Friday, November 4, 2016

Returning Minutes and Number of Players Who Were Former Top 100 Recruits

In our SI projections, we project every player and lineup to get our team projections.

But I still get lots of requests for a list of returning minutes. That isn't a direct input into our model, though it is something I can easily calculate with the roster data I have.

I cannot say that these numbers will be 100% accurate. We typically don't pull walk-on data unless those walk-ons are expected to play a lot. And we have to make some decisions about certain players. For example, this assumes Coastal Carolina's Shivaughn Wiggins will be able to return in the second semester. But it should be mostly accurate.

I also list the number of RSCI Top 100 recruits on each roster. This includes current RSCI Top 100 freshmen, former RSCI Top 100 recruits (who are now sophomores, juniors, and seniors), and players that we think were incorrectly ranked by RSCI because they changed classes.


Wednesday, October 12, 2016

Comparison of 2016 Preseason Rankings

(Quick note: This column is talking about last year's preseason rankings. The 2016-17 preseason rankings will be released soon.)

Ever since I partnered with Luke Winn at Sports Illustrated, we’ve been doing something rather unique when it comes to projecting the college basketball season. We project every D1 player, project every D1 lineup, and use those lineups to project every D1 team. I tend to think this is a unique and worthwhile exercise regardless of the accuracy, but every year we get questions about how our model has done in the past.

This year, we wrote a column that shows that we think we have had the most accurate projections for two years in a row. That said, as anyone who knows about statistics will tell you, there are often different ways to spin results. Our approach is to judge the preseason rankings based on the final ranking of teams 1-351 based on margin-of-victory (MOV). If you focus on NCAA tournament bids, NCAA wins, or conference wins, one of the other models may beat our model. But our feeling is that since the season-long MOV does a good job predicting those other outcomes, it is the best way to evaluate the rankings.

To judge the models I simply took the absolute value of the difference between each team’s preseason ranking minus the team's final MOV ranking and added up the absolute error for each model. (Taking the sum of the squared errors produced the same ordering of the various ranking systems.)
The column linked to above also highlighted some of the teams where the model did better than the other models and the teams where the model fell short. But I was asked on Twitter for a full comparison of all five preseason rankings from last year. In the interest of transparency, I list all five preseason rankings from last year below.

The first thing you will notice when you look at the full list of teams is that there were plenty of teams that surprised everyone. College basketball players are at a developmental stage of their career, and we only have a small sample of useful statistics, so not surprisingly there are positive and negative surprises every season. Still, all of the models meaningfully improved on simply running the final rankings from the previous year.

One final comment, the Final 2016 MOV 1-351 ranking is based on Ken Pomeroy’s final 2016 ranking as was on his website from April to August. I did not update this analysis after he recently made the decision to tweak his formula heading into this year.

Tuesday, February 9, 2016

Big Ten Player Development

I started this series to try to show the uniqueness of players like Colorado's George King and TCU's Vladamir Brodziansky. I look at the box scores every week and continue to be surprised to see a three-star bench-warmer (King) become a breakout star and to see a JUCO transfer who was not highly publicized (Brodziansky) lead his team. 

I also started this series to check in on our impact freshmen and breakout scorers lists from For that reason, I wish I had time to show all the high major leagues we previewed at SI. But due to time constraints, I am going to wrap this up with the Big Ten. Part 6 of the series looks at every player in the Big Ten playing roughly 20% of is team's minutes when active.

Big Ten Freshmen
PPG, Min%, Poss%, ORtg, Player, Team

15.9, 82%, 27%, 98.8, *Corey Sanders, Rutgers

12.9, 54%, 29%, 115.9, *+Diamond Stone, Maryland

12.3, 70%, 24%, 108.1, +Ethan Happ, Wisconsin

12.1, 56%, 22%, 125.1, *+Thomas Bryant, Indiana

10.7, 59%, 24%, 105.5, Jordan Murphy, Minnesota

9.9, 70%, 25%, 94.6, *+JaQuan Lyle, Ohio St.

9.8, 66%, 24%, 94.1, *+Caleb Swanigan, Purdue

9.8, 59%, 19%, 103.7, *+Jalen Coleman, Illinois

9.1, 50%, 19%, 127.4, Dererk Pardon, Northwestern

9.0, 61%, 17%, 116.7, Aaron Falzon, Northwestern

8.8, 55%, 20%, 113.7, Michael Finke, Illinois

8.6, 56%, 22%, 104.5, *+Glynn Watson, Nebraska

7.8, 45%, 19%, 117.6, *+Deyonta Davis, Michigan St.

Players marked with a * were consensus Top 100 recruits. Players marked with a + were players we picked as instant impact freshmen at

I am still kicking myself about our SI Corey Sanders projection. The stats model loved Sanders to be a major scorer in the preseason. He was Rutgers best player by far. But our SI scouting said that Sanders was attempting to be a pass-first PG. And that might have been true at first. But there is only so many times you can pass someone the ball and watch them miss a wide open shot. At some point in the season, the team just decided to give Sanders the ball and get out of the way, as seen by his team freshman record 39 points in a recent game.

And even if the scouting led us astray on Sanders, scouting has substantially improved the projections. Despite the fact that he was just a 3-star prospect, Luke Winn's personal scouting said that Ethan Happ was going to be a star. And my limited scouting moved Minnesota's Jordan Murphy up in the preseason. Murphy's recruiting rank suggested he would barely play, but I at least moved him up into Minnesota's Top 7. So even though the scouting caused us to miss on Sanders, if you look across the board at all the teams, 95% of the time scouting made our projection better. A few other notes:

-The big question for Indiana's Thomas Bryant wasn't whether or not he would score. The question was whether or not he could improve Indiana's defense. And he has delivered as Indiana's two point defense has improved from 283rd nationally to 137th. Overall, Indiana's defense has improved from 214th to 49th.

-A lot of people thought this might be the year that Northwestern finally made the NCAA tournament, but in a down year in the Big Ten, their performance isn't quite good enough. And distressingly, a lot of points will likely graduate along with Tre Demps and Alex Olah. But that's why Dererk Pardon's story is so awesome. His 28 point, 12 rebound performance soon after debuting mid-season bring a lot of hope for next year.

-Illinois' Michael Finke was a 3-star red-shirt player, the kind you might not expect that much from. In fact, among Illinois forwards he had the 6th best statistical expectation when I first ran the team's numbers this summer. But as three other big men left the team or were injured, he moved up. Finke's development is one of the only things that has gone right for Illinois this year.

PPG, Min%, Poss%, ORtg, Player, Team

6.7, 54%, 17%, 96.3, Jonathan Jean Laurent, Rutgers

6.7, 45%, 25%, 85.9, Kevin Dorsey, Minnesota

5.7, 56%, 20%, 84.6, Josh Reaves, Penn St.

5.1, 37%, 15%, 124.9, Nicholas Baer, Iowa

4.9, 53%, 19%, 83.8, Dupree McBrayer, Minnesota

4.9, 41%, 17%, 106.7, Jack McVeigh, Nebraska

4.5, 41%, 14%, 121.2, Michael Jacobson, Nebraska

4.2, 29%, 17%, 118.1, Ogugua Anunoby, Indiana

4.1, 46%, 16%, 86.4, *Daniel Giddens, Ohio St.

4.0, 31%, 16%, 106.5, Aaron Jordan, Illinois

3.9, 41%, 14%, 113.2, *Matt McQuaid, Michigan St.

3.8, 34%, 15%, 113.4, Ryan Cline, Purdue

3.8, 34%, 17%, 98.9, *Ed Morrow, Nebraska

3.5, 35%, 15%, 104.6, Deividas Zemgulis, Penn St.

3.1, 28%, 18%, 93.4, Charlie Thomas, Wisconsin

3.0, 35%, 15%, 90.9, Khalil Iverson, Wisconsin

2.9, 33%, 17%, 88.5, *A.J. Harris, Ohio St.

2.9, 22%, 19%, 100.4, *Victor Wagner, Michigan

2.8, 24%, 13%, 128.2, Ahmad Wagner, Iowa

2.8, 27%, 14%, 118.1, Alex Illikainen, Wisconsin

2.0, 37%, 9%, 96.3, Justin Goode, Rutgers

1.9, 25%, 12%, 111.3, Kenny Goins, Michigan St.

1.8, 22%, 18%, 82.5, Mickey Mitchell, Ohio St.

1.7, 22%, 17%, 70.5, *D.J. Williams, Illinois

1.5, 20%, 11%, 102.5, Juwan Morgan, Indiana

Four names I am not listing above: Wisconsin's Brevin Pritzl, a Top 100 recruit forced to redshirt due to injury; Ohio St.'s Austin Grandstaff, a Top 100 recruit who is transferring; Penn St.'s Mike Watkins, a 4-star prospect who was not eligible, and Rutger's Ibrahima Diallo, who played quite a bit early, but was injured.

Big Ten D3 Transfer

PPG, Min%, Poss%, ORtg, Player, Team

12.1, 70%, 18%, 132.0, Duncan Robinson, Michigan

There are not many players that transition from D3 to D1, so I can't really show you comps to explain how special Duncan Robinson's season has been. But there is no question he is very unique. The experts were right that he is a lights out three point shooter. But I'm not quite sure they were right about his ability to play in the paint. Michigan's interior defense has been pretty poor this year and teams expose Robinson inside when they can. I think Michigan views him as a stretch-four, but I think he would be better off as a true-wing. Or perhaps he will be a better interior defender and rebounder with another year of bulking up in the weight room.

Big Ten JUCO

PPG, Min%, Poss%, ORtg, Player, Team

2.1, 19%, 15%, 114.4, Jaylen Brantley, Maryland

No debuting JUCOs have played a big role in the Big Ten this year. Rutger's DeShawn Freeman and Iowa's Dale Jones might have made a bigger impact but they both got hurt.

Big Ten D1 Transfers

PPG, Min%, Poss%, ORtg, Player, Team, LastPPG, LastTeam

17.2, 72%, 24%, 123.3, Andrew White, Nebraska, 2.6, Kansas

13.1, 65%, 24%, 116.7, Robert Carter, Jr., Maryland, 11.4, Georgia Tech

10.9, 80%, 18%, 114.1, Rasheed Sulaimon, Maryland, 7.5, Duke

9.7, 51%, 25%, 110.6, Eron Harris, Michigan St., 17.2, West Virginia

7.9, 43%, 20%, 123.7, Max Bielfeldt, Indiana, 5.1, Michigan

7.9, 52%, 18%, 107.4, Omari Grier, Rutgers, 7.9, Bradley

6.5, 42%, 19%, 113.0, Trevor Thompson, Ohio St., 5.0, Virginia Tech

5.3, 45%, 18%, 114.9, Johnny Hill, Purdue, 9.6, UT Arlington

4.1, 33%, 20%, 96.3, Joey van Zegeren, Northwestern, 9.8, Virginia Tech

3.5, 56%, 14%, 88.9, Khalid Lewis, Illinois, 5.8, La Salle

1.8, 29%, 10%, 100.4, Alex Austin, Illinois, 7.0, Eastern Illinois

Andrew White was expected to be a major scorer for Nebraska, but he has done more than that. He has become the efficient superstar Nebraska needed to have a top flight offense again. (According to, this is the best Nebraska offense since 2004.) Unfortunately for Cornhusker fans the defense has fallen off. The best news I can give is that Nebraska's two point defense has been better than its three point defense and free throw defense. And two point defense tends to have more predictive power going forward. (FYI: the same 2Pt/3Pt/FT defense split holds true for Wisconsin, Ohio St., and Penn St.)
I saw people in the preseason saying that Rasheed Sulaimon could average 15 to 20 PPG. And of course he could if he was the only good player on his team. The problem is that on a good team, shots are at a premium. That is why my model tries to account for the number of possessions available to each player. Sulaimon has not been a statistical disappointment; he is just one part of an offense with a lot of good players.

I am not listing Mike Thorne from Illinois due to his injury, but the play of Illinois' two guard transfers makes me think the team really does miss Tracy Abrams. Khalid Lewis and Alex Austin are just not as versatile. The losses at forward and point-guard have clearly caused Illinois to drop from a bubble team to a team that is just hanging on.

Finally Playing
This group of players were technically on the roster last year, but they didn't play meaningful minutes.

PPG, Min%, Poss%, ORtg, Player, Team

10.2, 71%, 22%, 99.2, Payton Banks, Penn St.

8.6, 57%, 22%, 102.1, Vitto Brown, Wisconsin

7.5, 79%, 13%, 115.8, Zak Showalter, Wisconsin

5.0, 50%, 14%, 102.9, Bakary Konate, Minnesota

4.8, 44%, 16%, 104.6, Julian Moore, Penn St.

4.3, 51%, 13%, 103.2, Devin Foster, Penn St.

3.7, 34%, 17%, 106.1, Javon Bess, Michigan St.

3.5, 19%, 24%, 105.2, D.J. Wilson, Michigan

3.3, 26%, 14%, 112.4, Brady Ellingson, Iowa

3.0, 27%, 17%, 107.1, Gavin Skelly, Northwestern

3.0, 36%, 13%, 101.3, Jordan Hill, Wisconsin

2.6, 18%, 20%, 111.1, Alvin Ellis III, Michigan St.

2.4, 20%, 18%, 102.2, Nick Fuller, Nebraska

2.3, 27%, 16%, 94.7, Jake Hammond, Nebraska

1.9, 23%, 19%, 69.9, Isaiah Washington, Penn St.

Banks and Brown are the leading scorers in this group, but I'm not convinced either is playing well. Banks misses way too many threes and Brown is neither the outside shooter nor defensive rebounder that Wisconsin could really use.

Returning Players
Once again, when interpreting changes in PPG, you should keep in mind the changes in pace and opponent defense. Indiana's schedule has been easier defensively this year than it was on the whole season last year. Meanwhile, Rutgers games have had over 6 more possessions than their games had on the full year last year. Obviously that is going to inflate scoring totals.

ChOppDef, Team, ChRawPace

3.4, Indiana, 3.2

2.6, Rutgers, 6.4

2.5, Michigan, 5.7

2.5, Minnesota, 0.2

2.4, Northwestern, 4.8

2.2, Wisconsin, 5.1

1.8, Michigan St., 4.5

1.8, Nebraska, 4.3

1.5, Purdue, 4.4

1.4, Iowa, 5.4

1.3, Illinois, 4.8

0.9, Penn St., 0.6

0.7, Ohio St., 2.2

0.6, Maryland, 2.5

Here are the biggest changes in PPG:

ChPPG, ChMin, ChPoss, ChORtg, Player, Team, Class, CurrentPPG

8.4, 15%, 4%, 19.8, Peter Jok, Iowa, Jr, 15.5

7.9, 52%, 1%, -6.4, +Keita Bates-Diop, Ohio St., So, 11.7

7.5, 17%, 8%, 10.7, +Brandon Taylor, Penn St., Sr, 16.6

6.0, 23%, 3%, 19.3, Tai Webster, Nebraska, Jr, 10.0

6.0, -1%, 5%, 6.6, Jarrod Uthoff, Iowa, Sr, 18.4

5.7, 17%, 5%, 7.0, Mike Williams, Rutgers, So, 12.1

5.7, 3%, 4%, 12.2, Bryn Forbes, Michigan St., Sr, 14.3

5.6, 11%, 3%, -0.1, Kendrick Nunn, Illinois, Jr, 16.7

5.2, 22%, 3%, 31.2, Dominique Uhl, Iowa, So, 7.3

4.7, 8%, 7%, -13.6, Nigel Hayes, Wisconsin, Jr, 17.1

4.7, 14%, 3%, -11.4, +Bronson Koenig, Wisconsin, Jr, 13.4

4.5, 11%, 3%, -3.1, Malcolm Hill, Illinois, Jr, 18.9

4.4, 15%, 4%, -1.1, +Shep Garner, Penn St., So, 13.6

4.4, 22%, 1%, 1.3, Maverick Morgan, Illinois, Jr, 7.0

4.1, 22%, 2%, 6.9, Anthony Clemmons, Iowa, Sr, 8.9

4.1, -2%, 4%, 9.6, Denzel Valentine, Michigan St., Sr, 18.6

Players marked with a + are players we projected as breakout scorers at With DJ Newbill graduating, Brandon Taylor seemed like a good bet to increase his workload and he has. Taylor has increased his usage rate more than any other player in the Big Ten. Nigel Hayes has the next biggest increase in usage, but unlike Taylor, Hayes has seen his efficiency plummet now that he has a larger role in the offense.

There are not a ton of players in the Big Ten who have seen big improvements in efficiency (at least relative to the other top 5 conferences). But most of the big improvements are on Iowa. Dominique Uhl has had the biggest jump in efficiency in the Big Ten, Adam Woodbury (next table) has had the third biggest jump in efficiency in the Big Ten, Peter Jok has had the fifth biggest jump in efficiency in the Big Ten along with a big jump in usage, and Jarrod Uthoff has seen his efficiency improve slightly despite the third biggest jump in usage in the Big Ten. Even Anthony Clemmons is passing better and finishing better around the rim.

Don't let anyone tell you we should have seen this coming with the Hawkeyes because they have a veteran team. Iowa had a veteran team last year and they didn't become elite. The difference this year is that Iowa's players have put in the work and improved across the board.

I am a little nervous about Iowa's defense going forward. Their three point and free throw defense looks a little unsustainable given that they are not a dominant team at defending two point shots. But that is just my way of saying I am not sure Iowa is the #1 team in the country (the current rank). Because of massive offensive improvements by multiple players, Iowa is miles better than their preseason bubble status.

ChPPG, ChMin, ChPoss, ChORtg, Player, Team, Class, CurrentPPG
3.7, 26%, 0%, 4.5, D.J. Foreman, Rutgers, So, 7.5

3.6, 16%, 4%, -4.8, +Nate Mason, Minnesota, So, 13.4

3.5, 20%, 3%, -1.1, Mark Donnal, Michigan, So, 7.0

3.5, 28%, 4%, -19.8, Marc Loving, Ohio St., Jr, 13.0

3.4, 5%, 2%, 8.9, Bryant McIntosh, Northwestern, So, 14.8

3.4, 26%, 0%, 40.9, P.J. Thompson, Purdue, So, 5.8

2.8, 5%, 4%, 2.8, Matt Costello, Michigan St., Sr, 9.8

2.6, -1%, 0%, 9.9, A.J. Hammons, Purdue, Sr, 14.5

2.6, 1%, -1%, 19.9, Isaac Haas, Purdue, So, 10.2

2.5, 26%, -1%, -8.8, Charles Buggs, Minnesota, Jr, 6.1

2.4, 14%, 1%, -2.2, +Kam Williams, Ohio St., So, 7.8

2.4, 16%, 1%, -6.3, +Jae'sean Tate, Ohio St., So, 11.2

2.2, 10%, 0%, 0.6, Tre Demps, Northwestern, Sr, 14.8

2.0, 11%, 0%, 12.3, Scottie Lindsey, Northwestern, So, 6.4

2.0, 17%, -2%, 19.0, Muhammad-Ali Abdur-Rahkman, Michigan, So, 6.5

1.9, 8%, -3%, 21.2, Adam Woodbury, Iowa, Sr, 8.5

1.9, 13%, 2%, 16.7, Tum Tum Nairn, Jr., Michigan St., So, 4.1

Other than Iowa, the other big jumps in efficiency are concentrated on the Purdue Boilermakers. PJ Thompson has seen the biggest jump in efficiency in the Big Ten. Thompson still isn't consistent and he and Johnny Hill have split starts this year, but Thompson's huge drop in turnover rate and improved shooting have been a great asset for the team. And Isaac Haas has improved his free throw percentage from 55% to 69% allowing him to become an efficient player.

Meanwhile Ohio St.'s Marc Loving has had the biggest drop in efficiency in the Big Ten. Loving has had to take on a larger role in his team's offense and he has struggled.

ChPPG, ChMin, ChPoss, ChORtg, Player, Team, Class, CurrentPPG 

1.7, 2%, 1%, 10.6, Joey King, Minnesota, Sr, 11.4

1.6, 0%, 3%, -0.1, Vince Edwards, Purdue, So, 10.4

1.2, -4%, 2%, 10.4, Aubrey Dawkins, Michigan, So, 8.1

1.0, 10%, -1%, 9.5, Mike Gesell, Iowa, Sr, 8.4

0.9, 3%, 0%, 10.2, Derrick Walton, Jr., Michigan, Jr, 11.7

0.9, -2%, 2%, 0.6, Yogi Ferrell, Indiana, Sr, 17.3

0.5, -7%, 1%, 12.1, Benny Parker, Nebraska, Sr, 4.7

0.5, 0%, 2%, -8.9, Donovon Jack, Penn St., Sr, 4.0

0.4, 8%, 2%, -8.4, Nathan Taphorn, Northwestern, Jr, 4.5

0.4, 1%, 0%, 8.3, Jordan Dickerson, Penn St., Sr, 3.0

0.3, -2%, 0%, 16.5, Dakota Mathias, Purdue, So, 5.1

0.3, 10%, -1%, 1.1, Bishop Daniels, Rutgers, Sr, 8.7

0.2, -12%, 1%, 7.5, Shavon Shields, Nebraska, Sr, 15.7

0.0, 4%, 0%, -0.2, Sanjay Lumpkin, Northwestern, Jr, 4.4

0.0, -1%, 1%, -9.7, Nick Zeisloft, Indiana, Sr, 6.6

These tables do not include Michigan's Caris LeVert, Indiana's James Blackmon, Illinois' Leron Black, and Michigan's Spike Albrecht. Injuries have kept all of these players out of a significant number of games. LeVert and Blackmon were projected to be stars and Black was one of our projected breakout players at, meaning all three injuries were critical. The injury to Albrecht has been tempered by Derrick Walton's big improvement in efficiency this year. Walton was really hampered by injuries last year but he is back to being an efficient and effective PG this season.

I really thought heading into the season that Penn St.'s front-court would be an asset, but it hasn't been. First super-frosh Mike Watkins was declared ineligible. Meawnhile Donovan Jack stopped making his free throws and started turning the ball over more leading to the 9 point drop in efficiency you see here. But my biggest disappointment is Penn St. senior Jordan Dickerson. A lot of big men are late bloomers and the 7 footer seemed like he was making strides last year. It seemed quite possible he would anchor the defense and provide some surprising offense in his final season. The best I can say is that he has remained a solid rim-protector and shot-blocker. But he has been too foul prone and he hasn't been able to stay on the floor and take that next step.

ChPPG, ChMin, ChPoss, ChORtg, Player, Team, Class, CurrentPPG

-0.1, 6%, 1%, -12.3, Collin Hartman, Indiana, Jr, 4.8

-0.3, -3%, 2%, -5.4, Michal Cekovsky, Maryland, So, 2.3

-0.3, -2%, 1%, -5.9, Troy Williams, Indiana, Jr, 12.8

-0.6, -2%, -2%, -10.0, Damonte Dodd, Maryland, Jr, 3.4

-0.7, 4%, -2%, -6.9, Jared Nickens, Maryland, So, 5.4

-0.8, 1%, 0%, -11.6, Carlos Morris, Minnesota, Sr, 10.3

-0.9, -11%, 2%, -4.8, Gavin Schilling, Michigan St., Jr, 4.2

-1.0, -7%, 0%, 3.9, Marvin Clark Jr., Michigan St., So, 3.5

-1.1, -7%, 1%, -14.9, Greg Lewis, Rutgers, Sr, 4.2

-1.1, -17%, -2%, 8.2, Alexandru Olah, Northwestern, Sr, 10.6

-1.1, -6%, -1%, 9.0, Robert Johnson, Indiana, So, 7.8

-1.3, 0%, -4%, -3.0, Jaylon Tate, Illinois, Jr, 2.3

-1.5, -4%, 0%, 2.0, Melo Trimble, Maryland, So, 14.8

-1.6, -16%, 0%, 3.6, Kendall Stephens, Purdue, Jr, 7.2

-1.6, -9%, -3%, 1.0, Rapheal Davis, Purdue, Sr, 9.1

-2.0, 2%, -6%, 2.3, Jake Layman, Maryland, Sr, 10.5

-2.2, -12%, -1%, -11.3, Ricky Doyle, Michigan, So, 4.0

-3.1, -10%, -2%, 3.5, Zak Irvin, Michigan, Jr, 11.2

Maryland's Jake Layman has had to sacrifice a lot of shots with the additions of Robert Carter and Rasheed Sulaimon. Layman has seen the biggest drop in usage rate in the Big Ten.