Wednesday, March 5, 2008

How good is the Big Ten in the tournament?

Has the Big Ten been a bust or a surprise in recent tournaments? Before you get into the same discussion with your pals while watching the first two rounds, make sure to read this.

Point: Andy Glockner of ESPN has a nice article where he claims that the Big Ten has been receiving too many bids and that the committee should consider giving more bids to the ACC. His evidence is that despite a relatively high proportion of teams receiving bids each year, the Big 10 has sent a relatively low proportion of teams to the Sweet Sixteen.

Counterpoint: There is no question that this is true, (and I had a hard time picking a Big Ten team for the Sweet Sixteen in the column I wrote a few days ago), but I think it is still a little misleading. The issue is that the final Big Ten teams that receive bids also receive poor seeds, so that they are not expected to make the Sweet Sixteen.

Not Favored to Make the Second Week
Proportion of Bids Seeded 5+ in last 10 years
64.3% - 36/56 - Big East
63.6% - 35/55 - Big Ten
54.9% - 28/51 - Big 12
54.4% - 31/57 - SEC
53.3% - 24/45 - Pac10
38.3% - 18/47 - ACC

In fact, the Big Ten teams can often live up to seed even without making it to the second week!

I’ve discussed expectations and tournament performance many times in the past. Recall previous posts where I have discussed the expected wins for each seed:

EW Seed
3.36 - 1
2.43 - 2
1.79 - 3
1.52 - 4
1.17 - 5
1.26 - 6
0.87 - 7
0.67 - 8
0.59 - 9
0.63 - 10
0.50 - 11
0.48 - 12
0.24 - 13
0.18 - 14
0.04 - 15
0.00 - 16

If I take actual tournament wins minus expected wins and add this up for each conference within the last 10 years, things look a lot different:

Actual Wins minus Expected Wins, Last 10 Years
10.88 Big Ten
7.00 Big East
0.98 ACC
-2.00 SEC
-2.61 Pac 10
-3.94 Big 12

While Duke has been earning a 1 seed virtually every year and not quite delivering, teams like Michigan St. have been exceeding seed expectations on an almost annual basis.

Actual Wins minus Expected Wins
7.01 Michigan State (24 wins as 1, 1, 1, 4, 5, 6, 7, 9, and 10 seed)
3.50 Purdue (9 wins as 2, 6, 9, 9, and 10)
1.52 Wisconsin (12 wins as 2, 5, 5, 6, 6, 6, 8, 8, and 9
1.13 Penn State (2 wins as 7)
0.92 Indiana (10 wins as 4, 5, 6, 6, 6, 7, 7, and 7)
0.38 Illinois (16 wins as 1, 1, 4, 4, 4, 4, 5, 5, and 12)
0.21 Ohio State (12 wins as 1, 2, 3, 4, 4, and 5)
-0.79 Michigan (1 win as 3)
-1.46 Iowa (3 wins as 3, 5, 7, and 10)
-1.54 Minnesota (0 wins as 7, 8)
Northwestern (0 appearances)

3.81 Maryland (19 wins as 1, 2, 3, 3, 4, 4, 4, and 6)
1.98 North Carolina (20 wins as 1, 1, 1, 2, 3, 3, 6, and 8)
1.74 Georgia Tech (6 wins as 3, 5, 8, and 10)
0.61 Boston College (3 wins as 4, 7)*
0.52 Florida State (1 win as 12)
0.49 North Carolina State (5 wins as 3, 7, 9, 10, and 10)
-0.17 Virginia Tech (1 win as 5)*
-1.26 Clemson (0 wins as 6)
-1.93 Duke (28 wins as 1, 1, 1, 1, 1, 1, 1, 1, 3, and 6)
-1.69 Virginia (1 win as 4, 5)
-3.12 Wake Forest (5 wins as 2, 2, 4, 7, and 7)
Miami (0 appearances)*
*Since joining ACC

Moreover, while the Big Ten has been pretty poor at sending teams to the Sweet Sixteen, a main reason they’ve exceeded seed as a whole is that the Big Ten has been very good at sending teams to the Final Four.

Teams in Final Four
9 ACC
9 Big Ten
6 Big East
5 Big 12
5 SEC
4 Pac10
2 Other (MWC, CAA)

And again, the Big Ten does this while earning 1 seeds about half as often as the ACC.

Teams with 1 Seeds
12 ACC
7 Pac10
6 Big Ten
5 Big 12
4 Big East
4 SEC
2 Other (CUSA, A10)

I’m certainly not trying to argue that the Big Ten has been a better conference than the ACC, only that the numbers Andy mentions are fairly irrelevant. The fact that the ACC's 1 and 2 seeds often reach the Sweet Sixteen says nothing about whether the ACC actually deserves more poor seeds.

To actually answer Andy’s question of whether the Big Ten or ACC should get those last slots, we could compare the performance of those conferences when they get the last NCAA tournament slots:

1998 #10 Purdue 2 wins
1999 #8 Wisconsin 4 wins
2001 #10 Michigan St. 0 wins
2001 #8 Wisconsin 1 win
2002 #9 Purdue 1 win
2004 #10 Iowa 0 wins
2004 #8 Minnesota 0 wins
2005 #9 Wisconsin 0 wins
2006 #9 Michigan St. 1 win
2006 #12 Illinois 0 wins
2006 #9 Purdue 1 win

1997 #12 Florida State 1 win
1999 #8 North Carolina 4 wins
2000 #8 Georgia Tech 0 wins
2002 #9 North Carolina State 0 wins
2004 #10 North Carolina State 2 wins
2005 #10 North Carolina State 1 win
2006 #10 Georgia Tech 0 wins

Here the ACC does look a little better, but apart from 1999 when both conferences sent an 8 seed to the Final Four, neither conference has been that successful with its last teams in. More importantly, there really isn’t enough data here to reach any strong conclusions. If I were to make an argument, it would be to allow more mid-majors in the field. After all, George Mason of the CAA is the only 11 seed to make the Final Four in the last decade. But should we really conclude from George Mason’s incredible run that we need to let in every borderline CAA team?

I think that any attempt to base bids on past tournament performance is foolish. The numbers don’t support the idea that any conferences bubble teams are clearly inferior or surperior. More importantly, each team should be evaluated on its own merits. Ohio St. should be left out of the field this year, but not because the Big Ten never makes the Sweet Sixteen. Ohio St. should be left out because they are 1-9 vs the RPI Top 50.