dermite
03-31-2008, 08:33 AM
A great website, http://sports.espn.go.com/ncaatourney08/bracketmaster has compiled NCAA tourney seeds and finishes. They have come up with a value of PASE (performance against seed average); statistically compiling the number of wins that each seed averages; #1 seeds average 2.33 wins, etc. I have looked at the numbers differently, comparing expected wins/tourney for seed: 0 wins for seeds 9-16; 1 for seeds 5-8; 2 for seeds 3-4; 3 for two seeds, and 4 for a one seed. The plus/minus number e.g. for a 2 seed going 3-1 and losing in the regional final is 0; FF and losing (4-1)gets a 1+. Teams with a one seed are expected to make the FF, get +1 for the finals, +2 with a NC.
Here is a summary from 1985 (1st yr of seeding) through 2008:
Team Yrs I/O +/- #1 seeds FF NC
Duke 23/1 -1 10 10 3
UNC 22/2 -2 9 8 2
Kentucky 21/3 -1 7 4 2
Kansas 23/1 -7 8 7 1
UConn 15/9 +4 4 2 2
Indiana 21/3 -5 2 3 1
UCLA 19/5 +2 3 4 1
Louisville 17/7 +6 0 2 1
Arizona 24/0 -12 5 4 1
Florida 14/10 +7 1 4 2
The plus minus column is current as of 3/31/08 for UNC, Kansas, and UCLA.
This compendium reflects 1985 on up; obviously that leaves out the bulk of UCLA’s work; Kentucky’s rich past; UNC’s better teams in the early 1980s; and our four previous final four teams. At least it provides one objective way of comparing programs.
Not all plus/minus against the seeds are the same. A 4 seed losing to a 5 seed (-1) doesn’t hurt as much as a heavyweight team that could go much further (think our 1998 #1 seed) losing to a 2 seed (-1).
The striking analysis of the plus minus? Duke, UNC and UK have been on a fairly even keel over 24 years. The splits stick out: we were a whopping +11 (85-94) and -12 from 1996-2008. We had stunning success (or good fortune) from 1985-1994 against expectations. Our tourney results from 1996-2008 did not measure up to our regular season expectations. Bad coaching or choking? No. Bad luck? Perhaps. Regression to the mean? Yes.
The perception that Arizona, and to a lesser extent Kansas, disappoints in the tourney rings true with the numbers.
The low number of one seeds for some of the programs is really surprising.
Teams like UConn and Florida that miss the tournament a lot are harder to compare with statistically. After all, you would much rather make the tourney and bow out earlier than not have a chance at all.
I don’t think that K was any smarter the first 10 years than he is now: still a HOFer, still the best. I think that we are regressing to the statistical mean. We overachieved statistically, particularly in 1988-1990 where we made three FFs; the fact that we were beaten rather handily in each of those years for our final game suggests that we did go farther than we “should” have. There’s no greater way to enhance your reputation than what we did in 1991: win a NC without having the best team in the country. (Unlike years where we did have the best team and didn’t win, such as 1986, 1999, 2004, at a minimum). Our tremendous NCAA tourney success 1986-1994 revived our national reputation a lot more than our “underachieving” of 1996-2008 has hurt it. Our overall plus minus is comparable to the other heavyweight powers, except for notorious underachievers like Kansas and Arizona.
The best way to look at the health of the program is to look at the number of one seeds. That puts you into the best possible situation before the crapshoot known as a six game single elimination tournament.
Here is a summary from 1985 (1st yr of seeding) through 2008:
Team Yrs I/O +/- #1 seeds FF NC
Duke 23/1 -1 10 10 3
UNC 22/2 -2 9 8 2
Kentucky 21/3 -1 7 4 2
Kansas 23/1 -7 8 7 1
UConn 15/9 +4 4 2 2
Indiana 21/3 -5 2 3 1
UCLA 19/5 +2 3 4 1
Louisville 17/7 +6 0 2 1
Arizona 24/0 -12 5 4 1
Florida 14/10 +7 1 4 2
The plus minus column is current as of 3/31/08 for UNC, Kansas, and UCLA.
This compendium reflects 1985 on up; obviously that leaves out the bulk of UCLA’s work; Kentucky’s rich past; UNC’s better teams in the early 1980s; and our four previous final four teams. At least it provides one objective way of comparing programs.
Not all plus/minus against the seeds are the same. A 4 seed losing to a 5 seed (-1) doesn’t hurt as much as a heavyweight team that could go much further (think our 1998 #1 seed) losing to a 2 seed (-1).
The striking analysis of the plus minus? Duke, UNC and UK have been on a fairly even keel over 24 years. The splits stick out: we were a whopping +11 (85-94) and -12 from 1996-2008. We had stunning success (or good fortune) from 1985-1994 against expectations. Our tourney results from 1996-2008 did not measure up to our regular season expectations. Bad coaching or choking? No. Bad luck? Perhaps. Regression to the mean? Yes.
The perception that Arizona, and to a lesser extent Kansas, disappoints in the tourney rings true with the numbers.
The low number of one seeds for some of the programs is really surprising.
Teams like UConn and Florida that miss the tournament a lot are harder to compare with statistically. After all, you would much rather make the tourney and bow out earlier than not have a chance at all.
I don’t think that K was any smarter the first 10 years than he is now: still a HOFer, still the best. I think that we are regressing to the statistical mean. We overachieved statistically, particularly in 1988-1990 where we made three FFs; the fact that we were beaten rather handily in each of those years for our final game suggests that we did go farther than we “should” have. There’s no greater way to enhance your reputation than what we did in 1991: win a NC without having the best team in the country. (Unlike years where we did have the best team and didn’t win, such as 1986, 1999, 2004, at a minimum). Our tremendous NCAA tourney success 1986-1994 revived our national reputation a lot more than our “underachieving” of 1996-2008 has hurt it. Our overall plus minus is comparable to the other heavyweight powers, except for notorious underachievers like Kansas and Arizona.
The best way to look at the health of the program is to look at the number of one seeds. That puts you into the best possible situation before the crapshoot known as a six game single elimination tournament.