In a couple of weeks, the Washington Interscholastic Activities Association’s first venture into using a ranking system to seed teams into the state basketball tournaments will be put into action.
And as we’ve seen, the state’s use of a Ratings Percentage Index, or RPI, leaves a lot to be desired.
Part of the appeal of RPI is its simplicity. In Washington, RPI is made up of three components: a team’s win percentage (worth 25 percent), the win percentage of its opponents (50 percent) and the win percentage of its opponents’ opponents (25 percent).
Anyone can track down those numbers, plug it into the equation and figure out a team’s RPI.
Other ranking systems are far more complex, which makes them appear less transparent.
Then the WIAA took the simple RPI formula and made it more simple by eliminating actual win percentages of out-of-state opponents and making no adjustments for games between schools from different classifications.
The problem is the more simple you make a formula, the less accurate you make it.
The WIAA has said it will take a look at its RPI system after this first year and work to improve it year-to-year.
But maybe it’s better to look at lessons learned by other states that use a ranking system. We’ll look at a few now.
One other state uses just RPI to place and seed into state tournaments. Colorado launched the use of RPI and quickly realized the flaw in the system.
The biggest flaw is the overemphasis on strength of schedule. The issues that we are seeing this season in Washington during the basketball season — which has a regular season of 20 games — would actually become monstrous when applied to football and its season that consist of half as many games.
Last fall in Colorado, an 8-2 team missed the state playoffs in favor of three 3-7 teams, because of strength of schedule.
So last month, Colorado voted to modify their RPI equation from the aforementioned 25-50-25 split to 37.5-37.5-25 for football. For baseball and boys soccer, the equation would be 35-35-30.
Other sports will remain with the 25-50-25 equation. However, many of those sports are experiencing their first season with RPI and perhaps some of those flaws have not become completely visible.
But Colorado’s use of RPI differs from Washington’s in two key ways.
First, Colorado counts out-of-state winning percentage with the actual winning percentage. It’s only when going to opponents’ opponents’ winning percentage that the default .500 number is used — the default number Washington uses for both figures.
Secondly, for football only, Colorado uses a weighted formula for wins based on games between teams from two different classifications.
In Oregon, a combination of RPI and the Colley Index is used. The Colley Index was developed by Wes Colley, a statistician with a Ph. D in Astrophysical Science from Princeton.
Like RPI, the Colley Index was originally developed to rank college teams. So its application to high school sports may not be ideal.
I can’t explain how the Colley Index works because I would need to be wicked smart to explain it and you would have to be wicked smart to understand it.
But, basically, it takes a team’s winning percentage and adjusts it according to the team’s strength of schedule.
But in Oregon, the Colley Index only measures games against in-state opponents from the same classification.
Oregon doesn’t have split-classification leagues. Washington does, so the Colley Index might not work in Washington.
Also, Oregon uses a different RPI formula than Washington. Oregon’s RPI is comprised of two components: a team’s weighted win percentage (worth 35 percent) and its opponents’ winning percentage (worth 65 percent).
The weighted win percentage is derived by applying value to wins based on if they came at home or on the road.
Also, out-of-state winning percentages are counted in Oregon. However, it is up to the teams playing out-of-state foes to supply information on those opponents to the state for ranking purposes. The winning percentage of opponents’ opponents was removed from Oregon’s RPI formula because tracking that information, particularly for out-of-state opponents, proved too difficult.
There are other ranking systems out there. For example, Arizona uses a ranking system developed by Max Preps. You can see what those ranking would look like in Washington by going to MaxPreps.com.
But going forward, Washington likely will stick with RPI, although expect a modified version of RPI, particularly before it gets applied to other sports.
Tim Martinez is the assistant sports editor/prep coordinator for The Columbian. He can be reached at 360-735-4538, tim.martinez@columbian.com or follow him on Twitter at @360TMart.
