Basketbullets: Brackets, Shot Volume, Wisconsin, Stephening Comment Count

Brian February 13th, 2018 at 2:19 PM

40180070712_f939439201_z

BRACKET MATRIX HAS US WHERE [Patrick Barron]

Brackets: large differences. Michigan is a 9 seed on the Bracket Matrix, but Joe Lunardi has a more optimistic take:

image

Lunardi doesn't pay enough attention to avoid rematches in the first two rounds, thus the potential UNC game in round two. Let's hope he's right, and not everyone else. (Everyone else is probably right.)

I don't dive into the numbers like bracket guys do, but it's completely bonkers to me that a team like Alabama is ahead of Michigan on the matrix. Alabama has Ls to Minnesota, UCF, and Texas in the nonconference and is 7-5 in the SEC. They've got a smattering of good wins but they're 43rd in Kenpom.

Sucks being a Big Ten team this year, I guess. Crashing The Dance's entirely algorithmic take barely has Michigan out of the first four—they're the #29 at large—and has them as a ten-seed. Woof. (It's missing Michigan's most recent win, but I don't think Wisconsin is moving that needle very much.)

Let's all try to not think about what Michigan's tourney profile with one of those games against Purdue in the W column.

Shot volume: we have some of it. Michigan's always been okay at getting shots up, and this year they're fairly good. Via John Gasaway:

Gluttonous	         TO%     OR%     SVI
1.  North Carolina      16.1    41.0    102.8
2.  Villanova           12.9    28.4    100.7
3.  Duke                17.2    37.4     99.8
4.  Ole Miss            15.9    32.9     99.8
5.  Florida State       16.2    34.2     99.5
6.  USC                 14.4    29.5     99.5
7.  TCU                 16.8    34.9     99.1
8.  Notre Dame          16.7    34.0     98.8
9.  Arizona State       14.1    27.1     98.7
10. Auburn              15.7    30.8     98.6
11. West Virginia       18.1    36.9     98.5

Normal                   TO%     OR%     SVI
12. Florida             14.5    27.2     98.3
13. Virginia            14.2    26.3     98.2
14. Iowa State          16.9    33.2     98.2
15. NC State            16.8    32.5     98.0
16. Michigan            13.9    25.0     97.9
17. Ohio State          15.5    28.6     97.8
18. Purdue              14.3    25.4     97.7
19. South Carolina      17.4    33.4     97.7
20. Butler              14.3    25.2     97.6

Michigan checks in 16th amongst 75 major-conference teams, largely on the strength of their TO rate, which is typical Beilein, and an OREB rate that, while last in the Big Ten, is not cripplingly low. Teams like Creighton and VT and their sub 20% OREB rates get sucked into the bottom here despite solid TO rates. And TO rate is where it's at:

Since turnovers are way more important to shot volume than offensive boards (you can’t rebound your miss if you’ve already coughed up the ball), you can generate a UNC-like shot volume with nowhere near the Chapel Hill-variety emphasis on the offensive glass. Shot volume is small-c catholic on how you get the job done, it just measures results.

Michigan doesn't focus on OREBs, exchanging them for excellent transition defense, but they're not hurting themselves on offense much by doing son.

I'd be interested to see a defensive version of this, especially to see Michigan's uptick in it over the last couple seasons.

33228616242_695667cc71_z27937215809_e4758438ec_z

Cowan and Mason are UA3 shooters [Paul Sherman/Marc-Gregor Campredon]

The Stephening. Kevin Pelton takes a look at Trae Young, the insane-usage Oklahoma guard who is the face of a vanguard of players. Curry has created a subclass of three-point launchers unprecedented in NBA history...

Curry's 3-point attempts increased only marginally to 7.9 per game in 2013-14, albeit in fewer minutes. It wasn't until two seasons later, after winning MVP, that Curry fully unleashed his full 3-point arsenal. In 2015-16, Curry zoomed past double-digit 3-point attempts, averaging an incredible 11.2 per game ... and becoming the first player in NBA history to be voted MVP unanimously.

Curry's success helped pave the way for other stars to shoot 3s more frequently. This season, his 10.0 3-point attempts per game rank second in the league behind MVP favorite James Harden of the Houston Rockets, who is attempting 10.7 per game. Each of the past two seasons, both Harden and teammate Eric Gordon have attempted more 3s per game than any player in league history before Curry.

...and naturally the college game has also noticed.

The website Hoop-Math.com has used play-by-play data to track assisted and unassisted field goals at the college level back through 2011-12. The leaderboard of unassisted 3s is dominated by the past two seasons, and Young has an excellent chance to post the highest total in that span.

Young is on pace to hit around 100 unassisted threes this season. That is on another level from Derrick Walton's ability in this department; he hit 34 last year. It's on another level from everyone, but these kinds of guys are about to be a lot more common. Maryland's Anthony Cowan, who you may remember making some very frustrating shots against Michigan earlier this year, is hitting 40% from deep despite half of his makes coming unassisted. PSU's Tony Carr is hitting 46% despite having 38% of his makes unassisted; Minnesota's Nate Mason is at 42% and on 45% unassisted threes.

When the pull-up three is a good shot that changes your late clock offense significantly; not only do you get a decent shot at three points but the defense has to respect it, opening up other things.

(FWIW, Michigan's dip in late-clock offense this year is a little about reduced efficiency from three—Walton hit 40% on 64 late threes last year; Zavier Simpson is at 33%; as a team Michigan's FG% on late threes has dipped 4 points. But it's more about an inability to get anyone to hit a jumper inside the line. Michigan is at 22% on late two point jumpers; last year they were at 36%.)

This is particularly relevant for Michigan because you can't throw a brick around here without hitting a game video of David DeJulius in which he pulls up for a three, several times.

The evolution of this from BS high school offense to something you really want to have on your team has been fascinating, and rapid.

What's going on with Wisconsin. This doesn't have a ton to do with Michigan but I found this conversation about Wisconsin basketball to be interesting all the same. This is an excellent point:

TORVIK:

I do not think that Gard is intentionally changing the Ryan system—but I do think the formula may be outdated. Not because it doesn’t work, but because everyone is using it now. Bo was at the vanguard, and used shot volume to wring wins out [for] lesser-talented teams. After about ten years of doing this, other coaches finally grudgingly admitted that it was ingenious and started doing it (namely: protecting the defensive glass and limiting turnovers) themselves. At around the same time, however, Wisconsin got super talented and put together its best sustained run ever. So we didn’t notice that the underlying formula might not work as well without elite talent.

Let me explain this a little further. You correctly point out that Wisconsin used to rank among the elite in turnover percentage and defensive rebounding percentage, and now they rank as mediocre. That is correct, but it obscures the fact their actual performance hasn’t changed much. Look at the turnover numbers:

Year   TO% Rank

2018  18.0  120

2005  17.9   22

An even starker example, using last year’s full year numbers:

Year   TO% Rank

2017  17.0   71

2006  17.4    9

In other words, a turnover percentage that used to rank among the elite, now ranks among the mediocre. Wisconsin hasn’t really gotten that much worse at committing turnovers, it’s just that literally everybody else has gotten better.

A similar thing is happening with defensive rebounding.

Michigan's managed to stay ahead of these trends. They're still 4th in TO rate nationally, and their DREBs have improved a great deal this year. Wisconsin got a sudden reality check this year. Which is nice when you're playing at the Trohl Center. But I'm not sure it's good for the league overall. Wisconsin putting together a good program comprised largely of random kids from Fond Du Lac helped the league perform without upping their recruiting. And there's really no way to up the Big Ten's recruiting without activating the Bag Man Wonder Powers that have an even bigger influence on basketball recruiting than they do football.

Who fills Wisconsin's spot in the league? Nobody, in all probability. And then you get slotted in as a nine seed despite being 20-7 with a win at MSU.

Finally. A dream too beautiful and vision-blocking to live.

Comments

stephenrjking

February 13th, 2018 at 2:40 PM ^

Seeing a piece of headline entitled "the Stephening" was, I confess, briefly troubling. Once I read closer i was annoyed, but for a different reason, because Steph Curry does not pronounce his first name correctly and I'll generally give him a pass for this but it has the potential to undo all of the benefit of proper pronunciation I've inherited from my famous namesake.

I do find that Wisconsin conversation fascinating. Might take a couple of years to see if a Wisconsin dip is real or just a transitory thing.

ST3

February 13th, 2018 at 3:14 PM ^

I have no problem with the Stephening. I also pronounce Stephen properly so it took me awhile to make the connection to Curry. I went to the dentist yesterday and got the "Steffening" from my dental hygienist. All I could do is shake my head in sadness at those living in a universe where Stephen is pronounced Steffen.

MH20

February 13th, 2018 at 3:02 PM ^

OSU has been a good-to-great basketball program for years.  The last two years of Matta were an aberration.

Maybe Penn State can replace Wisconsin.  Pat Chambers is recruiting his butt off in Philly and doesn't appear to be doing it the wrong way.

trueblueintexas

February 13th, 2018 at 3:06 PM ^

This is the problem with NCAA basketball recruiting. Is not appearing to do it the wrong way using hookers and blow or not? I don't know what the right and wrong way to recruit in basketball is anymore. Is it what the majority of programs are doing (cheating) or the minority (not cheating).

snarling wolverine

February 13th, 2018 at 3:45 PM ^

My question with PSU is, do they have the fan support to get to the next level?  Every time I watch a game there it looks like an old Piston game at the Silverdome, only their arena holds 15K and not 80K...

TrueBlue2003

February 13th, 2018 at 4:03 PM ^

is probably way premature. 

They just lost a bunch of seniors and have a couple guys hurt.  They're not unlike our 2015 team.  There are enough good players in Minnesota and Wisconsin (both states that they recruit well) for them to stay in the top third of the conference and be nationally competitive.

Gard may not be the guy to do it, but they'll find someone who is.

UVA essentially plays the same brand of basketball and is #1 in the country right now.  Not enough teams are doing it or are good enough at it (although that may change) to offset the advantages that system has. And speaking of UVA, a lot of Badger fans assumed they'd go hard after Bennett when Bo left, but he left in a way that pretty much ensured Gard the job.  If he has another one or two bad years, he'll be gone and they'll make a swing at Bennett that he might not be able to refuse.

snarling wolverine

February 13th, 2018 at 6:52 PM ^

Bennett was a good coach but outside of a miraculous Final Four run as an 8 seed, he was not at Ryan's level.  He had a losing B1G record (37-43) and went 0-2 in the NCAA tournament outside of that one run.  

I could see UW landing another guy like that.  Another Ryan is extremely doubtful.  Ryan went 172-68 in B1G and never finished worse than 4th in his 14 years.  Now, if Gard were to match Bennett's record in five years it might get him fired.

 

 

 

 

TrueBlue2003

February 13th, 2018 at 9:04 PM ^

He took UWGB from a 4-24 team to regularly winning conference titles and making the NCAA tournament - all firsts for the program.

He took over a terrible Wisconsin program that had only 8 winning seasons in the 40+ years before he arrived. In his first year they made it to the NIT, in his second year they had the first winning conference record in 23 years, they had their first ever 20-win season in his fourth year and made the NCAAs in his final three years, making the Final Four in his final full season.

He laid the ground work for Bo Ryan, and then took over a Washington State program that was ranked 208th (!!) in kenpom the season before he arrived and did this over the next 5 seasons:

2004: 126th

2005: 79th

2006: 99th

2007: 27th with a 26-8 record and OT loss in NCAA round of 32.

2008: 10th with a 26-9 record and sweet 16 loss to Final Four bound Tyler Hansbrough-led UNC.

This was at WSU!  He had that program at the level of Bo Ryan's Wisconsin by the time he left.  No reason to think he wouldn't have done just as well as Ryan had he stuck around longer.

Remember that for all of Bo Ryan's regular season success, he underacheived pretty badly in the tournament until the Decker/Kaminsky group broke through.

In any case, their previous two hires were absolute home runs.  It was extremely unfortunate at the time that Bennett left but they nailed his replacement.

B-Nut-GoBlue

February 14th, 2018 at 4:20 AM ^

*I think his kiddo was coaching Wazzu for at least that Sweet 16 season.

Otherwise yes, all valid and sort of my point with him. What snarling laid out is all factual but add the context of Wisconisn PRIOR to his arrival and hoo boy, he did a helluva job and seemed to have laid a foundation...for Ryan to continue (and continue he did!).

jbrandimore

February 13th, 2018 at 2:52 PM ^

Who wants to see a guy dribble over half court and launch a three without ever passing all day?

BORING.

I suspect it's time to move the 3 point line back. Way back - so it becomes a stupid shot in open floor situations.

TrueBlue2003

February 13th, 2018 at 3:24 PM ^

Judging by the insane popularity of guys like Trae Young, Stephen Curry and James Harden.  Most people like the faster pace that jacking up a three allows (rather than passing it around the perimeter for 30 seconds waiting for a crease or backdoor).  It's directly led to much higher scoring in the NBA and an increasingly popular product.

If there's a reason to hate data analytics and the use of them in basketball it's that teams crash OREBs much less than they used to in the attempt to limit transition opportunities.  And that teams are far less likely to make risky (but exciting) passes for fear of turning it over.  Bombing threes early in the shot clock is thankfully offsetting those trends that slow the game down.

stephenrjking

February 13th, 2018 at 6:42 PM ^

A lot of people DO want to watch that. And many of the people who bomb on it haven't actually watched that much, either.

I'm a bit tired of the NBA now, but that's because the drama is limited to which two of three or four ultra-stacked teams are going to play each other and who is going to lose to Golden State. A couple of years ago, though, when GS was up-and-coming and it was the Spurs and the (insert name of whatever team Lebron plays for) that were the establishment, it was a lot of fun. When any of the Spurs, Warriors, Thunder, or Clippers could be a legit Finals team, I was really interested.

But the style of game is fun. Yeah, sometimes guys are just jacking up threes, but it's fun and when they get hot it's crazy. Also, jacking up threes 5 seconds into the shot clock is still quicker and more free-flowing than the 2-point equivalent, which is a 24-second slog in which Carmelo Anthony clears everyone out of the lane, backs his guy down, and shoots a fadeaway from the elbow at a 55% clip. 

When they're not jacking up threes, a lot of the actual basketball on the floor is actually quite attractive. The Warriors are a classic example, but even the late Duncan Spurs actually relied on great-looking play to win titles, and Lebron's teams are guaranteed to have some beautiful plays just because he's such a good ball distributor. 

 

snarling wolverine

February 13th, 2018 at 3:53 PM ^

So by making the three-pointer much tougher, you basically want 1) a more congested game with defenses just sagging in the paint (since the three will no longer be a realistic counter) and 2) fewer comebacks, since it's harder to come back with twos instead of threes.

Why not go for the trifecta of boredom and eliminate the shot clock?

bronxblue

February 13th, 2018 at 4:59 PM ^

So what are "closed floor situations" where it would be a viable shot?  

If they move the arc back, then you'll have a year or so where the percentages will go down and then guys will compensate and we'll be back to the same approach.  3 pointers are simply mnore efficient than long 2s, and about the only more efficient shot is one around the rim, which means either (a) a return to the dump-down to your big man and then 10 seconds of him trying to do moves, or (b) lots of drives to the hoops with players looking for fouls, which is always a blast to watch and totally doesn't irritate fans.

Wolverine In Exile

February 13th, 2018 at 3:27 PM ^

All kidding aside, increasing trends like this are interesting to me. Similar things happened in the 1960's with baseball pitching where by the 1968 season, basically all pitching stats were being obliterated (http://www.espn.com/mlb/columns/story?columnist=kurkjian_tim&id=6534824) and MLB's solution was to lower the pitching mound. Within a year, offensive statistics were back in line with historical norms and stayed there for another 20 years (until PEDs moved the pendulum way over to the offensive side).

Basketball at the college ranks, if not the NBA, might be soon be making / need to make a similar move by moving back the 3-pt line, and then the team that can exploit the market inefficiency (in this case, I'm guessing the classic "elbow jumper" shooter) will put together a wining streak until others catch up. Remember, the reason the 3-pt shot has exploded is that GM's and coaches realize the math works on shooting more 3's if the percentages stay above a certain level. If you move the line back to where percentages go down on average, you're still going to have your Harden and Curry sharpshooters, but the "stretch 4" and "stretch 5" may become endangered.

TrueBlue2003

February 13th, 2018 at 3:36 PM ^

this brand of basketball is very exciting and highly popular.  Chicks dig the deep ball.  I'd much rather watch guys bomb 25 footers than 15 footers.  Especially when the threat of three pointers spreads the floor such that there are lots of shots/dunks at the rim as well.

Baseball made changes because offense puts seats in the stands.  Basketball isn't going to make a change that makes offense more difficult and less enjoyable to watch.

TrueBlue2003

February 13th, 2018 at 3:55 PM ^

evolution under Beilein.

When he was at WVU, he could run the 1-3-1 and get enough TOs to offset the easy buckets that defense gives up (both because most coaches at that time weren't doing enough to prevent TOs and Big East coaches were behind the curve).

Came to the B1G ten and ran into coaches like Bo Ryan and others that were recruiting players that take care of the ball and coaching players to take care of the ball (and more and more have since adopted a low TO approach).  His 1-3-1 was torn to shreds and he had to move to man defense because it wasn't getting enough TOs to offset the high FG% allowed. 

In his last two seasons at WVU, their TO rates were around 24%. His first two seasons at Michigan it was around 20%.  That's a huge difference.  That's about 3 more shots per game (nearly four if you assume nearly one of those misses will be OREB'd) for the opposition and three fewer transition opps for his teams.  That's about a five to six point swing. 

Six points is the difference between being ranked 50th in kenpom and missing the tourney (which M did twice in his first 3 years) and being in the top 20 and a sweet 16 contender which M has been four of the last 6 years.

Beilein has been really good adapting to changes in the game and new information.  The emphasis on preventing threes and defensive rebounding in the past two years has been another evolution.

Now if we could get an offensive coordinator that is ahead of the curve and let him do his thing, that would be greaaaat.

TrueBlue2003

February 13th, 2018 at 9:12 PM ^

they had TO rate of 22% each year.  Still significantly higher than his first two years at Michigan.  So while there is likely a slight learning curve that argument doesn't explain the difference.  Teams in the big ten were - and teams everywhere were becoming - increasingly better coached, specifically in regards to protecting the ball.  That's verifiable by the data shown in this article.

Beilein saw this.  That's a primary reason he switched defenses.  If he thought all his players needed was a couple more years running the 1-3-1, he would've stuck with it.  But it doesn't take more than two years to learn to run that zone well.  It just didn't work in the big ten and was increasingly ineffective nationally.

MH20

February 13th, 2018 at 4:56 PM ^

Above I opined perhaps Penn State but Indiana seems like the obvious answer, doesn't it?  Kind of feel silly that it didn't even dawn on me to consider the Hoosiers.  Yes, they are going to lose several key contributors* but Miller has a nice class coming in and has proven he is a very good coach.

*Is Juwan Morgan a threat to leave?  Probably not because of being an undersized 4/5 but he has been very good for IU this season.

TrueBlue2003

February 13th, 2018 at 5:34 PM ^

IU has mostly been good and should be considered in the upper tier.

The thing is, Michigan might be the answer here.  It's just that we rose to the top tier six years ago and Wisconsin stuck around for five of those years thanks to a series of fortunate hits on Kaminsky, Decker, Happ, Hayes and Koenig.

During that time the B1G ten enjoyed unprecedented dominance (in the regular season) that may not be sustainable.

The other candidate that has actually been bad for ten years but is historically good is Illinois.  If they can get the right coach - one that can recruit Chicago (it would help them a lot for Calipari to leave the college game) - they should enter that top tier.  Their more plausible path is to do it through recruiting guys that UK and Duke currently poach from the Midwest.

Indy Pete - Go Blue

February 13th, 2018 at 4:50 PM ^

SVI - another data point in basketball I had not heard of - thanks Brian!  We have the highest SVI in the B1G (If I am reading this chart correctly).  Dylan Burkhardt wrote a nice article recently on the Athletic pointing out that Michigan gets off the highest quality shots in the B1G. That means: we get he most shots up AND we get the best shots up in the B1G!  We should be making a lot more looks; just don't quite have the shooters of previous years.  Still, when they are clicking, we can be explosive! (see Purdue and Wisconsin games)

Excerpt from DB on Athletic:
Using play logging data from Krossover, we can evaluate a team's shot quality. Krossover's model relies on shot location, shot type (off the catch or dribble), play type (half court or transition) and level of defensive contest to create an expected value for every shot. Using data from Big Ten conference games only, Michigan is leading the league in expected points per shot, but ranks just sixth in performance against expectations.

https://theathletic.com/228885/2018/02/02/why-michigans-shot-selection-…

bronxblue

February 13th, 2018 at 4:51 PM ^

Credit to Beilein for adapting to what his team is capable of this year and trying to manufacture points at the end of possessions despite not having a Walton-type PG who can do it by himself.  He's definitely given the okay for MAAR and Simpson to attack the rim when the clock is running down, and that seems to be a slight change to past seasons when he wanted his perimeter players to look for a late 3.

As for Wisconsin, I agree their style hasn't changed but others have caught up.  But they've also stopped producing those dead-eye shooters in bunches like they used to.  Aleem Ford is a good outside shooter, but he's basically it.  And what made those Frank Kamisky teams so good was that you have guys like Brust, Koenig, etc. who could hit open shots.  Happ draws in defenses enough to get some open looks, but when no one else is an option teams aren't being punished.  And because nobody else can score, it makes sense to let Happ waste 15-20 seconds a possession for a shot that is worth about 1 point per possession.  I know people got worked up about Michigan not doubling him toward the end of the game, but even with the refs giving his WIDE latitude to push and hook defenders, he wasn't being all that effective.

Also, it's a minor complaint, but it amazes me that Michael Jordan, LeBron James, Tim Duncan, Larry Bird, Magic Johnson, etc. never won a unanimous MVP, but Steph Curry got that distinction despite being, at best, an average defender that year.