2010 NCAA Tournament Retrospective

Last night, Duke won the 2010 NCAA basketball tournament in the best final game in at least a decade.  If you didn't see it, Butler barely missed a half court shot at the buzzer that would have given them the game.  In some ways it is too bad they missed as it would have been a fitting end to the season.  Either way, it was a nice cap to an otherwise mediocre year in basketball.



One thing I usually like to do after a season is to compare conferences in the tournament to see how the conferences stack up against one another.  This was the result for the 2010 tournament in order of winning percentage.

	Teams	W	L	Pct
Horizon	1	5	1	0.833
Ivy	1	2	1	0.667
Missouri Valley	1	2	1	0.667
ACC	6	9	5	0.643
Big Ten	5	9	5	0.643
Pac 10	2	3	2	0.600
SEC	4	6	4	0.600
West Coast	2	3	2	0.600
Big 12	7	9	7	0.563
Big East	8	8	8	0.500
Colonial	1	1	1	0.500
Mid-American	1	1	1	0.500
Ohio Valley	1	1	1	0.500
Atlantic 10	3	2	3	0.400
Mountain West	4	2	4	0.333
Conference USA	2	0	2	0.000
Other	12	0	12	0.000
Sun Belt	1	0	1	0.000
WAC	2	0	2	0.000

As you can see from this listing, the non-BCS conferences had a great year winning a total of 19 games in the tournament lead by Butler, Cornell, and Northern Iowa.  This was the most wins by non-BCS schools in the history of the tournament (they won 18 in 1998 and 2006 and 17 in 1990, 1996, and 2008).  There performance certainly justified their attendance.

One thing these numbers don't tell is how each conference did vs expectations.  For example if you look at the Big East, the .500 record doesn't look too bad.  The truth underlying these numbers is the Big East was by far the highest ranked conference at the end of the regular season so they had a disproportionate share of high seeds.  In theory a high seed should win more games than a lower seeded team as they face easier competition.  After thinking about it a bit, I decided to try to come up with a way to create  way to compare expected tournament results with actual results.  What I came up with was to look at the NCAA tournament for the last 25 years and see how many games a typical team won at various rankings.

Rank	W	L	Pct	W/Year
1	345	84	0.804	3.450
2	242	97	0.714	2.420
3	187	96	0.661	1.870
4	148	99	0.599	1.480
5	112	101	0.526	1.120
6	121	100	0.548	1.210
7	85	100	0.459	0.850
8	66	99	0.400	0.660
9	58	100	0.367	0.580
10	64	100	0.390	0.640
11	48	100	0.324	0.480
12	52	100	0.342	0.520
13	25	100	0.200	0.250
14	17	100	0.145	0.170
15	5	99	0.048	0.050
16	0	100	0.000	0.000

What this means is in the past 25 years, if a team was ranked #1 then they won an average of 3.45 games.  A #6 ranked team only won 1.21 times.  If you use the W/Year as the standard expectation for the teams going into the 2010 tournament then you come up with the following:

	Teams	Wins	Expected	Diff.
Horizon	1	5	1.1	3.9
Big Ten	5	9	7.0	2.0
Pac 10	2	3	1.1	1.9
West Coast	2	3	1.3	1.7
Ivy	1	2	0.5	1.5
ACC	6	9	7.6	1.4
Missouri Valley	1	2	0.6	1.4
Mid-American	1	1	0.2	0.8
Ohio Valley	1	1	0.3	0.8
Colonial	1	1	0.5	0.5
Sun Belt	1	0	0.1	(0.1)
Conference USA	2	0	0.8	(0.8)
SEC	4	6	6.8	(0.8)
WAC	2	0	1.0	(1.0)
Other	12	0	1.2	(1.2)
Atlantic 10	3	2	3.2	(1.2)
Mountain West	4	2	3.9	(1.9)
Big 12	7	9	11.0	(2.0)
Big East	8	8	15.0	(7.0)

This again shows the success of the mid-majors and the relative success of the Big Ten / Pac 10 vs expectations.  The most amazing statististic is at the bottom.  Based on seeding alone the Big East was expected to win 7 more games then actually occurred.  To put this is perspective here is the all time worst performances before this year.

	Year	Teams	Wins	Expected	Diff.
Pac 10	2000	4	3	9.0	(6.0)
Big Ten	1996	5	2	7.8	(5.8)
Big Ten	2006	6	3	8.8	(5.8)
Big 12	1992	7	5	10.4	(5.4)
Big Ten	1995	6	1	6.0	(5.0)
Pac 10	1999	4	1	5.9	(4.9)
Big 12	2001	6	3	7.7	(4.7)
SEC	1989	7	1	5.6	(4.6)
Big Ten	1986	6	4	8.6	(4.6)
Pac 10	1989	4	3	7.4	(4.4)
SEC	2002	6	5	9.3	(4.3)

The Big East beat the previous post season collapse in NCAA history by 1 whole game.  They should thank god for West Virginia because it could have been much worse.