To me, it was always intuitively obvious that a team’s record versus the RPI Top-50 was a key stat to use when evaluating a team or when evaluating a coach over time. Anyone can beat the Stetsons of the world (current RPI – 326), but it takes something more to beat good teams. However, some people claim that my use of this stat is misleading. So, let’s look a little closer at RPI rankings and won/loss records against the RPI Top-50.
From the NCAA, I have the Final Four teams from 1999 through 2005. From Ken Pomeroy’s site, I have the final RPI for each team and their record versus the RPI top-50. Let’s see if we can find any correlation:
|
|
1999 |
RPI |
Record vs |
|
Champion |
|
3 |
10-2 |
|
Runner-Up |
Duke |
1 |
16-0 |
|
|
|
2 |
12-4 |
|
|
|
20 |
9-3 |
|
|
|
|
|
|
|
2000 |
|
|
|
Champion |
|
13 |
10-6 |
|
Runner-Up |
|
18 |
6-7 |
|
|
|
41 |
3-8 |
|
|
|
32 |
8-8 |
|
|
|
|
|
|
|
2001 |
|
|
|
Champion |
Duke |
1 |
15-4 |
|
Runner-Up |
|
8 |
9-4 |
|
|
|
22 |
7-8 |
|
|
|
3 |
10-4 |
|
|
|
|
|
|
|
2002 |
|
|
|
Champion |
|
2 |
9-3 |
|
Runner-Up |
|
13 |
7-7 |
|
|
|
1 |
7-1 |
|
|
|
5 |
7-4 |
|
|
|
|
|
|
|
2003 |
|
|
|
Champion |
|
9 |
8-4 |
|
Runner-Up |
|
6 |
7-6 |
|
|
|
10 |
5-3 |
|
|
|
4 |
8-6 |
|
|
|
|
|
|
|
2004 |
|
|
|
Champion |
|
5 |
10-6 |
|
Runner-Up |
Georgia |
16 |
8-7 |
|
|
Duke |
1 |
10-4 |
|
|
|
6 |
8-2 |
|
|
|
|
|
|
|
2005 |
|
|
|
Champion |
|
6 |
6-3 |
|
Runner-Up |
|
2 |
11-0 |
|
|
|
21 |
6-4 |
|
|
|
12 |
7-2 |
Observations from these seven years:
– 18 of the 28 Final Four teams were ranked in the RPI Top-10 (64%).
– 26 of the 28 Final Four teams were ranked in the RPI Top-25 (93%).
– 10 of the 14 teams in the championship game were ranked in the RPI Top-10 (71%).
– All 14 teams in the championship game were ranked in the RPI Top-25.
– All seven NCAA champions had winning records versus the RPI Top-50.
– Only 3 Final Four teams had losing records versus the RPI Top-50.
– The worse rated team to make the Final Four was UNC in 2000 with a RPI ranking of 41 and a 3-8 record against teams in the Top-50.
– The worse rated team to make the championship game was Florida in 2000 with a RPI ranking of 18 and a 6-7 record against teams in the Top-50. This was the only time that a team appeared in the championship game with a losing record versus the RPI Top-50.
Conclusions
Good teams (by any measure you want to use) lose early in the NCAA tournament every year. However, bubble teams and those with poor records against the Top-50 do not make the Final Four either. I think that we remember the big upsets, but forget that the bracket-busters usually lose rather quickly after their big win.
The RPI is not the end-all and be-all of college basketball statistics. It may not even be the best computer formula to use when ranking teams. However, the RPI calculation is far from useless. It is no surprise to me that teams that do well in the Big Dance also have a good record against teams in the RPI Top-50. Doing well in the NCAA tournament is a function of good coaching and good players…not lucky bounces, getting “hotâ€? at the right time, or by getting breaks from the refs.
The idea that any team in the NCAA tournament can “get hot� at the right time and make the Final Four is a fairy tale just like Cinderella. Somehow, too many people forget that Cinderella stumbled and fell at midnight…just like every March when Cinderella teams stumble and fall right after their big entrance. Those addicted to predicting brackets might want to keep all of this in mind next March.
