ESPN's BPI projections

[fomalhaut]

The historical record of 14 through 16 seeds in the NCAA tournament, through last year, is 1 win and 323 losses:

Records for every seed

However, according to ESPN's BPI projections, the 1, 2, and 3 seeds have first-round winning probabilities ranging from .736 (3 Indiana vs. Charlotte) to .982 (South Carolina vs. play-in winner):

BPI predictions

In the first place, it's hard to imagine how they could project such a precise win rate for South Carolina without even knowing its opponent.
In the second place, ESPN is predicting there is just above a 20% chance that all twelve 1/2/3 teams will win: nearly a 1-in-8 chance for another 16 vs. 1 upset, better than 1-in-3 for a 15 vs. 2 upset which has never occurred, and almost 2-in-3 for a 14 vs. 3 which has also never occurred.

I'm certainly not claiming that all of the 1/2/3 seeds are guaranteed winners, and sooner or later there will be a 14 or 15 seed upset. But these win rates seem ridiculously low. By comparison, Massey has the 1 and 2 seeds at 99% or even 100%.

 

[Sifaka]

Just to persuade myself that this wasn't about the Bank of the Philippine Islands, I looked it up.

“The College Basketball Power Index (BPI) is a measure of team strength that is meant to be the best predictor of performance going forward. BPI represents how many points above or below average a team is. Strength of Record (SOR) is a measure of team accomplishment based on how difficult a team's W-L record is to achieve. Game predictions account for opponent strength, pace of play, site, travel distance, day's rest and altitude, and are used to simulate the season 10,000 times to produce season projections. “

 

[and one]

The BPI has UConn's chances of winning against Mercer at 90.4%. I suspect the BPI model is less accurate at the extreme ends of the curve. If UConn plays Mercer 100 times the chance they win 10 is very close to zero. I'll get a good look at Mercer Saturday and see if that changes my mind. 99% chance it doesn't.

 

[RegisteredUconn]

Just to persuade myself that this wasn't about the Bank of the Philippine Islands, I looked it up.

“The College Basketball Power Index (BPI) is a measure of team strength that is meant to be the best predictor of performance going forward. BPI represents how many points above or below average a team is. Strength of Record (SOR) is a measure of team accomplishment based on how difficult a team's W-L record is to achieve. Game predictions account for opponent strength, pace of play, site, travel distance, day's rest and altitude, and are used to simulate the season 10,000 times to produce season projections. “

Interesting, but what is the starting point? Is the assumption that all teams begin the season on equal footing? Surely that can't be the case.

 

[RoadMaster]

Just to persuade myself that this wasn't about the Bank of the Philippine Islands, I looked it up.

“The College Basketball Power Index (BPI) is a measure of team strength that is meant to be the best predictor of performance going forward. BPI represents how many points above or below average a team is. Strength of Record (SOR) is a measure of team accomplishment based on how difficult a team's W-L record is to achieve. Game predictions account for opponent strength, pace of play, site, travel distance, day's rest and altitude, and are used to simulate the season 10,000 times to produce season projections. “

Rats!.. I had hoped that they'd wisened-up and gone with the Boneyard Prognostication Index..

 

[cwaduke]

I just want to watch a game.

 

[Wbbfan1]

Strange on some of ESPN's BPI Rankings show Team X winning but when you check the Tournament odds, team Y has better tournament odds. There's no correlation between the two.

Examples Are
UCF vs Florida
Washington St vs Kansas. St
South Dakota vs Ole Miss

 

[UcMiami]

Strange on some of ESPN's BPI Rankings show Team X winning but when you check the Tournament odds, team Y has better tournament odds. There's no correlation between the two.

Examples Are
UCF vs Florida
Washington St vs Kansas. St
South Dakota vs Ole Miss

I think the actual game odds are based on that single match-up while the tournament odds are based on the odds against the whole field and the ability to pull of a string of 6 'upsets' in a row. By the second measure power conference membership probably trumps an individual game match up.

History shows it is more than 20 years since a power 5 conference or old big east team has played in a championship game with the exception of Uconn. There have been surprise teams in the FF but they have all been power 5 or OBE teams.