How Mike Smith compares to B1G PG leaders
I hear a lot of comments about "too many turnovers" and doubts about his 3 point shooting this year. Thought it would be interesting to compare him with the presumed best of the B1G point guards statistically:
Name Assists Turnovers 3pt %
Carr 77 30 .363
Dosunumu 63 41 .422
Curbelo 60 33 .200
Bohanon 64 20 .396
Trice 46 18 .438
Davison 32 7 .412
Brooks 40 19 .372
Smith 59 29 .462
Just bothers me to hear a lot of comments about too many turnovers when his ratios of Assists to turnovers is 3 to 2 better than Brooks (who gets no such remarks) and stands up very well across the best of the conference.
January 14th, 2021 at 9:13 AM ^
Sorry that should read Brooks and Smith have the same ratio of assists to turnovers even as Smith accounts for a 1.5 times as many assists.
And I also did a disservice to Smith by not including another category he dominates
Total FG%
Smith .522
Brooks .425
Dosnunmu .498
Curbelo .479
Trice .461
Davison .365
Bohanon .417
Carr .422
January 14th, 2021 at 9:14 AM ^
PG has to be the most difficult position to transfer into and play immediately. You don't just have to fill a role in the offense, you have to run it... Not every assist is the same either. Drawing a double-team and finding the open guy for an easy dunk is a lot more impactful than just rotating the ball around the perimeter until someone shoots a 3-pointer. His assists are more the former than the later.
I expect he will continue to improve his assist:turnover ratio as the season progresses.
January 14th, 2021 at 9:25 AM ^
A great point on the rotating the ball assist men and explains why someone like Davison has an unbelievable assist to turnover ratio but also is not driving in to the box to get his assists.
January 14th, 2021 at 9:16 AM ^
This is an excellent comparison to show! I don't follow basketball stats, so constantly hearing about him being a turnover machine had me thinking he was - although I didn't really see it on the court. These numbers are impressive in that now I know, definitively, that people are just nitpicking a low fruit because there's nothing else to complain about.
Mike Smith is as good as any PG in the B1G. Full stop.
(Thanks, Valenti! I'm gonna get a lot of mileage out of that...)
January 14th, 2021 at 9:31 AM ^
Mike Smith is as good as any PG in the B1G. Full stop.
Eh, the advanced stats say Dosunmu's combination of usage and efficiency make him the best PG in the B1G by a pretty wide margin. Kenpom thinks he is the 3rd best player in the country.
January 14th, 2021 at 9:41 AM ^
As the OP, I would second Dosunmu as the best PG in the B1G, but not necessarily the PG that Michigan needs on this team.
We don't need a PG that leads the team in scoring, but someone who is looking to feed our bigs and is enough of a threat to keep the defense from sagging on our bigs or helping guard the wings.
Honestly, I think Curbelo is the true point on Illinois and their hope (with a little more confidence) to remain a top 10 team.
January 14th, 2021 at 3:50 PM ^
Yes, Curbelo is the PG and Dosunmu is something else. It will be interesting to compare Smith to the Illinois guards when we play them!
January 14th, 2021 at 10:16 AM ^
Just FYI, this kind of post is why we use tempo free stats over counting stats. Counting stats are fun for ESPN to push their favorite guys, but not very useful if we want to have a more meaningful discussion on a player’s performance.
His assist rate is 27.6 and his turnover rate is 27.6. This ratio leaves something to be desired I’m not really sure how you could argue otherwise. Even the most even keeled and intelligent Michigan fan you can draw up, Dylan Burkhardt of UMHoops, has mentioned that some of these turnovers are very unforced and need to get cut down a bit. I mean Michigan is second to last as a team in conference only turnover rate. Smith is a part of why this is true. That’s not some huge attack on him like some inexplicably take it as, he’s amazing for this team. Without him the team does not work. It’s a place to look for improvement. It’s a thing to discuss when talking about how far this team can go and how they can go all the way. I don’t know how to link comments, but ctr-F “gobluem” and look for his comment further down in the thread. He pulled some great Torvik numbers showing that, compared to other Big Ten point guards, Smith’s turnovers are indeed quite high.
I see this a lot on forums where it’s taken as some HUGE insult to discuss inefficiencies in a player’s game. Smith could definitely improve this aspect of his play, but we’re not supposed to talk about that? The point of these sports discussion boards is to, like, discuss sports. That includes where improvements can be made. Michigan is a fantastic team, not a perfect team. I don’t know why it’s not seen as a *good* thing that we have a potentially correctable area to improve.
EDIT: OP legitimately went to UMHoops to ask Dylan (cool!) while completely fabricating what I said (Not cool! I did not even come close to calling him a “turnover machine”) and Dylan’s response was this, for those interested:
Smith’s turnover rate in Big Ten games is up over 35% and is the worst in the league. That needs to improve (Simpson’s was 18% in Big Ten games last year) but it doesn’t need to lead the league or anything. It just needs to be closer to 20 than 30.
This is all I’ve said here, while also explicitly stating what Dylan implicitly did by using them, that tempo free numbers tell a much more accurate story.
January 14th, 2021 at 10:29 AM ^
Let me understand this correctly. You have chosen to interpret my OP of the B1G stats as somehow calling for all criticisms of Smith to stop....but then you proceed to explain how tempo free stats are beyond repute and that some of Smith's turnovers are unforced...which is nothing Tovrik measures??......and based on this none of us should question that Smith is a turnover machine. Got it. Where do I go to hand in my MgoBlog card?
January 14th, 2021 at 10:31 AM ^
Wow are you arguing in bad faith here.
A) yes I am appealing to the most respected Michigan basketball writer out there. This is because he is more intelligent in basketball matters than the this board combined. The writers here link to him often enough that they seem to respect him quite a bit too.
B) No, tempo free statistics are not God Himself, I never remotely said they are beyond refute, but they are 50x more valuable than per game stats and render the latter practically useless aside from letting newcomers digest stuff more easily.
C) Smith has something he can improve on and you are mad that people discuss it. Thus my inference.
January 14th, 2021 at 10:48 AM ^
Let's see perhaps when you stated:
This is why we use tempo-free stats here.......
....because Smith could definitely improve this aspect of his game, but we’re not supposed to talk about that?
As OP, I wanted to be the first to offer up a counter-argument....not suggest anyone cannot debate what I put up and then by the way stated it is an open topic how he rates among the other PGs in the conference.
January 14th, 2021 at 11:05 AM ^
You and Don have a cup o' coffee this morning? You both seem to have woken up ready to shit on things people say...
How do you compare this statement: "...people are just nitpicking a low fruit because there's nothing else to complain about", with your "That’s not some huge attack on him, he’s amazing for this team. It’s a place to look for improvement."
Seems like they're the same thing - but there is a difference. The OP made a claim that Smith is working out quite well for the team as compared to other PG's around the conference. What you said, though, is that he is the problem with the team - he is the cause of bad stuff, but it's okay because he can do other things.
Do you see the difference? One guy was positive, the other wasn't. One guy said 'he's pretty good, makes a couple of TO's, who cares'. The other guy says 'you're not telling everyone how bad he is at that!'.
You make some good points, and have clarified some details. But I read your post as a smarmy statement about how much smarter you are than everyone else. Instead of building up a player, you decided to tear him down in a backhanded manner.
January 14th, 2021 at 11:15 AM ^
The fact that you interpret a discussion of a player’s turnover rate (actually much more a discussion about why tempo free stats are preferable to per game stats), while even including my comment that he is a great and essential player for a great team, as “tearing him down” is quite literally my point with that portion of the comment, and knowing that it was coming is why I chose to include it despite it not being related to Mike Smith’s play or tempo free stat debate. A fairly notable subsection of fans find it offensive to even have this discussion. Which is a shame.
I mean... you seriously took my scathing remark of “his turnover rate leaves something to be desired,” as a “tear down of a player. You are exactly what I’m talking about lol.
”What you said, though, is that he is the problem with the team.”
Apologies for my crassness, but I don’t even know how to responsd to such bullshit. And there is zero chance you typed that without knowing exactly how disingenuous it was.
January 14th, 2021 at 11:45 AM ^
Talk about tone deaf... I get that his TO situation isn't the best, it's not a life threatening thing but it can certainly be improved. You ably explained that because you're so smart about stats. But that isn't the question here. It's not what you said - it's how you said it! Re-read my last paragraph... You're so focused on being the victim that you can't even see what I wrote.
January 14th, 2021 at 11:28 AM ^
Don't forget, he's young as well as smart. Not like those old guys who don't know how to stat.
January 14th, 2021 at 11:35 AM ^
If you’re like OP, think tempo free stats are “unfair,” and argue against them in favor of per game numbers then no. You have no idea “how to stat.”
You certainly don’t need to know “how to stat” to enjoy basketball, but you can’t get mad when people point out that you aren’t “statting” correctly if you try to, like this post.
Not sure why you think age is an excuse or whatever you’re getting at. John Beilein made his entire brilliant career off of being a genius at manipulating the tempo free four factors to how he saw fit. He was light years ahead of his time with seeing the inefficiency of the long two and the advantages to lighting it up from three. It really shouldn’t be a generational thing.
January 15th, 2021 at 8:33 AM ^
Nah, it's none of that. It's your condescension and hubris.
January 14th, 2021 at 9:27 AM ^
From what I have seen, many of his his turnovers are ones where he sees something that could be a highlight, something that many wouldn't see, and he tries to make a play.
He's one of those that I would hate if he weren't on my team, because he isn't always going to dazzle with athleticism or play like Westbrook, but he is quite quick and his bball IQ allows him to control the flow and the offense.
January 14th, 2021 at 9:30 AM ^
He’s a Columbia student. His IQ is off the charts!
January 14th, 2021 at 9:56 AM ^
Isn't he a Columbia graduate? An Ivy League degree is pretty special...
January 14th, 2021 at 11:39 AM ^
He's a Michigan student now...
January 14th, 2021 at 9:29 AM ^
He’s an excellent decision maker and rarely ever forces a shot. He does force passes sometimes leading to a few bad TO but he makes it up with his assists. He’s great at drawing defenders off of Dickenson and the dishing it to him for some easy baskets. Overall he’s been the most surprising part of this team IMO
January 14th, 2021 at 9:31 AM ^
Question for the board. With the NCAA allowing players a sixth year due to covid does anyone think Smith will come back next season? From what I understand he’s going to graduate school at Michigan so it seems like a legit possibility
January 14th, 2021 at 9:39 AM ^
I can see Smith sticking around and Zeb having a Brooks-like role next year.
January 14th, 2021 at 9:54 AM ^
Brown is having the best year of his basketball life and I want him to come back but he'll probably leave to chase his hoop dreams. #ThreadJack
January 14th, 2021 at 11:36 AM ^
It’s gonna be difficult for both to come back. Our scholarship limit is maxed for next year. If Dickenson and Wagner leave early then they can both come back. The fact that I am discussing this has me excited for years to come. Michigan is a basketball school!!
January 14th, 2021 at 9:37 AM ^
Unpopular opinion, but I feel that Smith might be an upgrade from Simpson. Don't get me wrong I loved Simpson's hook and his solid defense, as well as annoying my wife by singing X gonna give it to you every time he scored. With Smith, I see a better free throw shooter, three point shooter, and maybe an equally as good defender. Smith also made some jaw-dropping passes against Wisconsin as well.
January 14th, 2021 at 9:42 AM ^
I think his hair is better
January 14th, 2021 at 9:45 AM ^
So hard to compare, right? X also was an incredible floor leader, warrior, etc. But if one is to compare (which Coach Harbaugh eschews), Mike is also a better 2-point jump shooter. Mike is just much more of an offensive threat IMHO which may open up for possibilities for others, too.
January 14th, 2021 at 9:48 AM ^
May be unpopular because of how successful Michigan was with Simpson and how much fun he was to watch, but I think most people would take a PG who can shoot and isn't a defensive liability over a guy who is a great defender and can't shoot. Almost no team is successful without a PG who can shoot. Rondo-led teams (surrounded by all-stars) and the 76ers (Simmons is 6'10") being the only ones I can think of off the top of my head.
This team needs a PG who can shoot more than it needs a PG who can be a top 5 defender. Wagner, Brown, Livers, and Brooks are all ++ defenders. Even Dickinson is looking like it as well. Having 1 more ++ defender isn't worth the offense getting bogged down because the PG can't shoot outside of 5 feet.
January 14th, 2021 at 10:25 AM ^
The foul shooting too. Simpson was such a liability there. It was nerve-racking having him shoot fts particularly in close games.
January 14th, 2021 at 9:50 AM ^
It is closer than a lot of people would like to admit. The only areas where they greatly differ are steals in Simpson's favor (though he was never the best at stopping a big driving guard either) while 3 point % and free throws are much to Smith's favor. I think it is still early to make a call.
Total FG% 3% FT% A TO Steals Turnovers/game TOs/min.
Smith .522 .462 .826 59 29 4 2.7 .089
Simpson .476 .360 .574 236 96 30 3.2 .095
January 14th, 2021 at 10:08 AM ^
I will say that while Simpson's 3 pt average was decent in his last year, he needed the wide open look to get it down. Smith seems to be able to get a 3pt shot off quicker which adds to his offensive ability.
January 14th, 2021 at 9:52 AM ^
I agree. Simpson ran the offense extremely well, but wasn't much of a threat offensively. Smith seems to be able to score from just about anywhere, which really helps stretch the floor.
January 14th, 2021 at 9:56 AM ^
+1 for DMX
January 14th, 2021 at 10:38 AM ^
And now I have DMX stuck in my head. Which is fine.
January 14th, 2021 at 11:27 AM ^
I think you could make an argument that Smith is a better fit for this team this year. But I don't think he's close to Simpson as an individual player. He is a better player off the ball than Simpson off the ball and obviously a better shooter. But Simpson was asked to do alot more than Smith is because the talent around him was worse.
Dickinson > Teske, healthy Livers > unhealthy Livers, healthy Franz > unhealthy Franz, Sr Brooks = Jr Brooks; this year's bench > last year's bench
January 14th, 2021 at 12:41 PM ^
I agree Smith does a few things better than Simpson and some things not quite as well. We are doing great because everyone, in totality. is playing at a higher level, an extreme high level. Adding the two senior transfers, Chaundee and Smith really was a stroke of genius. Hats of the Coach Howard and his staff.
January 14th, 2021 at 9:42 AM ^
too many turnovers when his ratios of Assists to turnovers is 3 to 2 better than Brooks
Brooks has limited sets where he runs the point, a position has a lot of assist opportunities with P&R action as well as in transition where we've seen Michigan get out into their fast break and make 1 long lead pass to an open 3 (sometimes to Brooks specifically).
Brooks' AST:TO is 40:19, so almost the exact same ratio as Smith and according to Torvik, Smith's TO% is 7.5 points higher than Brooks'.
Not saying Smith is a walking turnover, but he is taking over for Simpson, who averaged under 20% across his career (his assist rate was over 40) so any average TO% will look a lot worse than what we're used to.
Also, people complain about Brooks and TOs. Thinking back to the Oregon game last year when he was dribbling in the backcourt with his head down and got the ball stolen for a layup when Michigan was leading and looking to end the game in regulation. He does those things from time to time. The thing is, again, his usage is lower so his raw turnovers are lower. His 1 or 2 turnovers don't really stick in our minds because Smith has 4 or 5 and it's hard to remember that Smith had the ball in his hands 20% more than Brooks did.
January 14th, 2021 at 9:43 AM ^
Did all 7 of Davison's turnovers come against Michigan the other night? ?
January 14th, 2021 at 9:51 AM ^
It's useful to look at these things with rates, to account for usage and tempo - that way it's much more apples to apples
I just plugged Big 10 players with a >25 assist rate into Torvik, as a quick way to pick out good guards in the conference in terms of assists. That spit out 11 players.
Mike Smith has the worst turnover rate (27.6) by far among those players, with only Curbelo from Illinois coming close with 24.6. The rest (Jacob Young, Conner McCaffery, Trevion Williams, Dosunmu, Bohano, Banton, Watts, Buie, Carr) range from 12-18.
If you back that up to high majors, Smith is the second worst player nationally with an assist rate >25% (n=59)
So yeah - Smith is turnover prone
That said, he's been amazing for us and he has been a godsend, especially after we were worried about the position due to depth and Simpson leaving
January 14th, 2021 at 9:57 AM ^
I disagree that your Tovrik comparison is apples to apples. Based on choice of players and statistical measures. First, McCaffery is not Iowa's point guard, and in what orchard is Trevion Williams the same apple as Mike Smith? As another poster pointed out earlier...this should compare guards whose primary role on the team is to penetrate and find the open teammate (that would be apples to apples).
Don't get me wrong...I don't think Smith is the next John Stockton, but I am trying to argue that he is a great balance between a good outside shooter (that gives us the Miami Heat spacing that Howard seeks) and a good penetrator/passer when required.
As for statistical measure. You are citing a TOR ratio that measures turnover per shots taken by the player. Unfortunately, we do not know the time on ball for players in college so the TOR uses a substitute for time on ball based on shots taken (FG or FT). This means a guy like Mike Smith that rarely shoots (he so far has 50% less shots on goal or FT as Zavier) unfairly ends up with a much higher TOR %. This likely explains how a guy with 2.7 turnovers per game (versus Zavier's 3.2) ends up with a TOR that is twice as high as Zavier's that senior year.
January 14th, 2021 at 10:24 AM ^
You are kinda living in the past here, man. There’s no reason to isolate “point guard,” they’re all playmakers. Stauskas wasn’t the “point guard” of 2014, but he was the lead guard. Positions are way more fluid than they were 20 years ago. Jokic is the offensive initiater of a damn NBA team.
Counting stats are fun for handing out post season awards, but practically useless for meaningful analysis. The thing he said was apples to apples was using tempo free stats over the bad “per game” stats. I do mean no offense by this, but disagreeing that this is the proper way to compare players shows a lack of knowledge regarding basketball lol. Go on UMHoops and tell them that it’s unfair to use tempo free and see what happens. This just is not the right way to look at player stats and tells us almost nothing about their true ability.
January 14th, 2021 at 1:21 PM ^
Well made points on a variety of fronts. Basketball is such a fluid game and while stats are useful in understanding "what" happened when it comes to explaining the "whys"-not so much.
Comparisons between Simpson and Smith keep coming up. The fact is that both function/ed quite well within their individual skill-sets. Choosing between the two is mostly a matter of personal preference. And because we can never know how direct comparisons with identical teammates would play out it is silly to draw any hard and fast conclusions here.
I love basketball...played and/or coached five different sports competitively...for me, basketball is the #1 sport to play-and it's not close. I played and coached the game way back in the "eye-test" days. This Michigan team passes most of the eye-tests quite well...am pretty sure they are exceeding everyone's expectations. Incredible coaching.
January 15th, 2021 at 4:17 AM ^
This fuggin boomer did. Laughed out of the room. Incredible patience shown on that thread though.
January 14th, 2021 at 10:36 AM ^
I stripped out the big guys (limited the height to 6'4" since Torvik has Pure PG and Shooting PG that makes it hard to get a view of all of the PGs at once) and Smith is still the worst by 9 points. I limited it to just Big Ten play and his TO% actually went up to above 30. AST% he's top 5, but he is a turnover problem. Not enough for anyone to be complaining about him as the reason for a loss (because they haven't lost!) or anything, but on a team that is currently 13/14 in turnover% in conference, he leads the way.
There aren't many problems with the current team, but his turnovers are a problem if we had to pick one out.
January 14th, 2021 at 10:56 AM ^
Since we are not on UMHoops, I apparently am still able to use discrete quantities as measures.
Therefore, I would like to throw out there that Mike Smith has precisely 2.7 turnovers per games played for UM, whereas Simpson had 3.2 per game.
Now I am quite aware of the usage differences (Simpson played much longer periods in most games than Smith is doing)...so I then decided to calculate turnovers per minute on the court. Here are those results:
Smith 29 TOs / 325 minutes = .089 turnovers per minute
Simpson 96 TOs / 1,010 minutes = .095 turnovers per minute
but I am just getting to a simple point. How is someone---charged with leading our offense when on the court----a "turnover problem" with just 1 turnover for every 10.5 minutes on the court.
January 14th, 2021 at 11:43 AM ^
Again, IF we had to pick out a problem on the current team, it would probably be the guy with the highest TO% among Big Ten PGs who plays significant minutes. Another would be Center depth behind Dickinson because Davis is a defensive issue (an his injury could flare back up) and Johns is a small ball 5. Those are the only real problems but you don't see people panicking about any of them really. It's ok to point out minor deficiencies. No one is attacking Smith or calling for his benching.
Your raw stats are useful to an extent, but you aren't using adjusted numbers based on usage and whatnot. I'm not a statistician, but the weighted numbers for Smith gives us more information about his TO and AST output.
You used Simpson's senior year for your TO/minute numbers, which just so happened to be his worst TO year. If you take his 3 year starting totals, he averaged 0.069 (nice) TO/minute. His junior season he averaged 0.059!
Simpson's AST/minute across his 3 years was 0.18 (the same as Smith's this year). Individual seasons Simpson had 0.23 (2019) and 0.19 (2018) AST/minute. Plus, that was Simpson's sophomore-senior seasons while Smith is a grad transfer.
Simpson's senior year increase to his TO% was actually in large part due to the beginning of the season. His TO rate was 21.2 but in conference it was 18.1 and against top 50 teams it was 18.3. Meanwhile Smith's rate has gone up in conference (small sample size).
So, yes Smith is the "turnover problem." His per 40 minutes and per 100 possessions TO rate is significantly higher than any other starter and the closest guys to him are Johns (who struggles putting the ball on the deck) and Davis (who... I mean that's expected).
We can hold 2 thoughts in our heads at the same time. He has been a revelation coming from the Ivy League to leading the Big Ten's best team (to date). He has also had a bit of a turnover problem that isn't catastrophic, but could be cleaned up a bit. If you watch the turnovers, a lot could be eliminated by just not driving to nowhere and leaving his feet. Throwing a pass to Dickinson on a P&R that is just a bit too high or gets tipped by a 6'11" guy is fine. Can't really avoid those. But there are avoidable ones that he can clean up and I think will as time continues.
January 14th, 2021 at 1:14 PM ^
I agree that it is important to hold 2 thoughts in our heads at the same time. But perhaps one such thought is that 3 turnovers per game for a point guard in the Miami Heat offense is on par with how most B1G and Miami Heat guards would perform in that position. And as you confirmed above, it is what Zavier averaged over his 5th year (my choice of the comparison with the 5th year player Mike Smith).
January 14th, 2021 at 3:03 PM ^
Here's a hypothetical:
I miraculously got on a D1 basketball court as a 40 MPG starter. I suck, so all the team does is stick me in the corner on offense and hide me so I don't fuck up. However, twice a game the ball ends up in my hands, and I turn the ball over every time it touches my hands. But that's only 2 TOs per game, or 1 TO every 20 minutes by your metric.
The starting PG on that team ends up bearing the brunt of me being absolutely useless. As a result he takes almost every shot or assists every shot that the rest of our teammates take. But with all the times he touches the ball, he ends up turning the ball over 3 times per game, while playing 30 minutes. Or once every 10 minutes by your metric.
So who's more TO prone? Me or the PG? By your metric it's the PG because he turns the ball over twice as often as me. But common sense says it's me, because obviously.
That's why we use TO rate instead of any per game or per minute stats.