Boyd's World-> Breadcrumbs Back to Omaha-> RPI Changes for 2004 About the author, Boyd Nation

RPI Changes for 2004

Publication Date: November 18, 2003

One Step Sideways, Please

I found out this week that there are some proposed changes in the baseball RPI on the board, and I want to go ahead and look at the likely outcome while there's still some discussion to go. At this point, the proposal is to be discussed at the upcoming ABCA meetings in San Antonio January 2-5 and then be voted on by the Division I baseball committee afterwards. It appears likely to pass, but lots could happen between now and then, and some of the loudest voices are those likely to be hurt by it.

The basic RPI formula is more or less sacrosanct, not having changed since the early 90's and having had the same form since its inception back in the mid-80's. Each sport, though, is allowed to add small bonuses or penalties to the base formula in order to guide scheduling the way that the governing committee wishes. For the past few years in baseball, there have been bonuses on a graduated scale for beating a team in the top 75 on the road and penalties for losing to a bottom 75 team at home. There are also penalties for playing too many games against non-Division I teams. The effect of the top 75 limit is that most road wins in the SEC or Pac 10 are bonus games. Since the intended effect was to get top programs to play more road games, it hasn't really worked the way it was intended.

The proposal in front of the coaches and committee now is to stop awarding bonuses (and stop charging penalties, which turns out to be significant to some leagues as well) for conference games. The intent (at least as I've heard it through the grapevine) is to get the top conferences, especially in the Southeast, to start scheduling better non-conference schedules. The theory is that if the SEC teams stop being rewarded for just playing their conference schedules, maybe they'll put some meat on the non-conference schedule.

How will this work out in practice? Well, probably not exactly the way they're intending, although the bath water does go out along with the baby. It's a simple code change, so I've run last season through and looked at it a few different ways to see what changes. First off, here are the top 25 from last year under the old and proposed systems:

     Old                        Proposed

  1  Stanford                1  Miami, Florida
  2  Florida State           2  Florida State
  3  Miami, Florida          3  Stanford
  4  Rice                    4  Rice
  5  Cal State Fullerton     5  Cal State Fullerton
  6  Louisiana State         6  Texas
  7  Auburn                  7  Georgia Tech
  8  Texas                   8  Auburn
  9  Georgia Tech            9  Louisiana State
 10  Baylor                 10  Baylor
 11  North Carolina State   11  Long Beach State
 12  Long Beach State       12  North Carolina State
 13  Arizona State          13  Arizona State
 14  Texas A&M              14  Clemson
 15  Clemson                15  North Carolina
 16  North Carolina         16  Texas A&M
 17  Mississippi State      17  South Alabama
 18  Southern Mississippi   18  Nebraska
 19  South Alabama          19  Southern Mississippi
 20  Nebraska               20  Alabama
 21  Alabama                21  Florida Atlantic
 22  South Carolina         22  Richmond
 23  Florida Atlantic       23  Mississippi State
 24  Houston                24  Houston
 25  Virginia               25  South Carolina

First of all, please note that the old listings don't match exactly to the ones I published at the end of the season. I'm going through my annual process of reconciling my version of the RPI's with the NCAA one, and I noticed that I had some of the regional and super-regional games listed as home games when my understanding of the official RPI is that they're counted as neutral-site games no matter where they're played. I'm still in the process of reconciling my results to what Baseball America published and could really use a copy of the official NCAA RPI report; if anyone out there has one and is willing to share, let me know. I'll be glad to pay copying and shipping costs.

Anyway, nobody here moves all that many places. In this range, the absolute magnitude of the RPI values spread out some compared to, for example, the bunch between #80 and #120, so the bonuses don't have as much effect. The independents gain a bit, since there are no conference results to ignore, but the last major independent is going away, so that doesn't matter much. The SEC teams move down a bit, but so do the Pac 10 and ACC teams.

So, who moves the most? Here are the ten (plus ties) teams who move up and down the most:

19 Central Connecticut State
13 Jackson State
12 Southeast Missouri State
11 Quinnipiac
 8 Southern
 7 Xavier
 7 Pepperdine
 7 Mississippi Valley State
 7 Northwestern State
 7 Liberty

-14 Oregon State
-13 Michigan
-12 Duke
-12 Kentucky
 -9 Creighton
 -8 East Carolina
 -8 Washington State
 -8 Brown
 -8 Gardner-Webb
 -8 Mercer
 -8 Arkansas State
 -8 Mississippi

Some of these are just one-year flukes. Michigan, for example, won three of four at Ohio State last year. They lose those bonuses, and they were sitting right at the top of a tight clump, so they fall thirteen spots. The difference in #94 and #107 isn't significant, really. Note that a couple of Pac 10 teams make the most hurt list, which isn't the intent (and which makes the RPI even less useful in evaluating Pac 10 teams).

The list of teams that moves up makes a useful point: It's hard to avoid penalties under the old system when almost your entire conference is in the bottom 75. To illustrate that and some other things, here are the total moves for all conferences:

 Sum Delta  Conference

 -14   24   ACC
   3   17   America East
  14   30   Atlantic 10
 -11   27   Atlantic Sun
  -9   23   Big 12
   9   17   Big East
   8   20   Big South
 -13   31   Big Ten
  -6   18   Big West
 -20   32   C-USA
  15   19   CAA
   6   10   Horizon
 -12   22   Independents
  -4   26   Ivy
  11   17   MAAC
  15   23   MAC
   0   10   MEAC
 -13   29   MVC
   1   11   Mid-Continent
  -5   15   Mountain West
  50   54   NEC
   6   22   OVC
 -39   39   Pac 10
   5   11   Patriot
 -38   48   SEC
  43   45   SWAC
  19   29   Southern
  -4   34   Southland
 -18   26   Sun Belt
  -9   15   WAC
  10   18   WCC

Sum = sum of all changes
Delta = sum of absolute value of all changes

The SEC and SWAC, for example, move almost the same amount, but in opposite directions. The intent to hurt the SEC in the short term in order to affect their scheduling decisions is accomplished (and that's not an unreasonable goal, no matter how it makes some folks feel), but the Pac 10, which already schedules tremendously, is hurt just as much, while the ACC, which schedules almost as badly as the SEC, is less affected.

I don't really have a recommendation on this matter one way or the other, if anyone cares. These changes aren't going to remedy any of the fundamental problems with the RPI, and the system isn't likely to be any more fair with the changes included. On the other hand, the desired goal of getting big-conference Southeastern teams to improve their schedules is a good one, and if the political act of changing the rating system gets that message across, it could be a good thing. Those within the game will have to decide if this tactic is likely to be successful.

Money, Part Three

On a side note, I want to open up the floor for you to tell me what you think about something. While it's not overwhelming, I put a good bit of time and effort into the site. I get a lot out of it, from the enjoyment I get from corresponding from readers to the fun I have doing it to the ability to contribute to a game I love, and I plan to keep doing it however this discussion turns out. However, I'm not allergic to money. As I've said before, the only reason I don't make any money from the site is that I haven't found a way to do so that doesn't take more effort than it's worth financially or that would drive readers away.

There are a couple of options that I've been thinking about, though, and I wanted to see what the general consensus is out there about them. The first of these, the one that pleases me aesthetically the most, is the page sponsorship model used by the marvelous Baseball Reference site. This is vaguely similar to the public broadcasting model, where users donate and get to have their name linked with a particular page. I have some new content in mind that might work well with that, so if you'd be willing to donate, say, $25 (or $10 or some other amount) to sponsor your team's page, for example, for a year, I'd like to know it. Nothing binding, just a survey of interest.

The other possibility is that, while I've always avoided banner ads because of the low payment rate and their visual worthlessness, Google has a new targeted advertising program that actually looks like it might be interesting. The ads are similar to those text ads you get down the right side of the page when you do a Google search and, if possible, are targeted to supplement the content of the page they're on. This strikes me as something that might actually add value to the site while bringing in a little money.

Usually, when I say I want to hear from you, I get one or two replies if I'm lucky. Since this discussion will affect the future of the site in ways we can't foresee just yet, I'd like to hear a bit more than that this time. If you can spare a minute, drop me a line at the address in the About... page below ( and let me know. Since I've always mentally deducted a point from sites I've used if they got too "commercial", I want to know if you think either of these ideas would significantly detract from the site, and I want to know if you think either of them are a good idea.

If you're interested in reprinting this or any other Boyd's World material for your publication or Web site, please read the reprint policy and contact me


Boyd's World-> Breadcrumbs Back to Omaha-> RPI Changes for 2004 About the author, Boyd Nation