Pages

RPI Matters

3/12/2006

Wow, I'm still in shock over the seedings. To say the Illinois got hosed is the understatement of the year. To get a '4' and end up possibly playing the one team that we match-up worst against in the S16 is a double whammy.

After my e-mail conversations with Craig Littlepage, I came away confident that the Committee was going to start heading in the right direction. The Illini aside, these brackets don't make a whole lot of sense at all. It's hard to determine what the Committee is trying to reward, is it 'body of work'? RPI? last 10 games? regular season or conference tourney? There's no clear cut answer because the factors weren't applied consistently across the board.

Even though Littlepage (and others before him) said during the telecast that RPI doesn't matter, it is clear that it definitely matters at least this year. I didn't hear Littlepage utter the term 'body-of-work' at all. Which is probably understandable because there is absolutely no way that you can justify teams like Tennessee or UCLA ahead of Illinois if you are compare a season's worth of wins and losses. No possible way. But...

RPI
Tennessee 6, SOS 6
UCLA 10, SOS 29
Illinois 14, SOS 54

Even if these numbers aren't anywhere close to true representation of a team's quality of season or its strength of schedule, they paint a much different picture in favor for Tennessee and UCLA. If you looked how a team actually performed relative to its RPI, most of the teams that outperformed their RPI (ie their RPI was underrated) were seeded lower than expected or left out completely, and many teams that underperformed their RPI were seeded much higher or got in at the expense of others.

(I'll compare RPI to my JCI rankings that was built to compare one team's W-L's versus another, and the Sagarin-Elo ranking, which is very similar. Both are very good body-of-work rankings)

Performance Exceeded RPI
1. George Washington, RPI: 37; JCI: 9, Sag: 9 (8 Seed)
2. Notre Dame, RPI: 95; JCI: 55; Sag: 59
3. Connecticut, RPI: 3; JCI: 1; Sag: 1 (1 Seed)
4. Penn State, RPI: 106; JCI: 67; Sag: 69
5. Florida St, RPI: 63; JCI: 37; Sag: 40 (Out)
6. Bucknell, RPI: 42; JCI: 22; Sag: 23 (9 Seed)
7. Louisville, RPI: 71; JCI: 47; Sag: 47 (Out)
8. Northwestern, RPI: 102; JCI: 74, Sag: 78
9. Nebraska, RPI: 96; JCI: 70; Sag: 69
10. North Carolina St, RPI: 51; JCI: 25; Sag: 25 (10 Seed)
11. Georgetown, RPI: 36; JCI: 18; Sag: 17 (7 Seed)
12. Georgia, RPI: 109; JCI: 81; Sag: 82
13. West Virginia, RPI: 38; JCI: 19; Sag: 20 (6 Seed)
14. Gonzaga, RPI: 9; JCI: 5; Sag: 5 (3 Seed)
15. Boston College, RPI: 22; JCI: 14; Sag: 14 (4 Seed)
16. Illinois, RPI: 14; JCI: 7; Sag: 7 (4 Seed)
17. DePaul, RPI: 91; JCI: 72; Sag: 72
18. Arkansas, RPI: 45; JCI: 26; Sag: 27 (8 Seed)

Performance Underexceeded RPI
1. Houston, RPI: 54; JCI: 88; Sag: 90
2. Montana, RPI: 61; JCI: 99; Sag: 91
3. Arizona, RPI: 23; JCI: 51; Sag: 50 (8 Seed)
4. Wis-Milwaukee, RPI: 53; JCI: 79; Sag: 76 (11 Seed)
5. Utah St, RPI: 46; JCI: 73; Sag: 65 (12 Seed)
6. Davidson, RPI: 94; JCI: 126; Sag: 131
7. Western Kentucky, RPI: 55; JCI: 77; Sag: 74
8. San Diego St, RPI: 56; JCI: 76; Sag: 73 (11 Seed)
9. Oklahoma, RPI: 16; JCI: 28; Sag: 30 (6 Seed)
10. George Mason, RPI: 26; JCI: 45; Sag: 45 (Out)
11. Saint Joseph's, RPI: 43; JCI: 57; Sag: 55
12. UNC Wilmington, RPI: 28; JCI: 44; Sag: 44
13. UCLA, RPI: 10; JCI: 21; Sag: 19 (2 Seed)
14. Butler, RPI: 82; JCI: 111; Sag: 106
15. Louisiana Tech, RPI: 68; JCI: 91; Sag: 87
16. Murray St, RPI: 65; JCI: 86; Sag: 81
17. Bradley, RPI: 33; JCI: 49; Sag: 49 (13 Seed)
18. Missouri St, RPI: 21; JCI: 40; Sag: 39 (Out)
19. Southern Illinois, RPI: 29; JCI: 43; Sag: 43 (11 Seed)
20. BYU, RPI: 67; JCI: 87; Sag: 86
21. Tennessee, RPI: 6; JCI: 16; Sag: 15 (2 Seed)
22. South Alabama, RPI: 66; JCI: 85; Sag: 85
23. Kent St, RPI: 48; JCI: 61; Sag: 57 (12 Seed)
24. Pennsylvania, RPI: 98; JCI: 119; Sag: 113 (15 Seed)
25. New Mexico St, RPI: 97; JCI: 117; Sag: 115
26. VCU, RPI: 81; JCI: 104; Sag: 105
27. Nevada, RPI: 19; JCI: 31; Sag: 26 (5 Seed)
28. UTEP, RPI: 76; JCI: 98; Sag: 97
29. Air Force, RPI: 50; JCI: 58; Sag: 58 (13 Seed)

Last year, I believe that the Committee relied on the Sagarin Elo ratings during the selection process because the home/road adjustments they implemented for the first time were causing some unintended results. Thus there were far fewer 'head scratchers' in the seedings. This year, even though the same unintended results could occur in the formula, I think they only had the new RPI in the room. Without the old RPI for context, there probably wasn't any motivation to go to the Sagarin or any other source. They probably just took the RPI at face value.

So if the RPI is here to stay, why do we continue to schedule the way we do? It's pretty clear that other teams have figured it out and have schedule to maximize their RPI. And they reaping the benefits from it. If our goal is to win national championships, then maximizing our RPI has to be at least one of the priorities when it comes to scheduling. A higher RPI can only help you. It can only lead to better seeds and thus better match-ups. If playing Wis-Green Bay, Tennessee Tech, Robert Morris, and Siena instead of S Dakota State, Coppin St, SE Miss St, and Tenn-Martin (a simple change that would little or no impact on our W-L record but boost our RPI from 14 to 7 and SOS from 54 to 29) is the difference between getting a '4' seed and getting a '2' or '3' seed, then I think we have to make that trade-off.

2 comments:

  1. Bubbleteams said...:

    comparisons of most of the brackets are up at http://bubbleteams.blogspot.com/

  1. pokerplayr said...:

    your conversations with craig littlepage ??

    must have missed that interview after the brackets went up

    did they squeeze you in between espn and sports illustrated ?

 
Wegoblogger #31 © 2011 | Designed by Bingo Cash, in collaboration with Modern Warfare 3, VPS Hosting and Compare Web Hosting