Some of you may have seen the announcement in your local paper on Saturday. The RPI has been improved, now with 150% of your daily recommended road-game value!
The Division I men’s basketball committee has altered the Ratings Percentage Index, used to evaluate teams for selection to the NCAA Tournament, to give more weight to playing and winning road games.
The RPI will continue to consist of 25 percent winning percentage, 50 percent opponents’ winning percentage and 25 percent opponents’ opponents’ winning percentage. The new formula will weight each road win at 1.4 and each road loss at .6, while home wins are .6 and home losses are 1.4. Neutral-site games will have a 1.0 value.
Previously, all games were given the same weight. Committee chairman Bob Bowlsby said home teams have won approximately two-thirds of all games over the last 20 years. [Dallas Morning News, 12/18/04]
Previously, I wrote about the new formula. I’ve also written about Mr. Bowlsby’s conclusion that because the home team has won 67% of the games over the past four years, that doesn’t mean that road games are twice as hard to win as home games. By using the factors mentioned above, a home win is worth about 40% as much as a road win. This is not commensurate with the difficulty of winning a road game, especially for an at-large quality team, but they didn’t ask me.
Now that the ingredients have been revealed I will be posting this version of the RPI on the main page. There is one possible bit of confusion. Will they do it the way I tested in that earlier post, or will they do it the way that makes sense? The way I talked about before is to only apply these adjustments to Part I of the RPI. The sensible way is to apply the adjustments to all three parts. To review:
Part I (25%) – Winning percentage.
Part II (50%) – Average opponents’ winning percentage.
Part III (25%) – Average opponents’ opponents’ winning percentage.
I showed earlier what happened to last year’s RPI when adjustments were only applied to Part I. Teams that were helped (or hurt) equally represented the power conferences and the mid-majors.
But when you apply the adjustments to all three parts, the results are what you would expect given the intentions here: teams at the top of mid-major conferences get a bonus. Teams in the middle of power conferences get penalized.
Here’s how this formula would have affected last year’s RPI
Teams most helped:
Austin Peay (94 in the basic RPI to 61 in the road-rage RPI)
Louisiana Lafayette (71 to 47)
Wisconsin Milwaukee (75 to 53)
George Mason (81 to 60)
East Tennessee State (61 to 42)
Illinois Chicago (87 to 69)
Troy State (66 to 49)
Murray State (58 to 41)
Air Force (70 to 54)
Western Michigan (44 to 28)
Nevada (29 to 14)
Kent State (76 to 63)
Pacific (65 to 52)
Utah State (43 to 31)
Power conference teams get killed by this formula. The biggest jump a team from a top-six conference made was four spots. Would this have changed the behavior of the selection committee? It’s hard to say, but knowing that Utah State was the last team left out, you’d have to think they would have made it with a 31 RPI. Now the teams that were hurt…
Teams most hurt:
Virginia (52 to 84)
Georgia (48 to 76)
Rutgers (56 to 75)
Louisiana State (38 to 55)
Notre Dame (49 to 66)
Saint Louis (64 to 80)
Villanova (67 to 83)
Michigan (55 to 70)
Alabama (26 to 40)
Florida State (53 to 67)
DePaul (37 to 48)
Washington (60 to 71)
South Carolina (45 to 56)
At least this list makes sense. Most of the teams on here were considered overrated, they failed miserably on the road or padded their schedule with a ton of early home games. Now South Carolina’s bid doesn’t look so safe. If the new formula was used last season, it’s possible, maybe even probable, that USC would have been replaced with USU in your brackets. With Notre Dame’s presence on this list, I am obligated to reprint coach Mike Brey’s reaction to hearing about the new formula:
"If we had that formula last year, we’re a seven seed." [South Bend Tribune, 10/23/04]
Just saying it doesn’t make it true, Coach B.
In 2004 – and I’d like to look at previous seasons when I get time – one mid-major school may have replaced a power conference school. While fans of the Gamecocks would not have been pleased with this development, the rest of America probably wouldn’t have minded. Even someone like me, who doesn’t buy all of the mid-major claims of discrimination, wouldn’t have a problem with a mid-major bubble team getting the benefit of the doubt.
The wierd thing about the road-rage RPI is the mid-major schools get rewarded for road wins in their conference, which are easier to get than for the same caliber of team in a power conference. Utah State was the last team left out last season, and the reason was that their non-conference schedule was weak. So it’s a little unusual that the NCAA is behind a formula that would have given the Aggies that last boost needed to get in the field because of their play in conference. Maybe this is an appropriate leveling of the playing field, though. I’m willing to have an open mind.
So, what’s the bottom line here? Assuming that the NCAA is using this version of the calculation, this is a step in the right direction. Maybe a step too far, but a step nonetheless. I was never a proponent of tinkering with the RPI. Hopefully we have not just witnessed the beginning of the "BCS-ization" of the formula, where each year a new formula is used based on the apparent weaknesses of the previous season.