That, my friends, is a chucker.I knew it in my bones the moment I looked at the Kenpom profile but there it is in the flesh. Fadeaway threes? Check. Shots from a distance freshman Stu Douglass thinks is outlandish? Check. Wild drives into tons of traffic with resulting chaotic results? Check. Way too many of those shots going down? Check.
Opposing a chucker is equal parts smug haterdom and fist-shaking frustration. If Cooper goes NBA Jam on us there's not much anyone can do. If he hews to his season percentages or is off, Michigan feels comfortable unless there's an equal and opposite force preventing it.
Given the chuckerdom and OHIO's generally poor three-point shooting, there is an obvious way to defend them. I'll let two more MAC Q&A answering expert-type guys execute the reveal:
Fitzwater: Teams that play a lot of zone have given Ohio problems. Teams with strong, tall forwards can also force a lot of fouls and dominate the paint. But nobody blew OU out this season; they were in a position to win every game and were up two possessions late at Louisville. The Bobcats thrive against teams that don’t take care of the ball, don’t deal well with high ball screens, give up offensive rebounds, or depend on the three.
Hmmm. Michigan takes care of the ball, does well against high ball screens only when Morgan is in the game, has had problems with rebounding, and takes a lot of threes. But they also can play zone.
Arkley: Typically teams that slow it down (lower possessions) give Ohio more trouble. Slower paced teams that play a lot of zone even more so. The Bobcats have been inconsistent at best from three and can have problems generating points if not getting any full-court opportunities
Ohio has played well against teams that don’t have good inside-out balance offensively. The Bobcats have been good enough defensively this season to take away opponents top option or two on the offensive end. Ohio wants to run as well, so they haven’t had many problems with high-pace teams.
Also zone. Zone zone zone. Michigan's gone away from the 1-3-1 and has mostly used the 2-3 this year; it will be an option if Michigan can't contain Cooper early.
Despite the above I assume man to man is going to be plan A. Once you go to a zone it gets harder to deal with specific three-point threats like Nick Kellogg, who's hitting 42%. Also, OHIO's four is a smooth-shooting 6'8" guy with range to exploit the gaps in a 2-3 around the free throw line. If Michigan can hedge and recover on the lighting quick Cooper they'll stick with man to man. Morgan's very good at that… when he's not picking up silly fouls on the perimeter. Smotrycz is crap at hedging and a zone will be preferable when he's in the game.
The other thing the Cooper tape shows is that he doesn't need a ball screen to get penetration. Most of Ohio's sets above are 1-4s where Cooper's isoed at the top on a single defender. He gets penetration and either shoots or finds an open guy. Stu Douglass is going to have his hands full.
Rebounding: Can We Do It?
Yeah, even against a MAC team. Michigan is tiny* and the combination of Cooper and a couple of 6'8" bigs with OREB rates over 12% could lead to issues. Cooper poses a couple problems: 1) as mentioned above, late or poor hedges are likely to lead to perimeter fouls and a lack of Morgan, and 2) since it's impossible to stay in front of him one on one he forces bigs over to help. Even if the resulting shot is wild, now you've got an uncovered big on the weakside. I'm watching Louisville-Davidson as I write this and Peyton Siva missed shots are all but assists as he pulls Wildcat defenders to him and leaves his center wide open.
The good news is that it doesn't look like anyone on the roster other than the center is much of a threat. Even if Ivo Baltic is a lot bigger than Zack Novak he's a stretch four whose bread and butter is pick and pop stuff or long twos off kickouts—he has the offensive game of an unathletic NBA power forward. Think late-stage Antonio McDyess. This means he's not often around the bucket to snatch up loose balls. His OREB rate is only incrementally better than Zack Novak. The difference doesn't seem likely to manifest.
On the other end, OHIO is a relatively poor defensive rebounding team (245th, about 70 spots below average but not awful). Michigan all but abandons offensive rebounding save for its center; it does seem like Morgan will be able to do some work.
This looks like a push. If it's not it's probably because Morgan's in foul trouble and Michigan is struggling to cope.
*[For the last time. Michigan loses a 6'2" guy and a 6'4" guy after this year and adds Mitch McGary, Jon Horford, Max Bielfeldt, Glenn Robinson III, and Nick Stauskas to the roster. All five of those guys are 6'5" or bigger. Prepare for vertigo when Michigan takes the court next year a huge, athletic outfit.
I wonder if this will alter Michigan's approach to offensive rebounds. Sending Novak to or Douglass to the glass is obviously less of a good idea than sending Smotrycz—soon to be liberated to the four again—and GRIII.]
Three Point Defense: A Real Thing?
I remember talking about this a long time ago with Big Ten Wonk when he was just some guy in his basement littering the internet with bolded, exclamation-point-laden sentences. At the time Michigan was enduring a rain of threes from opponents that seemed incredible. I took the position this was luck; Mr. Gasaway said the percentages were unsustainable but evidence of a real skill, or lack thereof, possessed by Michigan.
Gasaway moved on to ESPN, where the power of the WWL has made him not only wrong about all current things ("Michigan's 1-3-1 is difficult to prepare for!" —Digger Phelps literally every time he is asked to say anything about Michigan) but also all things he has said in the past. Kenpom has set out to make this true. You've probably seen this graph before:
That's how well your three point defense in the first half of your conference season correlates with three point D in the second half. The answer is "not well." The R-squared is zero. The graph is the classic blob you'd get if you compared two totally unrelated variables like the number of letters in someone's first name and his height.
I've been reading too much Phil Birnbaum to take the above absence of evidence as evidence of absence, but it is suggestive that the graph for attempts allowed is well correlated. Statheads correct me if I'm wrong, but those are the same data sizes and for one to correlate pretty well suggests that it's not a sample size issue when the other doesn't.
If it's just dumb luck that OHIO is a really good three point defense, Michigan will play its game and get its requisite number of open looks and either hit some or miss some. If they really do stick to shooters hard, Michigan has options with backdoor plays and screen and rolls. You'll be able to tell this is happening when Burke or Hardaway or someone gets the ball to the post on a pick and roll. Is it an uncontested thunder dunk? They're sticking to shooters.
That's probably not the case. Akron hit 9 of 12 threes in the MAC championship game, and they were often like this:
A nap and a double feature of Ishtar and Dances With Wolves later, that guy knocked it down. Michigan will get their looks.
Open court turnovers. Burke's coming off a terrible performance in which he got Craft'd. Ohio forces a lot of turnovers and tries to transition that into offense so they don't have to stand around waiting for Cooper to do something. Limiting Michigan's turnovers not only helps the offense but the defense, then, and it's much worse to get your pocket picked than to chuck a ball into the stands. Michigan's fantastic at the latter but only pretty good at the former.
They'll get a test in that department tomorrow. If we're sitting around muttering on Saturday the reason will probably be double-digit transition points.
Based on UMHoops' assessment of Ohio's defense, I think we can look forward to a game in which Burke has a lot of chances to drive into the lane and then find open guys on the perimeter. Hopefully those guys can knock down some shots.
"All of the doughnuts have names that sound like prostitutes."
This game truly scaes me. I know we're better than Ohio, but if we don't come out with energy (i.e. Iowa and Minnesota games) or Morgan goes all "Bad Jordan Morgan, Bad" on us and can't stay on the floor, it could be a long day
Ohio's tendency to play overaggressive and go for offensive rebounds makes me think we will look to RUN, RUN, RUN. If we can get those easy buckets and not turn the ball over in transition (cough cough Hardaway) I think we will steadily control this game from beginning to end...
The graph is the classic blob you'd get if you compared two totally unrelated variables like the number of letters in someone's first name and his height.
I note the emphasis on "first name", but I'm not quite sure that I'd say there's no correlation between length of first name and height. Since a big chunk of the world's population is Chinese (same holds true for Koreans I believe), and since Chinese names are monosyllabic, and since Chinese on average are shorter than world average, I think there's some correlation here. Now whether it's a meaningful correlation is another kettle of fish. But it seems like there's some correlation.
"Of course I care about that stuff. To the point of irrationality. It will always be Michigan first, cancer second." Jim Mandich (RIP)
The key thing is to keep the key thing the key thing.
As the earlier games are proving, the lower seeded teams are showing a lot of energy, nothing to lose. UM has to meet that energy, yet not let it get away with them. Burke has to remain patient and let OHIO's over aggressive nature lead to cracks in their game.
I feel pretty good. I think good level-headed coaching and a solid game-plan will help us deter any possible upset. If Ohio beats us, they will have earned it. Beilein has been on the other end of this equation before. He knows what its all about.
This Cooper guy is going to find it tough to get to the basket without someone stepping in and taking a charge tomorrow. We have to be one of the best team at taking charges!?!? If only Kenpom had a stat for that.....
As a "stathead" there are a couple problems here. One it isn't an issue of having enough sample but rather too much sample. Including teams from all of various conferences in D1 will produce significant noise...basically that is saying
(without running a regression to control) that you hypothesize the same trend exists for all conferences/groups of teams. Also, given that opponent three point percentage tends to stay within a certain narrow range, there is an issue with a restricted range in the variable of interest- insufficient variation to notice a pattern.
I apologize about the dress code, sir, but my next party is a hayride