Rankings?

[davidlee]

I went through the different rankings this morning and found the following for SLU:

25 Nolan Power Index(NPI)

31 Sanagrin

36 Pomeroy

37 Massey

39 BPI ESPN

55 RPI

Could someone explain the difference in these rankings and which ones are more creditable?

[kwyjibo]

See post 34 and other posts in this thread:

http://www.billikens...ic=21823&page=2

All the rankings are created for different purposes although there is a general similarity between ranking types (like MOV/Power rankings and Efficiency rankings). There are even hybrid types like ESPN's BPI, I think Pomeroy deserves a lot of respect because he had an idea for a different type of ranking and tested it on years worth of data (most of the rankings are done with less thought by people who are just good at engineering problems). The LRMC rankings are also very good but they do not

These rankings may or may not intend to be predictive (Power rankings like Nolan Power Index were developed by gamblers in order to be predictive and generally are decent at that). With a lot of data rankings that are not intended to be predictive, like RPI which is intended solely as scoring of past accomplishments with a season's worth of games, can be somewhat predictive. RPI does not add in Margin of Victory by design and as such will always be less accurate predicting but better as a reward for winning games.