2/21/2012 – Michigan 67, Northwestern 55 (OT) – 21-7, 11-4 Big Ten
LEFT: A fateful moment in which our brave comrade fouled the opponent's forearm in the eyes of corrupt capitalist lackeys. RIGHT: The imperialists were forced into illegal measures in their failed attempts to deal with Comrade Morgan.
Let the reign of Beilein be long and glorious. He is our sun and star and moons. He has brought basketball back to Ann Arbor long after we had ceded our land to the imperialists of East Lansing and set about hoping we would not be Northwestern forever. The bubble is banished and all loyal Wolverines are required to have Mao-style paintings of not one but two Dear Leaders. This is right and just.
But we have to talk about something, Oh Great Back Cut of Heaven. That thing is what to do when Michigan's glorious but thin frontcourt, sabotaged by foreigners who broke Comrade Horford's foot—we have executed the traitors or at least given people who probably know the traitors harsh looks—is brought low by the machinations of imperialist pig-dogs with whistles.
Oh Thousand Shining Arcs From Behind The Line, your response in the Northwestern game was to bench Comrades Morgan and Smotrycz in favor of Comrades McLimans and Christian. They are a good loyalists who contribute all they can to the cause. Unfortunately for the Glorious Revolution, that is zero shot attempts and zero rebounds in fourteen minutes. "From each according to their ability, to each according to their need" suggests that Comrades McLimans and Christian are most needed in the towel-waving collectives of Ukraine, where they can fan our team to greatness.
When they are placed on the court, starvation ensues. Michigan led 11-3 when McLimans entered the game; Northwestern led 31-24 by halftime, when Comrades Morgan and Smotrycz returned to the floor. In that span of time, Northwestern had six offensive rebounds in eight opportunities*. In the other 31 minutes they had five in 24. Northwestern scored more points in the fourteen minutes without Morgan and Smotrycz (28) than they did in the other thirty-one(27).
Upon their return Morgan and Smotrycz promptly led a glorious charge into a lead foreordained by your divinity, Great Leveler. Unfortunately, rebel conspiracy sabotaged the bridge between Tim Hardaway Jr. and free throw makes, forcing the revolution into overtime. The people rose up and slew their purple oppressors with a thousand swords, as you decreed would always be the case.
Some of our less faithful comrades may have momentarily lost confidence, however. While the will of the people can never be defeated, it should be pointed out that basketball teams can and putting in comrades who are not very good at basketball could lead to a (temporary, meaningless) setback in Michigan's five-year plan.
When comrades Morgan and Smotrycz returned it took 12 minutes for one of them to pick up a third foul. If they were allowed to continue playing in the first half it is true they would be in danger of fouling out early. But what would the consequences be in that situation? In the worst vaguely plausible scenario, both Morgan and Smotrycz foul out five or six minutes into the second half, forcing the Striped Orange Sun to… play McLimans and Christian for 14 minutes. The wisdom of the Shining Beacon of Halftime Adjustments is unquestioned, but in this one situation it seems like it is not infinite.
Earlier in the year, a similar substitution pattern saw Comrade Burke confined to the bench for a long stretch against Iowa. Burke left with Michigan down four and returned with them down twelve. Nefarious play by oppressors made Michigan play poorly throughout, so this did not make an impact on the outcome, but it didn't help matters much.
I submit that with Burke averaging 1.8 fouls per 40 minutes at the time of his transgression and six additional calls available to a two-headed center playing a team without any size, it would benefit Michigan greatly to roll the dice on players in foul trouble instead of willingly accepting the worst-case scenario of doing so. Oh sun and moon and stars.
*[It was actually 7 of 9 but one was a OREB credited to Northwestern's team after McLimans blocked a shot out of bounds. I don't think that shows up on the box score I'm using.]
Defense! Zounds. UMHoops says Michigan had Morgan and/or Smotrycz for 40 possessions. On those possessions Northwestern scored 27 points, or 0.68 points per possession. That's outstanding. Northwestern has the country's 15th-best offense and the league's fourth-best; when Michigan wasn't going to the deep bench because of the aforementioned rigidity they annihilated the Wildcats.
The primary way they did this was by switching everything. IIRC there was a single breakdown for an open layup in the first half, then nothing the rest of the game. Everything else was contested. John Shurna was 2 of 5 from three and 4 of 11 from two with a couple of those twos ridiculous circus things; after the game Bill Carmody kind of called out his leading scorer for passivity:
"It just seemed the whole game that he was reluctant to do anything," Carmody said. "He had some pretty good looks and he passed them up to go to the next thing. It was a game he had to take over."
Northwestern never tried to punish Michigan for switching players as small as Trey Burke onto Shurna. That's either blind luck or great scouting.
Threes? Michigan hit 37% on 38 threes for 1.1 points per attempt. Are we happy with that? I have no idea. On the one hand, a lot of those were wide open when opponents sagged off Burke or left a corner three open in the 1-3-1. On the other hand, 38 threes. I'm guessing we would have had a much different opinion than confusion if Burke and Hardaway didn't put down back-to-back triples after Michigan found itself down four late. Those makes opened the door for the rat-tat-tat at the beginning of overtime. Before that the numbers were ugly.
1.1 points is not great. It sounds good as a shooting percentage but you have to take into account that way more twos than threes end up getting erased in favor of free throws. On the other hand, being willing to launch from deep really cut down on Michigan's turnovers (six to an uncharacteristic 14 for Northwestern) and would have led to some additional possessions via Morgan offensive rebounds if the refs hadn't gone from suck to blow in the second half. In the end, it worked. Worked authoritatively when Morgan/Smotrycz were in.
1-3-1 response. When Michigan's 1-3-1 was getting shredded early in the disappointing Harris/Sims post-tourney year it seemed like opponents were attacking it diagonally and getting to the basket. Michigan was hesitant to put the ball on the floor at all and ended up shooting over it on a large majority of possessions. When they did dump it in low, Morgan had a couple opportunities but didn't go up strong, as they say, and we got the obligatory missed bunny or two*. I wonder if Northwestern just runs the 1-3-1 a lot better than Michigan ever has in the Beilein era.
*[This should be less of a problem with McGary. When people are asking Morgan to go up strong they believe he can dunk a ball from a standing start under the basket, which I don't think he can. This should be no problem for McGary as long as he can catch cleanly—always an issue with big men.]
Hardaway. Yerg. Back to the salt mines: 2 of 9 from three, 4 of 10 from the line. Two of three from two… against a team that has no shot blocking. I don't think those distributions are going to get fixed this year; we can only hope the shots go down when Michigan really needs them to.
How. in the HELL. do we lose two games to the same ranked team in overtime? HOW? Why does this happen? THIS IS JUST THE WORST.
Northwestern has now played about 8000 close games this season and lost all of them. Here are my questions, and I am furious about each and every one.
You get the ball witha bout 50 seconds and a full shot clock. Instead of opting to go two-for-one and take the last shot, which ANYBODY WITH ANY SORT OF BASKETBALL SENSE IN THE WORLD would have done, Northwestern held for 35 seconds and had a possession end with a JerShon Cobb three, a shot which is about as efficient as repeatedly stabbing yourself in the face. YOU DON'T WANT TO PLAY ANOTHER FIVE MINUTES AGAINST A RANKED TEAM WITH ALL THE MOMENTUM. YOU WANT TO END THE GAME IN REGULATION. YOU HAVE A BETTER CHANCE OF BEATING A BETTER TEAM IN THREE POSSESSIONS (TWO OF WHICH ARE YOURS) THAN FIVE MORE MINUTES. This is inexplicable.
We will root for Northwestern from here on out. We have hurt them more than they deserve. AnnArbor.com on Vogrich:
"He's been a big part of this little surge we're having right now," Michigan coach John Beilein said of Vogrich. "You've seen all year long that we've struggled with our bench play.
"And we need that. He's done a good job."
During Michigan's current four-game winning streak, Vogrich has gone 9-for-13 from 3-point range, providing a spark off the bench that hasn't been there for most of the season.
Known as a 3-point specialist, Vogrich entered the Nebraska game on Feb. 8 shooting just 21.2 percent from behind the arc. But thanks to his recent hot streak, he's jumped up to a more respectable 33.3 percent on the year.
But what would the consequences be in that situation? In the worst vaguely plausible scenario, both Morgan and Smotrycz foul out five or six minutes into the second half, forcing the Striped Orange Sun to… play McLimans and Christian for 14 minutes. The wisdom of the Shining Beacon of Halftime Adjustments is unquestioned, but in this one situation it seems like it is not infinite.
That scenario would be worse than what we experienced. Second-half possessions are more valuable than first-half possessions due to their relative scarcity. When you sit Morgan in the first half, it sucks but you still have an abundant number of remaining possessions in the game to recover from it. If you sit him in the second half and NW goes on a run, that's it - game's over.
I know the counterpoint to this is that maybe we'd have built a comfortable lead before Morgan fouled out, but I don't know if that would have happened. A player on the floor in foul trouble - especially a big - is not going to be as effective as normal, because he can't play aggressively. A Morgan unable to throw his body around for fear of fouls might still be better than McLimans, but not by as much.
In the above scenario, we'd be dealing with a limited Morgan for several minutes, and then no Morgan at all for a long stretch to end the game. What we got instead was no Morgan for awhile, but when he returned he could play his usual aggressive game.
Co-sign. As someone who fouled a lot because he was playing as a very short power forward, trying to avoid fouls is the quickest way to kill a frontcourt players aggressiveness and therefore effectiveness. I do think Belein may be a bit to much of an absolutist, you've got to maintain contact, which killed us against Iowa, but in general id much rather have an aggressive Morgan in the game's back half than worry about losing a lead.
I bet the Mathlete, were he so inclined, could figure out a way to quantify whether it matters when possessions happen in the game. Intuitively, I say that 2 points in minute 10 means the same thing as 2 points in minute 30, but it's possible there are some game theory issues I'm not thinking through. This is an argument in favor of getting points now and not worrying about foul trouble so much.
On the other hand, your point about aggressiveness is well taken. That makes me wonder whether 80% of Morgan is better than 100% of McLimans. I'm not sure either way.
Finally, as a middle ground, I once heard Jeff van Gundy, I think, argue that it was silly to keep players with 2 (or 3 in the NBA) fouls out of the last few minutes of the first half, since they could just as easily pick up a third (or fourth) foul in the first few minutes of the second half. So if you're willing to start Morgan in the second half, you should play him at the end of the first half. If he doesn't pick up a third foul, you've stolen a few minutes. If he does, then you sit him for a few minutes to start the second half and you're back to square one.
I don't think you even need the Mathlete for that. Just go to KenPom (if you've got the subscription) and look at any win probability chart for any game that stayed pretty close throughout. You'll see that a two-point bucket that happens early might swing the probabilities by 2-3%. A two-point bucket that happens late - especially within the last two minutes - can swing them by as much as 5-8%, or even 10%. 15%, in some extreme cases. I think clearly, the time of the possession matters.
If you got blown out in the first half, the discussion is basically moot because you probably had your top guys on the court anyway, for at least part of the other team's run if not most of it. After all, it takes a lot of possessions to turn a regular game into a hopeless one.
...if you turn a potentially hopeless game into a regular one, what does THAT do for win %?
If you leave Morgan and/or Smot out there for more of the second half, and make a (reasonable) assumption is that Michigan's production would roughly equal to that of other similar situations in the game, then Michigan probably goes into the half up... and up big. The Look at the advanced stats betwen McLimans time and non-McLimans time. If you're up 25 when Morgan fouls out, the difference in KenPom win % for any future made baskets is pretty damn small.
I find many of the stats meaningless in the moment. This team is coming back when down. They are going to find a way. M did get blown out against Arkansas, on the road, and I'll be damned if they didn't find a way to come back and come within a rim rattling miss of winning the game. M did get blown out at Indiana and I'll be damned if they didn't scratch their way back to a near victory. More often than not this team finds away and that is difficult to quantify,
I'm saying that end-of-game situations are only high leverage if the game is close. If you're down 10 with 20 seconds left, a bucket isn't worth a damn thing. The assumption everyone is making here is that sitting players with 2 fouls is good because you have them for the stretch run. But that doesn't remotely hold if the "stretch run" involves being down 15 because you sat all your good players.
There is no time, sir, at which ties do not matter.
because the Northwestern game ended up close while the Iowa one did not.
Being down 7 at half is not nothing. I don't have a win probability chart in front of me, but I would guess Michigan was a fairly substantial dog going into the locker room for halftime. That Michigan went on a run in the second half doesn't alter that.
There is no time, sir, at which ties do not matter.
I do have a chart. At halftime, Michigan was down by 7 and their WP was about 29%. They hit the first bucket of the half to swing the WP% 6 points to 35%.
Just after 3:00 to go in the game, we were back down to a 29% chance, but with only a four-point deficit. That alone should tell you the value of the next possession. We then hit a three to swing it all the way to 42% - a 13-point swing on one basket.
We started the game with a 55% WP, and it took an entire half to get it down to 29%. At about 8:30, we had it back up to 65%, and instead of 20 minutes, it took five, to drop it back to 29%.
I agree somewhat with Brian's point, but that doesn't allow for the possibility that a coach can look at the referees and say "man, they are calling 'em chintzy tonight. I better protect my best players. Especially the ones in the frontcourt who are more likely to pick up fouls." And it's plainly evident the value of a single possession in the first half is nowhere near the value of one later on.
And it's plainly evident the value of a single possession in the first half is nowhere near the value of one later on.
Not true. That's the same logic that says that early season at-bats matter more to your season-long batting average than do early ones, because it's easier to move your average in your 10th at-bat than in your 500th.
The value of possessions in a given game may seem higher late than early, and the KenPom percentages may be more easily swayed late rather than early, but the numbers aren't independent. The value of the late-game possessions are hugely dependant on the results of earlier possessions. The value of Trey Burke's last lay-up against OSU would have been vastly different if Morgan hadn't shut down Aaron Craft on the previous possession.
This makes sense when you consider the fact that an early bucket is worth the same # of points.
Ask OSU, who fell behind 6-0, and lost by 5, or NU, who fell behind 5-0, and lost in overtime, if it's "plainly evident the value of a single possession in the first half is nowhere near the value of one later on."
Not a valid comparison at all. The goals are completely different. A basketball game is a win-or-loss proposition. Black-and-white. A batting average is simply that - an average. All you have to do is ask yourself a question: would you rather be down by three points with 10 seconds elapsed, or 10 seconds left? What does the next possession mean in each case?
Of course your point about Burke's layup and being dependent on previous possessions is partially correct - but then, I noticed you instinctively chose the immediate previous possession instead of one that happened early in the first half.
Wouldn't my point be equally valid if Ohio State had made a random lay-up in one of their first few possessions?
Conduct a brief thought experiment: reverse the first and second halves of the box score. By your argument, moving the horribly unproductive late-first-half stretch to the END of the game would be somehow more harmful. Yet we know that simply flipping halves doesn't change the outcome.
Instinctively, it FEELS like late possessions are worth more. But mathematically we KNOW they aren't.
Yes, what you say is technically true (until the last paragraph), but you're forgetting one important thing: Decisions are made within the game, not after it. So you have things like win probability to give that idea some voice.
In return, try this thought experiment:
- It's a tie ballgame. You're about to go on a 10-0 run. You can choose between one of two situations: it can either be 2-2 with 38 minutes to go, or 52-52 with three minutes to go. Which would you prefer? It'd be totally illogical to pick the first, because in that case, the opponent has most of the game to catch up. In the second, you're almost certain to win. Since coaching decisions are made in-game instead of afterwards, it's only logical to look at plays and situations and possessions from the same perspective, instead of in hindsight.
It's similar to the questions I asked above: You're down 6-0. What can you do to win? Answer: a lot of different things. The game could unfold with near-infinite possibilities. What happens on the very next possession is relatively unimportant because of how many possibilities remain. You're down by six with 45 seconds left. What can you do to win? Very little. The next possession is crucial. That's why this is different than your batting average analogy.
A logical extreme helps. If you're down 2-0 and you hit a three-pointer, you're up 3-2 with the whole game in front of you. Is it that important to hit that three? It might be. It might not be. The game might be a blowout. It might be close. It might be anything. But later, you're down by two at the buzzer and you hit a buzzer-beating three, what does that mean? There's no longer any "might". You won the game. Then, in retrospect, you can say every possession mattered. But from the perspective where decisions are actually made, you don't have the benefit of hindsight, and therefore, win probability is important.
You never know if you're going to go on a ten-point run. I think everyone can agree, though, that it's better to be up by ten with three minutes left than it is to be up ten with 38 minutes left.
I understand the idea that each point becomes more valuable as the game goes on if you look at each successive point as a representation of a larger and larger percentage of the scoring opportunities that a team has left (which makes sense b/c you can't score points without at least some time going off the clock). There is a problem with looking at things that way, though, in that it treats the scoring that got you to whatever end-of-game scenario that you're looking at as faits accomplis. They aren't, of course. What happens when you're down 3-2 will affect the end-of-game scenario and therefore you're chances of winning.
The thing is, that end-of-game scenario could've been accomplished any number of ways. The closer you get to the end of the game, the more different things could've happened (which dilutes the effect of each individual one) and the fewer different things can happen. When you hit that three to go up 3-2, there are still so many things that could override it and make it meaningless.... or not. Nothing can override the buzzer-beating three.
I don't know how many more ways I can put this, but I'll try one more. Let's say you're down by four, which is two possessions. Would you rather be down by two possessions in a game with 20 possessions left, or two possessions left? In the latter, if you don't score on the next possession, you lose. In the former, if you don't score on the next possession, you still have lots of chances. Therefore, the next possession is much more important at the end of the game. And this whole debate started in the context of coaching decisions, which again, are made under this kind of thinking rather than postgame and in hindsight. So in which situation would you rather have your starters on the floor?
The only disagreement I have with what you've said above is with the idea that something could happen that would make whatever happens after 3-2 meaningless. The next event can't be meaningless. It may have very small effect on the outcome of the game or a very, very small effect, but it does matter. It gets you to the place where your buzzer-beating three ends the game instead of tying it, ties the game instead of making you a loser by one point, makes your a loser by one point instead of four, makes you a loser by four points instead of seven, etc.
Looking at it in a microcosm, I think everyone would agree that you'd rather be down four with twenty possessions left rather than two. The problem with that hypo, though, is that assumes that you're down four. How did you get to be down by four? You got to be down by four (at least in part) by your use of your players.
A side note: There's probably a graph or calculation (the creation of which would be beyond me) that could tell you how valuable time is vs. points on average. For example, am I more likely to win if I'm up 15 with seven minutes left or up four with two minutes left (on average)?
i agree with you 100%. a point is a point no matter when you get it. these dudes that overanalyze just throw out kenprom, saying a point is worth more later in the game. the problem is, this theory isn't in a vacuum. they fail to realize that team's with leads milk the clock with a lead later in the game. this reduces possessions. and the losing team becomes more desperate and takes more bad shots.
so on the surface one can say a point is worth more in the 2H. but they fail to realize that having points in the 1H can completely change the game stragegy as the game progresses.
I have to think Beilein has thought long and hard about this stance and just disagrees with your worst-case scenario. He uses the facts that all minutes are not equal and that a foul ridden player is not playing at peak aggressiveness and determines it's better for the team for them to sit.
Logical arguments, like yours, to the contrary exist and are plausible, but you're dumbing down the argument above. I'm sure we could try to quantify all of the factors to determine a best practice, but let's not treat this decision by Beilein as something that he's just too stuborn to relent on. I think he's proven during his time here that if something isn't working, he'll try something else (1-3-1 being a prime example).
On the issue of fouls, one thing that's going to worry me going into the NCAA tournament is getting some quick whistles from non-Big-Ten referees who aren't used to seeing the contact. That's always been a concern for Big-Ten teams and has hurt more than a few in the past. Especially early in games Morgan and Smotrych are going to have to be very careful.
I don't think it has anything to do with the NCAA tournament specifically - even some of the worstrefs don't seem to call more fouls in March - but more with adapting to a crew that the players might not have seen before, and doing so in a single-elimination environment. A bad call or two in January doesn't have much of an impact unless you're looking at seeding; a bad call in March can dig a hole for a team that it can't escape.
Crew, game, environment are all admittedly different
But I agree that if anything, it seems more times than not they call the game in a Big Ten way, not wanted to affect the outcome of games EVERYONE is watching, and other teams have to adapt to the more physical style of the Big Ten than the other way around.
So you're saying the refs let the big, physical team, push us around? Because we weren't the most physical team in the world, and they didn't exactly protect us. One of the reasons we are doing so well this year is even though we're not the biggest team, we not letting ourselves get out-toughed too often. In the game you cite, the more physical team was allowed to play, and beat the less physcial one. So if you were trying to bolster my point, thanks.
I think we're pretty much saying the same thing - the issue is refs who aren't used to the Big Ten style. A couple of quick whistles and you're in trouble fast in a single-elimination tournament. They just have to have the discipline to be cautious until they see how the game is going to be called. It's been an ongoing issue for both players.
And while 1.1 points per 3pt attempt isn't GREAT, 37% from 3 is actually a pretty good percentage.
These are equivalent. What are you trying to say?
"It would be a travesty, it would be ridiculous to all of a sudden come back and get the feeling back, get the health back, feel good again and then all of a sudden go throw some other colors on my shirt and go coach."
Vogrich and Douglass have completely different strengths...
Vogrich may end up a much better shooter than Stu, and his defense is improving, but he'll never be an on the ball defender like Douglass. Stu might be one of the 3 best on the ball defenders in the conference (Craft is clearly #1).
Vogrich will never have a handle like Stu either, Stu can play the point if there isn't full court pressure, Matt should never play the point.
Lastly, Stu has learned how to drive to the basket and get his own shot. He also has a nice step back jumper he displayed in the Ohio game.
Vogrich is a set shooter. We may be able to get him to run off some screens and catch and shoot. But shooting is where he's most efficient. Not at bad thing, just different players.
“True loyalty is that quality of service that grows under adversity and expands in defeat. Any street urchin can shout applause in victory, but it takes character to stand fast in defeat. One is noise — the other, loyalty.”
Vogrich and Douglass are both guards who shoot the three pretty well. The poster didn't mention Novak. Vogrich and Douglass are different players, but Vogrich could improve to be more like Douglass. It wasn't a crazy suggestion.
I don't see how anyone can be offended with that being the case. Who knows what kind of player this unholy amalgamation of Douglass and Novak would be? He could be a lot like Vogrich...though I bet he'd have Novak's eyes.
Actually, I concluded that since Stu and Zack are the 2 scholarship seniors this season and that Vogrich will be the only scholarship senior next season (who has gotten any meaningful playing time), he could be the SENIOR LEADER who does a lot of the little things that don't always show up in the box score.
Dang, you guys are so racist... /sarcasm
But seriously - senior leadership is what i was going for.
For my privacy, my new username is "non-Oriental non-Andrew"