Illinois/Tourney Bullets

Submitted by Brian on March 15th, 2014 at 1:02 PM

Primary computer is currently nonfunctional; operating at suboptimal levels while trying to convince a man that he needs to take my computer and fix it. Please bear with me.

3/14/2014 – Michigan 64, Illinois 63 – 24-7, 15-3 Big Ten, BTT semifinalist


The duality of man! [Dustin Johnston]

A good way to escape, I guess. The Illinois game existed in two phases: a man to man phase in which Michigan eventually ran out to a 13 point lead and a zone phase in which Michigan attempted zero(!) two point shots that just about cost them the game. Groce's inexplicable decision to return to a man to man phase on Michigan's last couple possessions was decisive.

It is difficult to overstate how completely Michigan failed to attack the Illinois zone. From the 14:47 mark to the Stauskas free throw against man D with 55 seconds left, Michigan attempted four free throws, zero two pointers, and 15 threes. Most of those were terrible contested looks, with occasional exceptions.

In the aftermath, a couple of people pinged me on twitter, saying that's why they didn't want Syracuse. (That was before Syracuse's yakety sax final possession against NC State.

I'd still take them.)

And, yeah, that was alarming. But the thing about the How To Shut Down Michigan book is that it works until it doesn't work. It was deny Stauskas on the wing until it wasn't possible, and then it was put a point guard on Stauskas until it wasn't. Michigan will work it out. A zone wake up call is a good thing to get right now, especially when you pull the game out anyway. Much better to get that out of the way before next week.

It looked to me like Illinois was overplaying the free throw line and was leaving corner threes open, but Michigan did not take advantage. There's only five guys and Michigan is really good at shooting; they'll work it out.

Meanwhile in that's over now. Stauskas was a FTA machine against their man coverage. He hit both his two point attempts and went 9/10 of the line, all on drives. Whenever Illinois attempted to put Abrams on Stauskas things went not well for them, and the instant Illinois went back to man, Stauskas got to the line an assisted on the decisive Jordan Morgan basket.

It's worth noting that Illinois's late season surge was based on superior man to man defense. In their 6-4 stretch at the end of the year they held all but one non-Michigan opponent under 1 PPP (Iowa got 1.1 in the season finale) and held a number way under. They had a four-game stretch in which opponents could not crest the 50-point mark, and those were all good teams: OSU, Minnesota, Nebraska, and Michigan State. They're up to 15th nationally on Kenpom.

As concerning as the zone ineptitude was, a second consecutive torching of a defense that has been giving the rest of the league fits was further proof that Michigan cannot be contained on offense.

Survive and advance. And if you were feeling kinda bad about what went down, last night provided a reassurance. Syracuse went down; Duke had a near-identical one-point escape against Clemson; St. Louis ate a last second three from St. Bonaventure; North Carolina got rebounded to death by Pitt. Conference tourneys are full of chaos.

Falling into place. With Kansas's loss to Iowa State, Michigan (probably) controls its own destiny in their attempt to lock down the final #1 seed. Beating OSU and Wisconsin/MSU would give them another pair of wins against tourney teams, and it seems like everyone currently putting Villanova on the #1 line is just waiting for someone to take it from them. Meanwhile, Duke and Louisville probably can't catch Michigan without an M loss—Lunardi already has Michigan the top two seed. '

Louisville keeps coming up in these discussions because they're annihilating folks in their conference tourney, but the Bracket Matrix has them a four—nowhere near the conversation.

About what the ideal is. Is it a big deal to get the last one seed instead of a non-Florida/Arizona #2 seed? Not at the hypothetical-regional-final-if-top-seeds-hold level, where you're probably facing down the same team you bumped. But, yeah, it is a big deal. The first, second, and third rounds all feature worse opponents, especially at the Sweet 16 level. There you're facing down a four seed 35% of the time and a five or worse the rest.


Big difference between a probable matchup against a near equal (current 3s: Iowa State, Virginia, Syracuse, Creighton) and a probable matchup against someone in the 5+ range. Current fives: OSU, UConn, Oklahoma, North Carolina.

What was that? Caris LeVert drew the primary defensive assignment on Tracy Abrams on the last play, which was drawn up with about four seconds left. LeVert got super aggressive on Abrams, got beat, and was fortunate not to watch his decision get Michigan beat.

When you consider what kind of player Abrams is, that decision looks even more baffling. Abrams is bad at all kinds of attempts to put the ball in the basket but he's really, really bad at jumpers. He was just 30% on two point jumpers this year and 28% on threes. If you sag off him a bit and then come up to contest when he takes the shot you know he has to take, you're probably looking at Abrams putting up a 20% shot instead of a… well probably not 60% since Abrams is an impressively bad scorer, but way too good of a look.



March 15th, 2014 at 1:31 PM ^

Dylan at UMHoops did a good job of breaking down the probability of Abrams scoring on that last play. He shoots something like less than 30% on those floaters. That's not to say that forcing a 2-point jumper wouldn't have been better, but Abrams has been consistently missing that shot all year.


March 15th, 2014 at 1:33 PM ^

In the post-game floor interview, Beilein said of the last defensive possession something to the effect that Caris did a great job and they got what they wanted.


March 15th, 2014 at 1:37 PM ^

Except when you look at the numbers it's not so baffling. I doubt they wanted Caris to get blown by at the top of the key like he did but, again, Abrams hits that shot less than 30% of the time. He actually has a better shooting percentage on 2-point jumpers (with fewer attempts) than on shots in the lane. Also, of the three Illinois guards (Rice, Abrams, Bertrand), he has the worst shooting percentage in the lane and is therefore the one Michigan wants taking that shot.


March 15th, 2014 at 2:30 PM ^

In the post-game floor interview, Beilein said of the last defensive possession something to the effect that Caris did a great job and they got what they wanted.

Eh.. not really.

A) Beilein didn't even have a good look from his vantage point, so any analysis from his immediately after the play is flawed.

B) He's not going to say anything else publicly.  Privately, after watching the tape, he'd probably not be thrilled.


March 15th, 2014 at 1:37 PM ^


I recall reading that you have a computer engineering degree from UM.  Good gravy man, fix the computer yourself!  I have a lowly computer science degree from Arizona State, and I wouldnt be caught dead letting anyone besides myself fix my computer.


March 15th, 2014 at 6:03 PM ^

Thanks for putting yesterday's offense vs the zone issue into perspective. Whatever happens with new defensive sets in the rest of the BTT I'm sure Beilein will adjust for the tournement.

Also if you can't get your computer fixed you should consider doing a kickstarter for a new one. I'm sure mgousers could raise up to $2000 to keep content going. 

EDIT: What KOK Law said.


March 15th, 2014 at 2:02 PM ^

I saw it as him trying to use up some time by pressuring the ball immediately.  There were less than 4 seconds left, so time seemed like a bigger issue than shot selection, especially considering if they shoot quickly, it's a mess of rebounding and tip ins.  


March 15th, 2014 at 5:32 PM ^

I think the idea of comparing entire groups of seeds may be flawed in that it ignores the actual (i.e. S-curve) of the tournament seeding.

For example, seeds 1-4 have an average of 2.5 on the 1-64 S-curve, while 4-8 have an average of 6 (and so on). In such cases one would expect a drop off regardless if the decrease in quality of teams (or increased quality of opponents) is linear or not.

If you really want to show that a one seed is fundamentally better, do a comparision between seeds (on 1-64 S-curve)  4 and 5, 8 and 9, and so on, where you are comparing the lowest N and highest N+1 seed. Then I would be convinced.

I would do this myself, but I am a lazy freeloader off the Mathlete.




March 16th, 2014 at 9:14 AM ^

That is a really nice chart, it shows that a 10-12 seed is more favorable than an 8-9 seed. This is likely because they almost always play a 1 seed in their second game, as opposed to 10-12 only playing a 2-4 somewhat less often than an 8-9 plays a 1. Randomizing the seeding cohorts (3-4, 5-8, 9-16) would be fairer.