NCAA Tournament Projections & BPI

[Xavier_Musketeers]

http://espn.go.com/mens-college-basketball/bpi/_/type/tournament

Xavier is projected to win the big east and be 4 seed in the tournament

[waggy]

If they win on the road they can earn those things. Right now I just hope they beat Auburn.

[Masterofreality]

The BPI is a garbage stat made up because ESPN has to create something unique.

[MCXU]

2 things about that BPI stat that prove it is as useful as how many towels a team brings with them on road games.

1) ACC has 3 #1 seeds (Surprising since it is an ESPN stat).

2) The SEC is projected to get 8 teams in.... not happenening. The range of SEC teams getting in right now is at best and I mean in a perfect SEC dream world 5 and if things fall the wrong way 2. In the end they will likely send 3 teams.... Not 8... Ever... Ever.

[nuts4xu]

What in the hell is a BPI? A new version of the RPI? Is this something else the NCAA committee will say they don't look at - yet all decisions will factor it in?

[STL_XUfan]

2) The SEC is projected to get 8 teams in.... not happenening. The range of SEC teams getting in right now is at best and I mean in a perfect SEC dream world 5 and if things fall the wrong way 2. In the end they will likely send 3 teams.... Not 8... Ever... Ever.

are they dividing Kentucky into 3 teams?

[XUFan09]

There's a reason Kenpom is still weighted by preseason predictions and Sagarin used to not even post new numbers yet. It's too early in the season to make use of these metrics with any seriousness. I haven't looked at BPI's formula to see if it's a "garbage stat," as MOR says, but he's definitely right that it is ESPN trying to compete.

[gladdenguy]

If they win on the road they can earn those things. Right now I just hope they beat Auburn.

I agree. They lose to Auburn and its gonna be a big black eye on the resume. They need to win the next 2 games or else its an unsuccessful Non Conference season.

[PMI]

This is definitely the best ranking system out there, without a shadow of a doubt, as evidenced by the fact that it is extremely favorable toward Xavier. The other rankings should take note.

[PMI]

I agree. They lose to Auburn and its gonna be a big black eye on the resume. They need to win the next 2 games or else its an unsuccessful Non Conference season.

I agree. I think we will win the next two games, but if we don't, that would be really bad. If only we could've held on in that UTEP game. Then, we are either undefeated or our only loss is to an undefeated top 20 UW team. Then we actually could reasonably be argued as the 15th best team maybe. Dammit. Lot of good opportunities in league play this year though.

[Xtemporaneous]

What in the hell is a BPI? A new version of the RPI? Is this something else the NCAA committee will say they don't look at - yet all decisions will factor it in?

Well here's ESPN's explanation - http://espn.go.com/mens-college-basketball/story/_/id/7561413/bpi-college-basketball-power-index-explained

[paulxu]

Well, any system that uses Melo as a basis for their calculations has to be great. Right?

[xu82]

This is ridiculously early, and I have zero idea what a BPI is or how it's calculated, but since I like the outcome I think it's GREAT!

[toledodan]

What in the hell is a BPI? A new version of the RPI? Is this something else the NCAA committee will say they don't look at - yet all decisions will factor it in?


have no idea but i think we are in the top 20 so i will take it. :homer:

[X Factor]

Kenpom has Xavier #25

Sagarin has Xavier #22


My take, a lot of the computer models like Xavier so far this year. If the defense improves, Xavier could be really, really good.

[xu82]

Kenpom has Xavier #25

Sagarin has Xavier #22


My take, a lot of the computer models like Xavier so far this year. If the defense improves, Xavier could be really, really good.

I wonder if these computer models are as nervous about Auburn as I am.

[bleedXblue]

I wonder if these computer models are as nervous about Auburn as I am.

LOL....I'm right there with you.

[XUFan09]

One thing that is a cool idea with BPI is de-weighting games with key players missing. It's a cool idea, but it has the potential to be a statistical mess, which is likely why Kenpom and Sagarin still avoid it. I'm skeptical of how well BPI can measure players' absences and suspect the resulting model would be pretty clunky. Remember, the data set for college basketball is much smaller than the NBA, both in minutes played and in games played, so individual player impacts are harder to do. This is why adjusted plus/minus is somewhat worthwhile in the NBA but next to worthless in college.

One thing that they do that sounds logical is including diminishing returns for blowouts. It sounds logical, but I read up on some studies by Ken Pomeroy last year where he showed that the idea doesn't bear out statistically. There is still a significant difference between a 40-point blowout and a 30-point blowout, for example, as it is really, really hard to beat up on teams proportionally more when you get to that point.

[Olsingledigit]

One thing that is a cool idea with BPI is de-weighting games with key players missing. It's a cool idea, but it has the potential to be a statistical mess, which is likely why Kenpom and Sagarin still avoid it. I'm skeptical of how well BPI can measure players' absences and suspect the resulting model would be pretty clunky. Remember, the data set for college basketball is much smaller than the NBA, both in minutes played and in games played, so individual player impacts are harder to do. This is why adjusted plus/minus is somewhat worthwhile in the NBA but next to worthless in college.

One thing that they do that sounds logical is including diminishing returns for blowouts. It sounds logical, but I read up on some studies by Ken Pomeroy last year where he showed that the idea doesn't bear out statistically. There is still a significant difference between a 40-point blowout and a 30-point blowout, for example, as it is really, really hard to beat up on teams proportionally more when you get to that point.

Statistics always in any setting eliminate the human element. I work in an area where a lot of statistical models are used by some but the results do not in any way factor in the professional judgment that is needed for a good answer. In the case of weighting blowouts, one problem I see in trying to prove whether statistically they diminish or not is the fact that some teams will pull the dogs off and end up with a much lower margin than had they kept the dogs on. In other cases, depth might skew the results, i.e. teams with a lot of depth might end up with higher margins. However, that doesn't necessarily translate to more wins in tighter games since coaches often play only 7-8 players regularly regardless of depth.

[waggy]

In my view the blowout games are a problem. I've never gotten the feeling that Mack has run up a score. But I think Calipari and Pitino have.

[Caveat]

In my view the blowout games are a problem. I've never gotten the feeling that Mack has run up a score. But I think Calipari and Pitino have.

It's not like Mack calls the dogs off either -- if anything, I sometimes think he leaves starters in too long in games where the outcome has been decided.

[waggy]

I don't know how you avoid it though either. The cat's out of the bag... The committee unofficially looks at the numbers that come from scoring margins.

[paulxu]

It's not like Mack calls the dogs off either -- if anything, I sometimes think he leaves starters in too long in games where the outcome has been decided.

It would be nice to have, oh, say, about 27 of these games every year.

[xu82]

It's not like Mack calls the dogs off either -- if anything, I sometimes think he leaves starters in too long in games where the outcome has been decided.

The better and deeper you are the harder this is to avoid if you even want to. Think about Kentucky as the extreme example.... they only have a few guys who aren't McDonalds All Americans or a lottery pick! That's just crazy talent, but the type of crazy I'd like more of at X. Heading in the right direction.

[mid major]

LOL....I'm right there with you.

I'll join you two. Losing to Auburn would suck bigtime and I'm a little nervous about this game.

[gladdenguy]

I'm really nervous about this Auburn game. Maybe all the nervous people could text me updates on Saturday because I will be recording and trying to spend my time doing other things.
If they win I will watch Saturday night when I get home. Lose and I will erase it so fast.

[xu82]

I'm really nervous about this Auburn game. Maybe all the nervous people could text me updates on Saturday because I will be recording and trying to spend my time doing other things.
If they win I will watch Saturday night when I get home. Lose and I will erase it so fast.

Were you nervous about the game before? Will you be nervous about the game after? There might be medications to help with that. I will be at the game (with an Auburn fan!) and I'll have to put up with him on the ride back if we lose. I'm not sure what would be worse - the game or the ride back. So I'm doubly nervous!!!

[xuwin]

It's not like Mack calls the dogs off either -- if anything, I sometimes think he leaves starters in too long in games where the outcome has been decided.

When do you think he has done that? He is constantly subbing his starters throughout the games.

[muskiefan82]

I'm really nervous about this Auburn game. Maybe all the nervous people could text me updates on Saturday because I will be recording and trying to spend my time doing other things.
If they win I will watch Saturday night when I get home. Lose and I will erase it so fast.

Just for the record about being nervous about this game- I think a point should be made.

32 years ago when X went on the road ANYWHERE I was nervous about the game. Now, I expect X to beat teams on the road from non-power 5 conferences, feel only slight apprehension for lower level power 5 teams, and reserve nervousness for the high major competition only.

X really has come a long way AND, if you think about it, your nervousness comes from you expecting X to win as well and worrying that they will not meet your current expectations.

[UCGRAD4X]

In my view the blowout games are a problem. I've never gotten the feeling that Mack has run up a score. But I think Calipari and Pitino have.

Did John and Rick tell you they think Mack runs up the score. Or do you just think they think?