Selection Committee Use of RPI

[sciencejay]

This is an excerpt from a column in today's Omaha World Herald in which CU AD Bruce Rasmussen (who is on the selection committee) discusses the committee's use of RPI, KenPom, etc. Interesting and enlightening.


» RPI: “The RPI is like a 5-iron on the golf course,” Rasmussen said. “It’s something I might use. And if I do use it, I hope it works. But we don’t use it all the time. We don’t use it as much as people think. We don’t ever talk about RPI.”

» Other metrics, such as the “Kenpom” rating, go on a “Nitty Gritty” sheet for each school. Rasmussen said the metrics are used by committee members as a guide to form a big-picture, total-résumé opinion of each school.

“For instance, I’m in charge of the Big Ten for the committee,” Rasmussen said. “I might get up and talk about Maryland.

“I won’t mention their RPI. But I’ll talk about their offensive and defensive efficiency, which we get from Kenpom. Good wins, road wins, nonconference schedule, shooting percentage. How often do they get to the line. Are they balanced?”

» Top 50 wins: “We talk about it, but it’s easy to get tricked by the top 50. What’s the difference between No. 47 and No. 53? Those two teams are basically the same.”

» Conference RPI: “It’s never mentioned. But that doesn’t mean it’s not important. Where it’s a factor is this: If you’re in a high-rated league, you have a better chance of getting good wins and less of a chance of a loss that hurts you.”

» Number of conference wins: “It doesn’t come up. It’s hard to measure because it goes from league to league. Again, if you’re in a good league, and you win 10 games, you more than likely have some good wins.

“But I go back to the Big Ten. Until this week, Indiana had not played any of the other top five teams in the Big Ten: Michigan State, Maryland, Iowa or Purdue. And they only play those teams a total of five times, whereas Wisconsin plays them nine times.

“It’s hard to get a read on number of league wins with all the unbalanced schedules.”

In the end, every school is placed on the board and compared to others based on their individual merits. Then, it comes down to the opinions of 10 individuals.


Here's the link: http://www.omaha.com/sports/shatel-febr ... ece00.html

[admin]

Very interesting insight. Thanks for sharing.

[DudeAnon]

How does he explain the comparison sheet that uses nothing but RPI?

[ta111]

Seems what he is saying is they don't look at each teams individual RPI, but use it to determine what a good win or bad loss is. I do agree that you can't just look at top 50, because there is no difference between 50 and 51. RPI is still the dominant metric used.

[billyjack]

Sensible article.
I like that he says that they don't make a big deal about if a team is #47 or #53 as far as the arbitrary "Top-50 wins" thjng. This means we don't have to overobsess about whether Marquette is #102 or #98.

[stever20]

Sensible article.
I like that he says that they don't make a big deal about if a team is #47 or #53 as far as the arbitrary "Top-50 wins" thjng. This means we don't have to overobsess about whether Marquette is #102 or #98.

True- BUT if Marquette is where they are right now- #116- it would matter.

The category that they said a lot last year was teams in the tourney.

I find it VERY interesting they didn't ask him about OOC scheduling impact, given Creighton's situation.

The Indiana/Wisconsin thing is a bit misleading. Indiana can't play themselves. Indiana played the other 4 teams 5 times. Wisconsin played those other 4 teams 7 times. It's not the 9-5 difference that he says....

[stever20]

1 thing- didn't want to start a new thread- but a thought on this year with the RPI.... With SMU and Louisville ineligible, when you see some team like Butler for instance at #58, they are for all intents and purposes #56. Or like Seton Hall- instead of 42, they're 40 for realistic purposes. Why I think Seton Hall is pretty much a lock, and Butler is in good shape.

[stever20]

bump....

[DudeAnon]

Only conclusion to be made based on this year is they will choose whatever stat they can to exclude the non P5 (plus AAC and BE).

Top 50, OOC SOS, RPI, TOP 200 losses. There are enough metrics that they can just cherry pick a bad one for each mid-major bubble team and use that as their excuse. Of course the p5 programs don't have to endure such scrutiny.

[stever20]

Only conclusion to be made based on this year is they will choose whatever stat they can to exclude the non P5 (plus AAC and BE).

Top 50, OOC SOS, RPI, TOP 200 losses. There are enough metrics that they can just cherry pick a bad one for each mid-major bubble team and use that as their excuse. Of course the p5 programs don't have to endure such scrutiny.

South Carolina says hi.

The fact is conferences like the MWC, WCC were down this year. MWC was 12 in RPI, WCC 14.

in Ken pom:
2016-
AAC .6764 #7
MWC .5426 #10
WCC .5254 #11

2015
WCC .5796 #9
AAC .5547 #10
MWC .5131 #12

so MWC improved some- but the WCC fell way back.... Meanwhile the AAC had a much stronger year.