photo Michigan-Display-Ad-728x90-Die-Hard-Fans-only_zpskcnarkrk.jpg  photo MampGB header 2015 v6_zpsdluogxnr.jpg

Inside the Numbers: Will defense keep Michigan from Sweet Sixteen?

March 20th, 2014 by Drew Hallett


NCAA Tourney media day(MGoBlue.com)

Michigan fans are buzzing with excitement. Tonight, Michigan will kick off its journey in the 2014 NCAA Tournament against the Wofford Terriers in the No. 2 vs. No. 15 matchup in the Midwest Region. Michigan has its highest seed in the NCAA Tournament since it was a No. 1 seed in 1993 and seems poised to make a run at a second straight Final Four.

But should Michigan fans temper their excitement? History says that they should.

This season, the Wolverines have been an offensive juggernaut, recording an adjusted offensive efficiency rating of 121.9. Not only is that rating the third best among all NCAA D-1 teams this year, behind only Creighton (125.0) and Duke (124.5), but it is the seventh best since 2002. It is testament to head coach John Beilein and the offensive talent on this Michigan roster that U-M’s offense has improved despite the departures of Trey Burke and Tim Hardaway, Jr.

Glenn Robinson III and the Michigan offense hope to perform well enough to ease any concerns about the defense (MGoBlue.com)

Glenn Robinson III and the Michigan offense hope to perform well enough to ease any concerns about the defense (MGoBlue.com)

However, offense is only one-half of the game. Michigan has struggled on the defensive end more this season than any other under Beilein. Currently, Michigan is ranked No. 110 in adjusted defensive efficiency with a 101.9 rating. This is the first time in seven seasons under Beilein that the Wolverines have an adjusted defensive efficiency rating higher than 100.0 and the first time they have been ranked outside the top 100 in this statistic.

This development places Michigan in a very concerning situation despite earning a No. 2 seed. Teams with similar profiles to Michigan’s have not gone very far in past NCAA Tournaments. Since 2002, there have been 23 teams in the NCAA Tournament that were ranked in the top 10 in adjusted offensive efficiency and outside the top 100 in adjusted defensive efficiency. Only six of those teams reached the Sweet Sixteen. And only one was victorious in the Sweet Sixteen: Marquette in 2003, which advanced to the Final Four with the assistance of now-NBA All-Star Dwyane Wade.

But that is not the fairest method to evaluate those 23 teams’ NCAA Tournament performance. That method has a giant flaw which is that it does not take a team’s seed into consideration. If those teams were seeded from No. 5 to No. 16, they likely were not a favorite to even reach the Sweet Sixteen.

Thankfully, there is another method that measures more effectively a team’s performance in the NCAA Tournament: Performance Against Seed Expectations (PASE). PASE measures a team’s performance by comparing it to how previous teams with the same seed performed. Therefore, PASE is calculated by tallying the positive or negative differences between actual and expected wins at each seed position.

For example, from 1985 to 2013, all 29 years when the NCAA Tournament fielded at least 64 teams, a team with a No. 2 seed has won 2.41 games, on average, in each NCAA Tournament. Therefore, in 2014, a No. 2 seed is expected to win 2.41 games. If that No. 2 seed wins three or more games, that team has exceeded expectations. On the other hand, if that No. 2 seed wins two or fewer games, that team has underperformed.

Avg wins by seed

I used PASE to measure the performances of the 23 NCAA Tournament teams with a top-10 offense and a sub-100 defense since 2002, and I decided to categorize the teams by seed. The first category included these teams that were seeded from No. 1 to No. 4 because, by seeding, they were expected to reach the Sweet Sixteen. The second category included these teams that were seeded from No. 5 to No. 8 because, by seeding, they were expected to advance to the Round of 32. The third category included these teams that were seeded No. 9 or lower because, by seeding, they were expected to lose their first game.

PASE for Teams with Top-10 Offense and Sub-100 Defense: Seeds No. 1 – 4 (Since 2002)

Seed

Year

Team

Offense Rank

Defense Rank

Expected Wins

Actual Wins

PASE

2

2012

Missouri

1

146

2.41

0

-2.41

2

2005

Wake Forest

1

134

2.41

1

-1.41

3

2006

Gonzaga

1

186

1.86

2

+0.14

3

2005

Gonzaga

8

132

1.86

1

-0.86

3

2003

Marquette

2

119

1.86

4

+2.14

4

2007

Texas

6

106

1.52

1

-0.52

4

2006

Boston College

6

108

1.52

2

+0.48

4

2006

Wake Forest

1

135

1.52

2

+0.48

4

2003

Dayton

5

186

1.52

0

-1.52

Average

1.83

1.44

-0.39

Michigan fans should be concerned most about the foregoing table. Since 2002, there were nine teams with similar profiles to Michigan given a top-four seed in the NCAA Tournament. All teams with these seeds are expected to win 1.83 games. Yet, the nine teams with a similar profile to Michigan won 0.39 games less than expected, on average. On its face, this may not seem like a significant difference, but this is a 21.3 percent reduction in NCAA Tournament wins.

Of these nine teams, all teams that were expected to reach the Sweet Sixteen based solely on its seeding, only four did. Further, neither of the teams that were seeded No. 2, that took the same path in their region that Michigan will take in its region, made it to the second weekend. In 2005, No. 2 seed Wake Forest fell in the Round of 32 to No. 7 seed West Virginia, 111-105, in a classic double-overtime thriller. And, in 2012, Missouri became just the fifth No. 2 seed ever to lose in the Round of 64, being upset at the hands of No. 15 seed Norfolk State, 86-84. Michigan hopes it can be the first No. 2 seed with a top-10 offense and a sub-100 defense to win both of its games in the first weekend since 2002.

PASE for Teams with Top-10 Offense and Sub-100 Defense: Seeds No. 5 – 8 (Since 2002)

Seed

Year

Team

Offense Rank

Defense Rank

Expected Wins

Actual Wins

PASE

6

2010

Notre Dame

7

131

1.17

0

-1.17

7

2004

Michigan State

7

117

0.82

0

-0.82

8

2013

Colorado State

7

135

0.69

1

+0.31

8

2013

North Carolina State

9

112

0.69

0

-0.69

8

2012

Creighton

6

190

0.69

1

+0.31

8

2010

California

5

101

0.69

1

+0.31

8

2007

BYU

10

105

0.69

0

-0.69

Average

0.78

0.43

-0.35

Teams with similar profiles to Michigan’s profile that were seeded from No. 5 to No. 8 have suffered a similar fate as to those given a top-four seed. Where teams with this profile that received a top-four seed lost 0.39 games less than expected, on average, these teams that earned a seed from No. 5 to No. 8 have won 0.35 games less than expected, on average. Only three of these teams exceeded expectations, and none of them appeared in the Sweet Sixteen.

PASE for Teams with Top-10 Offense and Sub-100 Defense: Seeds No. 5 – 8 (Since 2002)

Seed

Year

Team

Offense Rank

Defense Rank

Expected Wins

Actual Wins

PASE

9

2008

Oregon

7

180

.58

0

-0.58

9

2004

Arizona

8

172

0.58

0

-0.58

10

2013

Iowa State

6

133

0.64

1

+0.36

10

2012

Purdue

9

125

0.64

1

+0.36

12

2010

Cornell

4

174

0.54

2

+1.46

12

2009

Arizona

5

155

0.54

2

+1.46

13

2011

Oakland

6

132

0.27

0

-0.27

Average

0.54

0.86

+0.32

However, teams with a top-10 offense and a sub-100 defense have excelled when they are the underdog. Unlike these teams that were given a top-eight seed, teams with a similar profile that earned a bottom-eight seed actually won 0.32 more games than expected, on average. Four of these seven teams pulled off the Round-of-64 upset, including two No. 12 seeds that stunned the public with surprise berths in the Sweet Sixteen.

So what does this all mean? Teams with a top-10 offense and a sub-100 defense are high-variance teams that struggle to consistently perform well because they are one-trick ponies. These teams rely on their offense to win. However, hot-shooting can be touch-and-go and experience high fluctuations. A team can run the perfect play for a 40-percent three-point shooter take a wide-open shot from behind the arc, but, sometimes, that shooter cannot bury it in one game.

Therefore, teams with a top-10 offense and a sub-100 defense are much more vulnerable to upsets in the early rounds of the NCAA Tournament. If that team’s offense sputters for just one game, it may not have the defense to fend off a pesky low-seed opponent. On the other hand, a team with this profile and a low seed may be more likely to cash in on an upset or two in the NCAA Tournament if its elite offense gets hot at the right time.

John Beilein is the biggest overachiever in the NCAA Tournament according to PASE (MGoBlue.com)

John Beilein is the biggest overachiever in the NCAA Tournament according to PASE (MGoBlue.com)

So will Michigan be more vulnerable to an early upset than previous top-two seeds? Most likely. However, this does not mean that Michigan is doomed. There are a few factors working in the Wolverines’ favor that will help U-M be more like 2003 Marquette than 2012 Missouri.

First, Michigan’s offense has been very consistent since the start of the Big Ten season. In its last 20 games, Michigan has averaged at least 1.100 points per possession, which is an excellent rate against Big Ten competition, in 16 of them. Of the four games in which U-M failed to score that many points per possession, one was at the haunted Assembly Hall in Bloomington, one was at Carver-Hawkeye Arena before Iowa collapsed, and one was when Michigan played its third game in three days at the Big Ten Tournament. Michigan will experience none of these conditions during the NCAA Tournament.

Second, Michigan does not have an offense that just happened to fall into the top 10 like some of the teams with similar profiles listed above. The Wolverines have the seventh-best offense since 2002. Therefore, not only has Michigan’s offense been consistent, it also has been more explosive.

Third, teams with a top-10 offense and a sub-100 defense that received top-eight seeds may have underperformed in the NCAA Tournament, but none of those teams were led by Beilein. Using the same PASE metric as above, Beilein is the biggest overachiever in the NCAA Tournament among all active coaches who have made a minimum of five trips. Beilein’s teams have won 0.775 more wins than expected each NCAA Tournament, on average. So, if any coach is going to help ensure that Michigan does not suffer an early upset, it is Beilein.

And, finally, Michigan’s path to the Sweet Sixteen will be a challenge, but it is not formidable. Tonight, the Wolverines face Wofford, who according to Ken Pomeroy, is the worst of the four No. 15 seeds in this NCAA Tournament. Further, Wofford has played five games against teams ranked in Pomeroy’s Top 100. The Terriers lost all five of those games, with its lowest margin of defeat at 14 points. No game is safe in the NCAA Tournament, but Michigan should beat Wofford by a comfortable margin.

Assuming the Wolverines win in the Round of 64, they will face the winner of either No. 7 seed Texas or No. 10 seed Arizona State in the Round of 32. Similar to Wofford, according to Pomeroy, both Texas and Arizona State are the worst of their four respective seeds. Additionally, both the Longhorns and the Sun Devils have struggled lately, with John Gasaway of ESPN listing them as two of the three coldest teams ($) entering the NCAA Tournament. So, although Texas and Arizona State may have a few matchups here and there, the Wolverines appear to have the best odds of any No. 2 seed to reach the Sweet Sixteen.

So should Michigan fans temper their excitement? I think not. Although history suggests that teams with the Wolverines’ profile will be vulnerable to early-round upsets, Michigan seems to have the talent, the coaching, and the first-weekend tournament draw to distinguish itself from the others in the history books. I like the Wolverines to ease past Wofford and claw past either Texas or Arizona State to advance to the Sweet Sixteen for the second straight season.

Related
Michigan hoops preview: Wofford
Examining Michigan’s path through the Midwest region
The M&GB March Madness Five-Spot Challenge
A first look at Wofford