Predicting the Final Four

[SWC75]

On one thread a poster was talking about the predictability of NCAA tournament results. I pointed out that last year’s Final Four consisted of arguably John Calipari’s worst Kentucky team, a Connecticut team that was 9th place in the Big East, a rebuilding Butler team and a Virginia Commonwealth team that came out of nowhere. The poster admitted that that was a strange year but said that those results were unusual. Naturally, I had to find out how unusual- and how often the Final Four follows form. I used the seedings of the team to determine how predictable the Final Four was. The NCAA began seeding the tournament in 1979, although it didn’t go to 64 teams until 6 years later. In any year since 1979, a perfectly predictable Final Four would have a score of 4- the combined seedings of the four teams. That would mean that all four #1 seeds made it. The higher the combined seeds are above that, the less predictable the Final Four was.

1979 Pennsylvania a 9 seed, Michigan State a 2 seed, DePaul a 2 seed, Indiana State a 1 seed. Combined seed: 14
1980 Iowa 5, Louisville 2, Purdue 6, UCLA 8. Combined: 21
1981 Louisiana State 1, Indiana 3, North Carolina 2, Virginia 1. Combined: 7
1982 North Carolina 1, Houston 6, Louisville 3, Georgetown 1. Combined: 11
1983 Houston1, Louisville1, North Carolina State 6, Georgia 4. Combined: 12
1984 Kentucky1, Georgetown 1, Virginia 2, Houston 2. Combined: 6
1985 St. John’s 1, Georgetown 1, Memphis 2, Villanova 8. Combined: 12
1986 Louisiana State 11 Louisville 2, Duke, 1, Kansas 1. Combined: 15
1987 UNLV 1, Indiana 1, Providence 6, Syracuse 2. Combined: 10
1988 Kansas 6, Duke 2, Arizona 1, Oklahoma 1. Combined: 10
1989 Illinois 1, Michigan 3, Duke 2, Seton Hall 3. Combined: 9
1990 Georgia Tech 4, UNLV 1, Duke 3, Arkansas 4. Combined: 12
1991 UNLV 1, Duke 2, North Carolina a, Kansas 3. Combined: 7
1992 Cincinnati 4, Michigan 6, Duke 1, Indiana 2. Combined: 13
1993 Kentucky1, Michigan 1, North Carolina 1, Kansas 2. Combined: 5
1994 Arizona 2, Arkansas 1, Florida 3, Purdue 1. Combined: 7
1995 UCLA 1, Oklahoma State 4, North Carolina 2, Arkansas 2. Combined: 9
1996 Mississippi State 5, Syracuse 4, Massachusetts 1, Kentucky 1. Combined: 11
1997 Kentucky 1, Minnesota 1, North Carolina 1, Arizona 4. Combined: 7
1998 North Carolina 1, Utah 3, Stanford 3, Kentucky 2. Combined: 9
1999 Duke 1, Michigan State 1, Ohio State 4, Connecticut 1. Combined: 7
2000 Wisconsin 8, Michigan State 1, Florida 5, North Carolina 8. Combined: 22
2001 Arizona 2, Michigan State 1, Duke 1, Maryland 3. Combined: 7
2002 Indiana 5, Oklahoma 2, Maryland 1, Kansas 1. Combined: 9
2003 Marquette 3, Kansas 2, Texas 1, Syracuse 3. Combined: 9
2004 Georgia Tech 3, Oklahoma State 2, Duke 1, Connecticut 2. Combined: 8
2005 Illinois 1, Louisville 4, North Carolina 1, Michigan State 5. Combined: 11
2006 Louisiana State 4, UCLA 2, George Mason 11, Florida 3. Combined: 20
2007 Georgetown 2, Ohio State 1, Florida 1, Kansas 1. Combined: 5
2008 North Carolina 1, Kansas 1, Memphis 1, UCLA 1. Combined: 4
2009 Michigan State 2, Connecticut 1, Villanova 3, North Carolina 1. Combined: 7
2010 West Virginia 2, Duke 1, Michigan State 5, Butler 5. Combined: 13
2011 Kentucky 4, Connecticut 3, Virginia Commonwealth 11, Butler 8. Combined: 26

Let’s break it down by areas.
The combined seeding of the Final Four has been 4 once.
It’s been between 5-8, (8 would be four #2 seeds), twelve times.
It’s been between 9-12, (four #3 seeds), thirteen times.
It’s been between 13-16, (four #4 seeds), four times.
It’s been between 17-20, (four #5 seeds), one time.
It’s been between 21-24, (four #6 seeds) two times.
It’s been between 25-28, (four #7 seeds) one time.

2011 was the worst Final Four in terms of seedings, (and quality of play) we’ve had in the era of seedings. But some others, (1980, 2000, 2006) weren’t far off. And there were four times as many FFs in the 20’s as there were “perfect” ones. The 2008 FF is the only one with all four #1 seeds. Four of them have had three #1 seeds. Fifteen of them had two #2 seeds. Ten of them have had one #1 seed. The other three, including 2011, had no #1 seed.

So how predicable is the NCAA tournament? Most of the time, somewhat predicable. But sometimes it goes well off course. What will it do this year? Try to predict that!
　
　

 

[SoBristol]

Good start, SWC. (2008 line adds to 5 by the way.)

Now, you have to combine that data with data about close games vs blow-outs. Try to figure out whether the top seeds in those odd years (1980, 2000, 2006, 2011) struggled with close games in February, or had a key injury, or had some other factor that fore shadowed an odd result. Rebounding margin?

 

[arrow2arinzesknee]

I think he made a typo on Memphis they should be 1 so the total was in fact 4

If I had to make a prediction right now to pick the 4 teams I think will make a run in the tournament (not the 4 best teams in my mind right now) I think I would have to say

Syracuse- a little biased here but the zone and depth will give lots of teams serious matchup problems and I am more familiar with them and feel like anything less than that would be a vast underachievement

Missouri- guard play guard play guard play, Plus a two seed or a one seed draw helps

Michigan St- They always seem to get a good draw / favorable upsets which although has no bearing on this years team, gives Izzo the appearance of being a dominant tournament coach, the team is peaking at a good time and has a good chance

Florida- I told my dad after we got done with the game this year, that Florida is a team that scare me come tournament time and that I wouldn't want them in my draw, they have a legitimate low post threat, another post threat who can extend his game, and three guards who have in the gym range, donovan has more NT than JB his coaching in crunch time is not a negative, and they have shown that in some losses (Rutgers?) that they can lose to anyone but I think they are a dangerous team that if shooting hot can win the whole thing.

My predicted seeds for those 4 teams
SU-1
Mizzou- 2
Mich St.- 3
Florida-4

totalling 10 a number in the middle that the 9-10 range frequents quite a bit in your observation above

 

[arrow2arinzesknee]

reasons against some other favorites

Kentucky- they are so talented but also so young, and a crunch time situation in the second round against lets say a big east bubble team, a Pac 12 team like Cal who has something to prove, or even a hot shooting team like Murray st could get the wildcats in a close game, where as we saw at Indiana and recently vs. Vanderbilt, they can get a little tense in a lose and go home scenario ala the Wall, Cousins team a few years back.

Ohio St- could win it all, I just don't particularly feel Sullenger has progressed where he should be and that they lose a matchup in sweet 16 or elite 8 to a good team on a good run

UNC- anytime you get blown out the way they did against FSU, and I watched them struggle at Miami last night, I just don't think the pundits are right on these guys.

Duke- same as UNC, the ACC is down this year and Duke has lost 2 this year at Cameron, when is the last time that happened?

anyone of those 4 wouldnt surprise me, not would Marquete, L'ville, or Kansas

all of the other teams just don't do it for me

 

[Marsh01]

I think you will see a couple of number one seeds make it as well as a #2 and a team that completely noone had in there. Something like this:

Kentucky
Syracuse
Missouri/UNC
Florida State/Vanderbilt/UNLV/Marquette

Something like that. If you look at the teams that all of these bracketologists are picking to be 4-5-6 seeds, you will see that the field is pretty weak this year.

 

[MartyCuse]

SWC your post are always detailed and well researched. Thank you for the time... As long as there is one #1 in there this year and its Syracuse I hope the total score is a 49!

 

[SWC75]

Good start, SWC. (2008 line adds to 5 by the way.)

Now, you have to combine that data with data about close games vs blow-outs. Try to figure out whether the top seeds in those odd years (1980, 2000, 2006, 2011) struggled with close games in February, or had a key injury, or had some other factor that fore shadowed an odd result. Rebounding margin?

The nice thing about being called out on a typo is that it proves that someone is actually reading this stuff.

 

[OrangeMojo]

On one thread a poster was talking about the predictability of NCAA tournament results. I pointed out that last year’s Final Four consisted of arguably John Calipari’s worst Kentucky team, a Connecticut team that was 9th place in the Big East, a rebuilding Butler team and a Virginia Commonwealth team that came out of nowhere. The poster admitted that that was a strange year but said that those results were unusual. Naturally, I had to find out how unusual- and how often the Final Four follows form. I used the seedings of the team to determine how predictable the Final Four was. The NCAA began seeding the tournament in 1979, although it didn’t go to 64 teams until 6 years later. In any year since 1979, a perfectly predictable Final Four would have a score of 4- the combined seedings of the four teams. That would mean that all four #1 seeds made it. The higher the combined seeds are above that, the less predictable the Final Four was.

1979 Pennsylvania a 9 seed, Michigan State a 2 seed, DePaul a 2 seed, Indiana State a 1 seed. Combined seed: 14
1980 Iowa 5, Louisville 2, Purdue 6, UCLA 8. Combined: 21
1981 Louisiana State 1, Indiana 3, North Carolina 2, Virginia 1. Combined: 7
1982 North Carolina 1, Houston 6, Louisville 3, Georgetown 1. Combined: 11
1983 Houston1, Louisville1, North Carolina State 6, Georgia 4. Combined: 12
1984 Kentucky1, Georgetown 1, Virginia 2, Houston 2. Combined: 6
1985 St. John’s 1, Georgetown 1, Memphis 2, Villanova 8. Combined: 12
1986 Louisiana State 11 Louisville 2, Duke, 1, Kansas 1. Combined: 15
1987 UNLV 1, Indiana 1, Providence 6, Syracuse 2. Combined: 10
1988 Kansas 6, Duke 2, Arizona 1, Oklahoma 1. Combined: 10
1989 Illinois 1, Michigan 3, Duke 2, Seton Hall 3. Combined: 9
1990 Georgia Tech 4, UNLV 1, Duke 3, Arkansas 4. Combined: 12
1991 UNLV 1, Duke 2, North Carolina a, Kansas 3. Combined: 7
1992 Cincinnati 4, Michigan 6, Duke 1, Indiana 2. Combined: 13
1993 Kentucky1, Michigan 1, North Carolina 1, Kansas 2. Combined: 5
1994 Arizona 2, Arkansas 1, Florida 3, Purdue 1. Combined: 7
1995 UCLA 1, Oklahoma State 4, North Carolina 2, Arkansas 2. Combined: 9
1996 Mississippi State 5, Syracuse 4, Massachusetts 1, Kentucky 1. Combined: 11
1997 Kentucky 1, Minnesota 1, North Carolina 1, Arizona 4. Combined: 7
1998 North Carolina 1, Utah 3, Stanford 3, Kentucky 2. Combined: 9
1999 Duke 1, Michigan State 1, Ohio State 4, Connecticut 1. Combined: 7
2000 Wisconsin 8, Michigan State 1, Florida 5, North Carolina 8. Combined: 22
2001 Arizona 2, Michigan State 1, Duke 1, Maryland 3. Combined: 7
2002 Indiana 5, Oklahoma 2, Maryland 1, Kansas 1. Combined: 9
2003 Marquette 3, Kansas 2, Texas 1, Syracuse 3. Combined: 9
2004 Georgia Tech 3, Oklahoma State 2, Duke 1, Connecticut 2. Combined: 8
2005 Illinois 1, Louisville 4, North Carolina 1, Michigan State 5. Combined: 11
2006 Louisiana State 4, UCLA 2, George Mason 11, Florida 3. Combined: 20
2007 Georgetown 2, Ohio State 1, Florida 1, Kansas 1. Combined: 5
2008 North Carolina 1, Kansas 1, Memphis 1, UCLA 1. Combined: 4
2009 Michigan State 2, Connecticut 1, Villanova 3, North Carolina 1. Combined: 7
2010 West Virginia 2, Duke 1, Michigan State 5, Butler 5. Combined: 13
2011 Kentucky 4, Connecticut 3, Virginia Commonwealth 11, Butler 8. Combined: 26

Let’s break it down by areas.
The combined seeding of the Final Four has been 4 once.
It’s been between 5-8, (8 would be four #2 seeds), twelve times.
It’s been between 9-12, (four #3 seeds), thirteen times.
It’s been between 13-16, (four #4 seeds), four times.
It’s been between 17-20, (four #5 seeds), one time.
It’s been between 21-24, (four #6 seeds) two times.
It’s been between 25-28, (four #7 seeds) one time.

2011 was the worst Final Four in terms of seedings, (and quality of play) we’ve had in the era of seedings. But some others, (1980, 2000, 2006) weren’t far off. And there were four times as many FFs in the 20’s as there were “perfect” ones. The 2008 FF is the only one with all four #1 seeds. Four of them have had three #1 seeds. Fifteen of them had two #2 seeds. Ten of them have had one #1 seed. The other three, including 2011, had no #1 seed.

So how predicable is the NCAA tournament? Most of the time, somewhat predicable. But sometimes it goes well off course. What will it do this year? Try to predict that!
　
　

that 2000 final four was just brutal - I hated that final four

 

[skurey]

Minnesota was a 1 seed in 97, wow.

 

[SoBristol]

The nice thing about being called out on a typo is that it proves that someone is actually reading this stuff.

Always read your stuff. Always appreciate what you bring to this site.
Thanks!