Bracetology 2012

S

SteveDeace

Well-Known Member
Back when I was still doing sports talk radio, I used to make light of ESPN’s “bracketologist†Joe Lunardi for bragging about predicting 99% of the teams to make the NCAA Tournament every year. After all, how hard can it be given the amount of brackets he puts out during the season right up until the actual bracket is released?


To prove my point, two years ago I started predicting the NCAA Tournament bracket two weeks before all the conference tournaments were completed and the Selection Committee unveiled the actual bracket. In 2010, I got 86% of the teams that eventually made the field correct. Last year that number dropped to 79%.

This year I’m back again to demonstrate that Lunardi is dramatically
overpaid.


In picking my field I am going through the same selection process the Selection Committee undergoes each year. I weigh the same factors, but put the most weight on non-conference strength of schedule. The reason why is since the RPI was introduced to the selection process in early 1990s there have been two constants.


First, every Selection Committee has chosen whether or not to weigh factors such as record in your last 10 games, conference strength, or head-to-head competition differently. But every committee has put a premium on non-conference strength of schedule for the teams from the power conferences. The greatest example of this was Georgia in 2001, which earned an at-large berth despite owning just a 16-13 record because the Bulldogs faced the toughest schedule in the nation.


Second, while it is true the RPI is not the end-all-be-all for the Selection Committee, the criteria that makes up the formula makes the RPI a sort of framework for the process as a whole. So while the at-large teams and the seeding aren’t a perfect match with the RPI, you rarely find at-large teams with a RPI of higher than 70 making the field. It’s only occurred a handful of times in 20 years, and one of them was New Mexico when its athletics director was on the Selection Committee.


There are 31 automatic bids to the 2012 NCAA Tournament, and 37 at-large bids. Automatic bids are awarded to the winner of each conference tournament, except for the Ivy League which doesn’t have one. Instead, its regular season champion is awarded its automatic bid.


I project the 31 automatic bids to go to the following teams:



  1. AMERICA EAST—Vermont
  2. *ATLANTIC 10—Xavier
  3. ACC—North Carolina
  4. ATLANTIC SUN—Belmont
  5. BIG 12—Missouri
  6. BIG EAST—Syracuse
  7. BIG SKY—Weber State
  8. BIG SOUTH—UNC-Asheville
  9. BIG TEN—Ohio State
  10. BIG WEST—Long Beach State
  11. *COLONIAL ATHLETIC—VCU
  12. CONFERENCE USA—Memphis
  13. HORIZON—Valparaiso
  14. IVY—Harvard
  15. METRO ATLANTIC—Iona
  16. MAC—Akron
  17. MEAC—Savannah State
  18. MISSOURI VALLEY—Wichita State
  19. MOUNTAIN WEST—UNLV
  20. NORTHEAST—Wagner
  21. OHIO VALLEY—Murray State
  22. *PAC-12—Arizona
  23. PATRIOT—Bucknell
  24. SEC—Kentucky
  25. SOUTHERN—Davidson
  26. SOUTHLAND—Texas-Arlington
  27. SWAC—Mississippi Valley State
  28. SUMMIT—Oral Roberts
  29. SUN BELT—Middle Tennessee
  30. WEST COAST—Gonzaga
  31. WAC—Nevada
Each year there are at least two or three “bracket busters†during the conference tournaments. A “bracket buster†is when a team wins a conference tournament that wasn’t already considered an at-large team regardless, thus potentially taking away spots on the so-called bubble.


In my predictions, you’ll note there are three such “bracket busters†in the Atlantic 10, Colonial Athletic, and Pac-12. I chose those three leagues specifically because I anticipate they have the highest likelihood of producing a “bracket buster.â€


Now to the at-large selection process.


Based on my conference tournament projections, I predict the following at-large bids to be safe regardless of how each of these teams finishes the final two weeks.


ACC (2): Duke, Florida State
Big East (4): Marquette, Georgetown, Louisville, Notre Dame
Big Ten (4): Michigan State, Michigan, Wisconsin, Indiana
Big 12 (3): Kansas, Baylor, Iowa State
SEC (2): Florida, Vanderbilt
Missouri Valley (1): Creighton
Mountain West (2): San Diego State, New Mexico
Atlantic 10 (2): Temple, St. Louis
Conference USA (1): Southern Mississippi


According to my projections 21 of the 37 at-large spots in the field are spoken for. That’s actually the fewest this late in the season I can recall, which means there is still plenty of time for teams like Virginia and Connecticut to play their way out of the field, and time for teams like BYU and Northwestern to play their way in.


I project the final 16 at-large spots to be awarded as follows.


ACC (2): Virginia, Miami (Fla.)
Big East (3): Connecticut, Seton Hall, West Virginia
Big Ten (2): Purdue, Northwestern
Big 12 (2): Kansas State, Texas
Conference USA (1): Marshall
Pac-12 (2): California, Washington
SEC (2): Alabama, Mississippi State
Mountain West (1): Colorado State
Colonial Athletic (1): Drexel


Last four in—Miami (Fla.), West Virginia, Northwestern, Drexel
Last four out—South Florida, BYU, Dayton, N.C. State


Northwestern, which has never made the NCAA Tournament before, was the second to last team in my projected field. South Florida, which could finish as high as third in the Big East, was the first team out. The difference is strength of schedule. The Wildcats have played a top-10 strength of schedule, courtesy of playing in the toughest conference in the country and also faced four RPI top 100 non-conference foes.


In recent years the Selection Committee has shown mercy to respected mid-majors with gaudy records that don’t win their conference tournaments, hence Drexel being the last team in.


However, once I got down to the final five slots in the field there weren’t any teams objectively deserving of an at-large bid. There were just teams subjectively less undeserving than the others. At that point beauty is totally in the eye of the beholder.


The biggest surprise in my selection process was Marshall, which has a top-25 RPI thanks mainly to a top 20 non-conference strength of schedule that included wins over two teams that made my field of 68 – Iona and Belmont – as well as a win at Cincinnati.


With our 68 teams selected, next comes the seeding process.


EAST (Boston)
@Pittsburgh
1. Syracuse vs. 16. Wagner/UNC-Asheville*
8. Xavier vs. 9. Virginia
@Greensboro
2. Duke vs. 15. Vermont
7. Iowa State vs. 10. Seton Hall
@Columbus
3. Michigan vs. 14. Belmont
6. Vanderbilt vs. 11. Creighton
@Albuquerque
4. Baylor vs. 13. Iona
5. Temple vs. 12. Marshall


WEST (Phoenix)
@Greensboro
1. North Carolina vs. 16. Texas-Arlington
8. New Mexico vs. 9. Arizona
@Louisville
2. Ohio State vs. 15. Weber State
7. Kansas State vs. 10. Mississippi State
@Albuquerque
3. UNLV vs. 14. Oral Roberts
6. Florida State vs. 11. Washington
@Portland
4. Louisville vs. 13. Nevada
5. Gonzaga vs. 12. Texas


SOUTH (Atlanta)
@Louisville
1. Kentucky vs. 16. Savannah State/Mississippi Valley State*
8. Memphis vs. 9. Purdue
@Omaha
2. Missouri vs. 15. Bucknell
7. St. Louis vs. 10. California
@Nashville
3. Marquette vs. Davidson
6. Wisconsin vs. Colorado State
@Nashville
4. Murray State vs. 13. Northwestern/Miami (Fla.)*
5. Notre Dame vs. 12. VCU


MIDWEST (St. Louis)
@Columbus
1. Michigan State vs. 16. Middle Tennessee
8. Southern Mississippi vs. 9. Connecticut
@Omaha
2. Kansas vs. 16. Valparaiso
7. San Diego State vs. 10. Alabama
@Pittsburgh
3. Georgetown vs. 14. Akron
6. Indiana vs. 11. Long Beach State
@Portland
4. Wichita State vs. West Virginia/Drexel*
5. Florida vs. 12. Harvard


*: the last four at-large teams in the field, and the last four automatic teams, play in the NCAA first round game prior to the Field of 64
 
I used to do this on my blog. I did it for four or five years. My first two years I only missed a couple teams. It really isnt that difficult. I was at 98 percent, I believe. Havent looked back in many years.
 
 
Will considering before the bracket you know who all the automatic byes are and you know who 90 percent of who the locks are. If you really break it down there's probably like 8 teams on the bubble for 4 spots and if you get 4 right I guess you could claim 64/68 which is 94 percent. Seems pretty easy to me. Even Deace predicting two weeks before 86 and 79 percent is pretty bad, considering Deace is probably projecting and Lanardi is not.
 
Is anyone going to point out that 86% and 79% correct is horrible? My wife got 84% correct last year
 
busabus114 said:
Is anyone going to point out that 86% and 79% correct is horrible? My wife got 84% correct last year
Click to expand...

Smoooth Busabus. At first I kind of thought you were slamming your wife because she's a girl 'n all and what would she know bout basketball, but then I realized you're just pointing out how easy it is to be smarter than the OP.
 
Lunardi is extremely overrated. His job really isn't that hard. They will say he got 62/64 teams right. But really 31 of those are at-large teams that you can't get wrong. At least 28 of the remaining spots are no brainers. So he got 4/6 correct, big whoop.

Heres a good link if you want to follow better bracketologists.

The Bracket Project - Ranking the Bracketologists
 
busabus114 said:
Is anyone going to point out that 86% and 79% correct is horrible? My wife got 84% correct last year
Click to expand...

Take a look at Lunardi's bracket yesterday, and snapshot it...that's two weeks out. That is what Steve is saying. Lunardi claims genius on the bracket he submits about 15 minutes before the tournament field is announced...most can do that good
 
Lunardi keeps track of the bracket throughout the season. It's pretty much another ranking system that is updated regularly, like Kenpom, realtime RPI or even the AP or the coaches poll. It's a fun way to see where your team is compare to others throughout the course of the season.

Fact is, he is one of the "few" people who follow the game, and every team very closely and he is kind of good at it. It interests people and gathers hits to the website therefor he gets paid to do it.

Acting, and caring, like it's anything other than that is a waste of time.
 
bcl20 said:
Acting, and caring, like it's anything other than that is a waste of time.
Click to expand...

ESPN's talking heads spittle all over themselves saying what a genius he is, etc, etc...and its a joke. I almost think ESPN went out and created the computer program to emulate the RPI, then went out and hired the nerdiest person they could find to cast as their geek bracketologist.
 
JonDMiller said:
ESPN's talking heads spittle all over themselves saying what a genius he is, etc, etc...and its a joke. I almost think ESPN went out and created the computer program to emulate the RPI, then went out and hired the nerdiest person they could find to cast as their geek bracketologist.
Click to expand...

Well, yeah, but most talking heads are dumb ex-athletes who are not very analytically people. I don't really pay attention to anything they say regardless of the subject.
 
Hey, I just wanted to know how Stevie was privy to the contracts of employees at ESPN? Could you tell me what Cowherd and Van Pelt make? I think those two guys know great radio.
 
JonDMiller said:
Take a look at Lunardi's bracket yesterday, and snapshot it...that's two weeks out. That is what Steve is saying. Lunardi claims genius on the bracket he submits about 15 minutes before the tournament field is announced...most can do that good
Click to expand...

It's true, most can.

And, as I have proven with the empirical, date-stamped evidence of my own predictions, it's really not that hard. You figure out who is pretty much locked in. Some people, in the past, disagreed with my locks, but I stuck with them - and was only burned once (Syracuse). Then you go for the teams that are playing the best, beat the best opponents throughout the season, have the best L10, and then use RPI to finalize.

The point that you make is that it's REALLY, REALLY EASY to predict these teams if you follow college basketball. I would agree. Yet Steve is trying to demonstrate how easy this is by providing information that he got 79% of the entries correct. 79%. A C+ in school, if you will. That basically means he got all the "locks" correct and missed on every single bubble team available. That's not impressive, nor does it prove Lunardi is overpaid. It just shows that Steve isn't very good at predicting which teams will make The Dance.

I'm not trying to call out Steve on this one - I think he is very, very, very knowledgeable about sports, specifically college football, but to create a thread boasting that it's an easy thing to do, then failing at the aforementioned subject matter seems, well, setting yourself up.

Maybe I follow college basketball closer than Steve, I don't know. But if you were to take my results and show that Lunardi is overpaid fine, you can probably do that. But with Steve's results, it just proves that maybe Lunardi knows what he's talking about, or that Deace isn't good at predicting the field.

Just my two cents.
 
You must log in or register to reply here.

Latest posts

Share this page

Top