Ken Pomeroy added some data to his website last season showing the most common five-player lineups for college basketball teams. One of the things I'd love to see him add is the defensive rating when various five-player lineups are on the floor.

For example, last year I was very curious whether Duke's defense was generally better in five-player lineups that included the 7'0" Marshall Plumlee.

I don't know of a source that tracks five-player lineup defensive efficiency (or offensive efficiency) for every team, but @nuclearbdgr currently tracks this type of data for Wisconsin. And he was nice enough to share his data with me for last season. The next table shows Wisconsin's two most common lineups last year. This featured a trade-off of Frank Kaminsky and Nigel Hayes:

Starters

Minutes

Off. Eff.

Def. Eff.

Diff

Traevon Jackson, Josh Gasser, Ben Brust, Sam Dekker and Frank Kaminsky

485

1.20

0.97

0.24

Traevon Jackson, Josh Gasser, Ben Brust, Sam Dekker and Nigel Hayes

164

1.15

1.03

0.12

We could probably guess that because of Kaminsky's outside shooting that the Badger offense was better with Kaminsky on the floor. But not everyone would necessarily conclude Kaminsky was the better defender. Kaminsky is not the most agile defender, and Hayes was quite strong if undersized in the post. But the numbers suggest that having the 7 foot Kaminsky on the floor did make a big difference to the Badgers' defense. Wisconsin’s defense allowed 1.03 points per possession with Hayes vs 0.97 points per possession with Kaminsky with the same set of teammates on the floor.

Admittedly, this data isn’t adjusted for opponent. But the bigger problem with this data, as with most college basketball data, is simply the small sample sizes. These were the only five-player lineups that Wisconsin used for over 100 minutes last season. I can think of a lot more fun questions to ask with this data, but everything else in this column suffers from a significant small sample problem.

Many people project Traevon Jackson, Josh Gasser, Sam Dekker, Nigel Hayes, and Frank Kaminsky to be Wisconsin's five-man starting rotation this year, so a natural question is how these five played together last year. But this lineup played only 13.5 minutes together last season. I can tell you it was a dynamic group, scoring 1.54 points per possession while allowing 0.73 points per possession, but drawing conclusions based on 13.5 minutes of data is foolish.

To expand the sample size, we might ask how Wisconsin played last year whenever they played the three big men with any guards. When Dekker, Hayes, and Kaminsky were on the floor together, how did the Badgers perform?

Three Bigs With

Minutes

UW Pts

Opp Pts

UW Poss

Opp Poss

Jackson/Gasser

13.5

30

16

19.5

22

Gasser/Brust

15.1

25

29

21

21

Jackson/Brust

13.4

31

22

22

20

Koenig/Gasser

1.1

1

2

2

2

Jackson/Koenig

6.3

6

10

10.5

10

Koenig/Brust

7.6

21

9

11.5

12

3 Big Total

56.8

114

88

86.5

87

Dividing Wisconsin's points by Wisconsin's possessions we see that a lineup with these three big men was much better offensively, with basically no impact on the team's defense:

Wisconsin

Off Eff

Def Eff

Diff

Dekker/Hayes/Kaminsky together

1.32

1.01

0.31

All other Lineups

1.16

1.01

0.16

I find this fairly fascinating, even if the sample size is too small. Probably the biggest surprise is that the offense was so great last year with these three big men playing together. You might be surprised to see this since Hayes was Wisconsin's least efficient rotation player. But ORtg doesn't always explain a player's role in putting pressure on a defense.

For example, Traevon Jackson has never been Wisconsin's most efficient player, in part because he turns  the ball over on occasion. But that doesn't mean he isn't vital to making the Wisconsin offense work. Jackson is the best player at beating his man off the dribble and causing the defense to collapse. And when the shot-clock is winding down, Jackson is the one player who can create a shot other than a jacked-up three.

In the same vein, Hayes puts a real pressure on the defense whenever he is on the floor because he is such a great back-to-the basket player. Hayes is incredible at drawing fouls, and the attention he draws in the paint makes the Wisconsin offense better.

I think what you see here is that teams guarding Wisconsin faced a real dilemma with this bigger lineup. If they kept their big defenders in the paint to stop Hayes from posting up, that often means Dekker or Kaminsky were shooting over a shorter player, and three point shots are always easier without a hand in your face.

You might think with a taller group of players on the floor that Wisconsin would be better defensively, but the numbers don't support that. I suspect that with the bigger lineup that Wisconsin struggled to keep certain players from driving to the basket.

Of course, you may also wonder if these numbers are slanted because of the quality of competition. According to the data, Wisconsin used this lineup of three big men against a variety of opponents, as listed in the next table. The small sample size is a concern, but I don't think the quality of opponent is greatly impacting these numbers.  

Wisconsin played Minnesota three times, which is one reason they used this lineup the most against the Gophers.

Opponent

Minutes

Minnesota

11.2

Green Bay

6.4

Northwestern

6.2

St. Louis

5.9

Iowa

4.8

Michigan

4.3

Indiana

2.9

Oral Roberts

2.5

E. Kentucky

2.3

Milwaukee

1.9

Ohio St.

1.8

Purdue

1.4

Arizona

1.2

Michigan St.

1.1

Illinois

1.1

North Dakota

0.9

Penn St.

0.6

Baylor

0.5

Total

56.8

Wisconsin doesn't have to play three big men this year. I suspect we will see lineups with Traevon Jackson, Josh Gasser, and Bronson Koenig playing together as well. But according to Nuclear Badgers' calculations, those lineups were not quite as dominant, with an offensive efficiency rating of 1.16 and a defensive efficiency rating of 1.04, a difference of just 0.12. Those three guards played together 132 minutes last year so we have a little more data on that group.

We also need to remember that Hayes and Koenig were true freshmen last year. Typically players improve a lot in their second season. So whether the Badgers use three guards or three forwards, the experience that Hayes and Koenig gained last season should significantly improve their efficiency.

And the reality with this team is clear. Any lineup with Dekker and Kaminsky, two forwards that have a chance to play in the NBA, is going to be extremely dangerous.