collectively 1/8 from the floor, which is approximately what they're shooting away from the basket on the season. Wagner had no problem checking them. Haarms would have been no worse an option on offense and would have been a problem on the offensive boards. He would certainly have swatted enough Simpson shots to give Michigan a regulation L.

49359194593_f927989e89_k

mano a mano [Campredon]

On the other end of the floor, though. Not doubling Trevion Williams was potentially justifiable for the first ten minutes. Last year Michigan threw Jon Teske in to sink or swim against everyone and he came out very well. But after Williams hit 6 of his first 7 shots it was time to double. After he hit a career high it was time to double. By overtime it was time to double. Nobody ever doubled him.

I get that post offense is usually not super efficient but you have to look at the context. Purdue had at least one of Eastern (0 3s) and Wheeler (23%) on the floor at all times; Jahaad Proctor (29%) was also out there most of the time. If Williams is killing you, doubling off one of those guys seems to make a lot of sense. Especially if the guy coming over is Wagner with his crazy gumby arms that should help you force a high-arcing pass that allows you to recover.

Purdue doubled Teske virtually out of the offense after he picked up a quick eight points. Michigan was able to pick it up because they've got some other guys. Purdue doesn't, really.

Did something happen to our defensive aces? One of the more disappointing developments on the season has been Simpson and Teske seeming to lose a certain something on defense. As mentioned, last year Michigan let Teske sink or swim in the post because it worked. Various opposing C lines last year against teams that actually used their C to create in the post a lot:

  • Juwan Morgan: 25 points on 26 shot equivalents, 3 TO, 98 ORTG
  • Juwan Morgan: 16 points on 15 shot equivalents, 3 TO, 94 ORTG
  • Ethan Happ: 26 points on 24 shot equivalents, 7 assists, 1 TO, 117 ORTG
  • Ethan Happ: 18 points on 19 shot equivalents, 1 assist, 5 TO, 80 ORTG
  • Kaleb Wesson: 12 points on 12 shot equivalents, 1 A, 1 TO, 107 ORTG
  • Bruno Fernando: 12 points on 11 shot equivalents, 1 A, 3 TO, 97 ORTG
  • Bruno Fernando: 12 points on 14 shot equivalents, 1 A, 3 TO 92 ORTG

The one legit good game in there is from Happ and it's largely because he assisted on some buckets—not Teske's problem. The one guy to give Teske even a small portion of the business last year was Luka Garza, who put up 19 on 14 shot equivalents with one turnover at Carver-Hawkeye.

Teske started the year seeming to be about the same defender. He obliterated Filip Petrusev in the Bahamas. Then things got rough. Garza—who is now Shaq I guess?—put up 44 at Crisler, going 17/29 from two. Kofi Cockburn put up 19 en route to a 122 ORTG. And Williams just put up 36 on 31 shot equivalents with no turnovers for a 134 ORTG.

To be fair to Teske, Williams hit an inordinate number of shots somewhere between tough and "now my toddler is repeating swears to me."

49360312707_60655009a9_k (1)

mere moments before many children of Michigan fans learned a new word [Campredon]

Even so, three events make a trend. There is a potential mechanism:

We won't have to wait long to get further data: Daniel Oturu (64% from the floor in KP tier A games, monster OREB numbers) and a return game against Garza are up next.

49359664361_35ac38c9b1_k

creates shots on a team that needs that [Campredon]

Time to make the switch? I don't want to say this after devastating Ace with my "Eli Brooks is an upgrade on Jordan Poole" Fuegobox™ hot take earlier this year, but I think Brooks has turned back into a pumpkin. This was his second consecutive single-digit-usage game. He scored 4 points, one of them a transition layup. His ORTG in Kenpom A/B tier games is down to 82. Since getting back from the Bahamas he's 3/16 from three against high-majors, with all of the makes against Iowa and its abominable defense.

Meanwhile David DeJulius is coming off a six-assist game and seems to be rounding into an excellent defender himself. He suffered the worst call of the game when he got called for a blocking foul after taking a forearm to the chest, and most of his other defensive involvements were positive. If DDJ is in the same universe as Brooks as a defender, he should be starting once Livers comes back:

In practice this won't mean a ton—take 10 Brooks minutes and give them to DDJ. Still, it might be important. I think we know what Brooks is going to be like, and it's time to see if DDJ has some takeoff velocity.

49359874117_41778937e9_k

raaaaaaaakes [Campredon]

Franz emerges a bit. Wagner remains frustrating because he turns the ball over too much and hasn't quite found his shooting, but it says something that the game rescue lineup had him in it. He put up 15 and was reasonably efficient despite going just 3/9 from three. I'm not entirely sure he blocked Williams's shot at the end of regulation—looks like it went off the side of the rim first—but he was able to rush it and help influence the miss. He also picked up a couple of steals.

More than anyone else Franz could use Beilein-level spacing. When he goes to the rim you can see the ability; far too often he gets stripped by an off-ball defender raking down from the wing. Part of that is on him—he's got an assist rate of 2.7—and part of that is Michigan not having the requisite shooting, particularly without Livers.

Getting Franz working downhill and at the basket is doubly important because he is 16/17 on FTs after being a 90%+ plus guy in Europe last year. If he gets to the line it's automatic.

Franz's development is probably the most important story to track over the next ~9 months of Michigan basketball since he's got the biggest gap between his potential and his current output.

The bell should be tolling. Adrien Nunez got three minutes in which he missed a three, turned the ball over, ate a blow-by, and committed a foul. At this point the "he must be doing really well in practice" thing has got to be over. He's shooting 28% from three and is the worst defender on the roster. What does Cole Bajema have to do to get those minutes? 

Savage. Zack Novak:

Don't get used to Phil Martelli. From a Seth Davis Athletic article on coaches who could be on the move this offseason:

…while Martelli is enjoying his time in Ann Arbor and thinks the world of Howard, he misses his wife and family, who are still living near Philadelphia. Martelli did not promise Howard he would be around for a certain period of time (Martelli has a two-year contract with an option for a third), and while he told me he hasn’t given much thought to what will happen after the season ends — “The only thing I know is that I’m fully dedicated to Juwan coaching on a Monday night in April” — Martelli very much wants to be a head coach again. “This is an ego business,” he said. “Everybody wants to run their own program. Wherever I am, I want to be in position to pursue championships.”

This article also informed me that Villanova has an assistant named "Kyle Neptune," which sounds like a bagging alter-ego that overtook its host body.

We did it. Always notable in a 353-team Division 1 when you hit #1 in anything:

imageYou shall not shoot threes.

Comments

Dodort

January 10th, 2020 at 12:40 PM ^

While Garza gave Teske the business, I thought Williams was mostly just him making tough shots.  He got to the baseline a couple of times late, but for most of the game, I thought Teske's defense was good.  According to Torvik, he was 8-16 from mid-range (and 1-1 from 3), which is much better than you would expect from a post player.

J.

January 10th, 2020 at 12:48 PM ^

100% this.  Purdue got 0.97 points per possession.  I wouldn't care if 100% of those points were by Trevion Williams.  The defensive strategy worked.

If you're playing the percentages, and the opposition takes a bunch of low-percentage shots, and makes them, the correct answer is not to change what you're doing, presumably giving them higher-percentage shots.  The answer is to keep going and count on the law of averages.

joeyb

January 10th, 2020 at 1:05 PM ^

You've said it best. In all sports, there are components of skill and components of luck. Forcing opponents into low-percentage shots is skill and whether they make those shots requires some skill on their part, but is largely luck to exceed the percentages.

Nothing better exemplifies this than Williams hitting that 3 at the buzzer. A guy who was completely left alone in 3-point land and just would not shoot hit a step-back, buzzer-beating, fade-away 3. If that's not a shot that you want them taking, then I don't know what is. The fact that he hit it is inconsequential on future game-planning.

Phaedrus

January 10th, 2020 at 1:31 PM ^

And it's important to note that Williams missing shots at the end (of regulation and both overtimes) is what cost Purdue the game.

Fans (and sometimes coaches) put a lot of stock in the "hot hand," but when the hot hand is consistently taking bad shots the best strategy is to let him keep doing it. Mistaking an anomaly for a trend happens to us all while immersed in a the moment, but this is a case of not seeing the forest for the trees. If Williams took that same three point shot ten times how many of those do you think he actually makes? When an offense succeeds by doing the bad choice you have forced them into all you can do is tip your hat.

pescadero

January 10th, 2020 at 1:48 PM ^

" If you're playing the percentages, and the opposition takes a bunch of low-percentage shots, and makes them, the correct answer is not to change what you're doing, presumably giving them higher-percentage shots. "

 

Against a team with shooters - you'd be giving higher percentage shots.

 

Williams averages 1.2 pts per 2 point shot.

Purdue as a whole average less 1 point per 3 point shot.

More 3's for less 2's from Williams is a winning strategy.

TrueBlue2003

January 10th, 2020 at 7:03 PM ^

This isn't how three point shooting works.  They're a 33% shooting team.  That's on threes that they chose to take.  If you change your strategy to make them chose to take a three, that's the percent they're expected to take.  Three pointers generally only come in two types: one that is open enough to shoot and one that is not so that I don't take it.  That's why there is no statistical evidence of "three point shooting defense".  You either prevent them from going up or you don't.  The ones that go up will on average be made at the rate the offense is capable of making them.  Purdue is not an offense capable of making many of them.

And even if you think they'd get a slight bump, there's a huge gap in efficiency between what Williams was doing (which he was doing at 1.35 ppp) and 33% three point shooting.  Even if they shot 40% from three, that's 1.2 ppp.  Still better than he was doing.  If you're getting killed that badly, what you're doing is not working.

TrueBlue2003

January 10th, 2020 at 1:59 PM ^

0.97 points to Purdue at home is not good. It is barely better than what an average D1 team has done this year against Purdue and significantly worse than Michigan was expected to do.  

The strategy did not work. Whether that means it was a bad strategy is up for some debate I suppose.  If one thinks that Williams just had an out of body experience and that he should have cooled off and simply never did is...debatable but I lean heavily towards Brian's side of things: fine for 10-20 min but when it's clear the guy is on fire/has something with the matchup against Teske AND you've seen the story over and over again, it seems like you have to try something else.

A bad Purdue offense stayed in a game it had no business being in, especially given then Haarms was out.  So...it didn't work.

TrueBlue2003

January 10th, 2020 at 6:49 PM ^

It's not horrible but it is bad. They were expected to score about 0.92 (their offense is only slightly better than average for D1 which is very bad for a P5 team, so they'd be expected to score slightly more than M's adjD which is about 90.

For another take, go to Bart Torvik who actually shows you each game compared to expecations.  Purdue's AdjO for the game was 109.7, which is significantly higher than their 104.2 season AdjO.  And I'm not sure if that adjustment takes into account the fact it was on the road and if it doesn't, that's even worse.

Purdue has a bad offense and they were better than they should have been against M.

Michigan's offense was actually quite good, compared to expectations.  Another case where the context matters.  Purdue has an elite defense (although that's with Haarms).  Michigan's adjO for the game was 117.5.  Significantly higher than their 110.6 season adjO.

Despite shooting meh from three, Michigan hit a scorching 24/37 from two (thanks to Z) and only turned it over 9 times in 80 possessions which is EXCELLENT.  It was actually the best they took care of the ball all season. That's probably why your perception is off.  You're focusing on how they shot but the mere fact that so many possessions ended up in a shot more than makes up for some bad three point shooting.

And they were generally great after they went to the three guard lineup.   They had 49 points with about 5 minutes to go in regulation and scored a whooping 35 in the last 15 minutes.  Part of that is due to Purdue extending the game but it passed the eye test with Z and DDJ creating, finally some threes falling, Teske finishing...

jmblue

January 10th, 2020 at 12:49 PM ^

 Michigan is pretty short and not particularly athletic and the preseason expectations were closer to right than bouncing up to the #4 team in the country. 

I think it remains to be seen how far we can go with a healthy Livers.  We look like a bubble team without him, but with, I'm not sure where our ceiling is.

AC1997

January 10th, 2020 at 1:00 PM ^

Michigan - where we make your best player hit their career highs in Big Ten games.  (I guess it doesn't matter as long as we win more than we lose, but it was frustrating.)   

I'm not as convinced that we would have been in trouble with Harms healthy.  I actually think he takes some minutes and usage away from Williams and maybe he doesn't get as hot or hit as many circus shots.  Because they had no one else, they played him a career high in minutes and kept feeding him on every play.

As for Brooks, he did a great job on their designated shooter.  I don't even care if they keep starting him and use DDJ as a spark off the bench....I just want the minutes to shift more to DDJ.  We have to get better shooting out of Brooks, DDJ, and Wagner - that's the key to the season.  

No mention of Johns?  I thought he was.....okay.  Slightly more confident than against MSU, still a work in progress.  

Is anyone even SLIGHTLY concerned that none of the stud recruits we're in the market for are known as being good shooters?  

KTisClutch

January 10th, 2020 at 3:16 PM ^

Not sure why people are telling you that Christopher is a good shooter. He's passable. He takes a lot of tough shots so it probably remains to be seen how good he is at catch and shoot.

 

None are deadeye. But Livers wasn't deadeye before he got here. Zeb, Williams, and Todd all can be very good shooters. If Todd is mostly playing off ball of Christopher I'd expect him to be a good shooter. Wagner getting his % up next year is important though

ijohnb

January 10th, 2020 at 1:04 PM ^

It is odd that Howard is this averse to double teaming.  He player for Steve Fisher who basically invested the auto-double in the post.  Possibly the players secretly hated it and it scarred Howard for life with regard to double teams?

TrueBlue2003

January 10th, 2020 at 1:12 PM ^

That was nearly 30 years ago.  Basketball has changed a lot and Juwan has had like 1000 coaches since then. He probably barely remembers the way they played then. 

He is wayyyyyy too three point averse, though.  Needs to adapt his NBA style more for college.

TrueBlue2003

January 10th, 2020 at 1:08 PM ^

Yeahhhh, yikes. That stat is more proof amongst a mountain of it that Michigan is too concerned with preventing threes.  The line was moved back this year.  The calculus has changed. And this isn't the NBA.  They need to be smarter about helping off bad three point shooters, especially when they're helping on guys that aren't good passers.  Which brings up the next point:

You nailed it about Wagner being a black hole right now.  He is the next iteration of his brother (who learned to pass a little eventually) and Iggy.  That's an even bigger reason teams know they can help aggressively on him.  In Ace's post about Johns unwillingness to shoot, he highlighted a play on which Wagner drove and Tillman helped onto him.  That's more a result of Tillman knowing that Franz isn't going to find the open man and that he can help with abandon.  (Also a big reason Franz is struggling at the rim.  He has two to three guys surrounding him and will not kick it)  

Which brings me back back to our scouting.  These are the things other teams know and do.  Michigan steadfastly sticks to a singular defensive philosophy no matter how badly they're getting killed in the post, who is standing to on the three point line, who is driving, etc. 

Their come down since the Bahamas is as much on the coaching as anything.  In an early season tournament with back to back games, there is little from which to scout teams and no time to even do it or prepare a game plan.  So teams just play, and Michigan did well because they have a mostly veteran team with quite a bit of talent (they started three second team all big ten level guys and four upperclassmen which is nearly un heard of in P5 basketball). 

Now that coaching wins the day and the team is going against big ten coaching, they're getting their lunch handed to them, tactically. Will be interesting to see how the staff develops.  Juwan seems willing to learn and adapt.  I think they'll get better.

remdog

January 10th, 2020 at 1:13 PM ^

Actually, I think it helped Purdue and hurt Michigan to have Haarms go out since Williams was en fuego and got more shots.

Otherwise,  I agree with Brian's take.  DDJ is playing confidently and with great poise in his ballhandling, shooting and passing.  He is a sparkplug offensively and is improved defensively.  Meanwhile, Brooks has regressed.  So DDJ definitely has earned some of those minutes.  And Nunez's minutes need to go to DDJ and/or Bajema.

I also think that Wagner and Johns are improving - fighting through some rough spots and making some big plays.  Johns, in particular, is gaining some valuable experience in Livers' absence.

4th phase

January 10th, 2020 at 1:25 PM ^

I don’t want to go so far as to say Haarms being out helped Purdue, but I do think it made Purdue’s post offense more efficient. Haarms and Williams are both guys that play near the basket, they would have ended up having spacing similar to the type that everyone complains about on here, 2 marginal to bad 3 pt shooters hanging out in the lane. With Haarms out, Eastern, Stefanovic, and Wheeler are all guys who kind of stand around the perimeter leaving Williams in iso on Teske.

 

Also while it’s annoying that post players keep having career days against Michigan I think you have to keep with the strategy, it’s frustrating but it’s a lot better than letting the other team rain 3s while Michigan is struggling to shoot the 3. That’s a recipe for getting blown out.

InterM

January 10th, 2020 at 2:11 PM ^

I will continue to lobby for a redesign of Crisler into an underwater, poorly lit ballroom a la Atlantis.  Sure, some of you will insist on being eggheads and talk about opposing team scouting and such, but Michigan hasn't hit its threes since the team left the Bahamas.

AlbanyBlue

January 12th, 2020 at 8:38 PM ^

This column is especially prescient after today's game. 

Nunez: does not belong. Was -7 in his time on the floor in the first half.

Brooks: On O, yikes.

Still no doubling. Teske eaten up by Oturu.

Memo to Juwan: More minutes for DDJ. Force the ball out of the post at all costs. Profit.

Hannah Green

January 20th, 2020 at 7:46 AM ^

One of the thrilling matches of my life and this match was fully enjoyed by me in the stadium. Visit topics mill for getting best topic ideas. I would love to go and watch more matches of this team. Jerry did very interesting job for the Michigan players.