KenPom Wichita State Projection

[_kai_]

http://kenpom.com/team.php?team=Wichita%20St.

Well he has us projected at 20-7 however, he states "Projected record is based on cumulative probabilities of winning each game and may not equal the individual game predictions." If you go based on his individual game predictions, he predicts we go 25-2. With our only OOC loss being Tulsa, and our only MVC loss being at UNI.

Interesting is about all I can say.

[shock]

I like it.

[_kai_]

Something of note, he projects Creighton at 17-12.

Also the conference projection is very interesting.

[_kai_]

Based on his individual game projections, our RPI at large sheet would look like this:

Top 50 Wins: 2
Top 50 Losses: 0

51-100 Wins: 6
51-100 Losses: 2

[RoyalShock]

The thing to keep in mind with KenPom is that it is a predictive system, which is why it actually gives a projected score. The reason the projected record may be different than the record of individual game predictions is that he incorporates "consitency" and "luck" into the projections.

Take last season for example. I don't think we won any games we expected to lose, but we lost five that we were favored in (MSU, UNI, SIU, VCU at home and InSU in the tournament). So if you looked game-by-game, we would have expected only 3-4 losses (UConn, SDSU, MSU away and maybe the finals of the MVC tourney). Yet we ended up with eight.

[_kai_]

Maybe the luck will fall the other way this year, and we end up winning a few games he expects us to lose.

[The Mad Hatter]

Does anyone know how KenPom tries to deal with the issue of newcomers in making these predictions before a single game of the season has been played?

[RoyalShock]

Does anyone know how KenPom tries to deal with the issue of newcomers in making these predictions before a single game of the season has been played?

Here is his ratings explanation. Unfortunately, it doesn't go into players.

http://kenpom.com/blog/index.php/weblog/ratings_explanation/

[The Mad Hatter]

Here is his ratings explanation. Unfortunately, it doesn't go into players.

http://kenpom.com/blog/index.php/web...s_explanation/

I get how his system works, I just don't get how it works in the preseason. Once there is a sample of data for any given year's version of a team (which can vary dramatically from year to year with the turnover in college sports) the whole thing makes sense. But how do you even calculate things like offensive efficiency when you haven't seen this collective group perform any offense and can't adjust it for quality of opponent, since they have faced no opponents?

I grasp that you can estimate based on last year's teams performance minus lost players, but I'm skeptical of predictions without any data taken from the actual combination of players a team will actually be fielding in the current year.

[martymoose]

Keep in mind 3 games are not factored into the totals. 2 in Puerto Rico and the Bracketbuster.

[DoubleJayAlum]

Does anyone know how KenPom tries to deal with the issue of newcomers in making these predictions before a single game of the season has been played?

It doesn't. In fact, it completely ignores them all together. He has essentially admitted such in answers to tweets...

[RoyalShock]

FYI, Pomeroy's site will be subscription-based this year. Only the main ratings page and the blog will be free. Drilling down to team pages and the like will cost you.

http://kenpom.com/blog/index.php/weblog/an_important_message/

[SubGod22]

Then I guess it's a good thing I've never been into KenPom. Is his info really so great that people will pay for it?

[RoyalShock]

Pomeroy is kind of like the Baseball Reference of college basketball. He goes into deeper statistical analysis so you don't just get the usual player stats, but insight into them. In the link I included he says he's pretty much saturated his intended target audience.

[_kai_]

His site is great. His statistical analysis goes much deeper than just face value statistics of ppg, etc. He's actually really accurate with projections as well. It's very very interesting. He also writes great articles about everything college basketball. Would I pay for it? Probably not. I do enjoy reading it though.