The first state basketball tournaments with the WIAA’s new RPI ranking system has come and gone, and there seems to be a consensus developing on how it worked.
While most agree there are flaws in the RPI rankings system and how it was used this first year, there is also an agreement that it worked better than the previous random draw. It is considered a solid first step in the right direction.
Now comes the next step: tweaking the system.
I think the system requires more than a “tweak.” Yes, the RPI system performed admirably in this first go. But it would be misguided to ignore some significant holes in the system.
The most glaring is represented by how the Foss boys team, which won the 2A state title and was expected to do so by most metrics, could come in 20th in the RPI in Class 2A.
And what’s worse, even if the Falcons had finished 20-0 instead of 15-5, they would not have been ranked higher than eighth, based on their strength of schedule.
The Hoquiam boys, who ranked 35th, could not have been ranked higher than eighth. The Prairie girls, who ranked 17th, could not have been ranked higher than eighth. The La Center girls, who did finish the regular season unbeaten, ranked seventh.
It’s not a good system when teams are being held back by circumstances beyond their control.
But instead of purely criticizing the system, I’ll instead offer some suggestions on how to fix the RPI system to make it better for 2018.
FIX THE FORMULA: The RPI system is based on a formula the weighs a team’s winning percentage (25 percent), the winning percentage of its opponents (50 percent) and the winning percentage of its opponents’ opponents (25 percent). That means three-quarters of a team’s RPI is based on what other teams do.
The RPI system was originally developed to rank college teams. But high school sports are much different. High school teams have much less control of their schedule. Some were being hampered by playing in a weak league, which in the case of Foss represented 70 percent of its schedule. Other teams’ non-league schedule can be hampered by finances and geography. So the formula should be weighted to better judge teams on what they do. Perhaps a 40-40-20 mix, maybe even 50-30-20.
COUNT OUT-OF-STATE FOES PROPERLY: The WIAA’s RPI committee didn’t know how to deal with the accurate reporting of out-of-state teams’ winning percentage. So they decided to give out-of-state teams a default winning percentage of .500.
That hurt teams, like Foss, that seek quality competition with out-of-state holiday tournaments. The Falcons played very good teams from Las Vegas, Phoenix, California and Oregon at a holiday tournament, but didn’t receive credit for playing quality opponents. The information is there, and all it requires to attain is a phone call or email. Simply require that teams that play out-of-state foes to supply accurate information on those teams so their RPI can be factored correctly.
COUNT POSTSEASON GAMES: The RPI committee also became concerned that, because of different district tournament formats, some teams only play two postseason games between the regular season and state, while others can play as many as six. They were worried that this could give an unfair advantage in their RPI ranking.
I find this found funny, because this imbalance has existed for years, and no one seemed to care. The WIAA has a rule that restricts teams to a 20-game regular season. If a team accidentally plays a 21st game, they could be disqualified from advancing to state. But in the postseason, it’s perfectly fine for some teams to play six playoff games against quality opponents in the walk-up to state, while others only play two. That is a tangible advantage, and it’s been going on for years. Given that the RPI system is based on percentages, these extra games would have much less impact that the actual impact of getting to play an extra three or four high-intensity playoff games.
OK fine. If you still want to even this issue out, simply count the last two or three playoff games each team plays. If a team plays six postseason games, the first three would not be counted. But the last three — games like district semifinals and placing games — would be counted, as they should be.
WEIGHTED WINNING PERCENTAGE: One of the things that always bothered me is that games against 4A opponents were counted exactly the same as games against 1A teams. And I think we can all agree that those games are not the same. When I looked at teams that appeared to be over-ranked in the RPI, I noticed those teams played a lot of games with teams from lower classifications. Meanwhile, under-ranked teams played a lot of games from higher classifications. A system needs to be devised that gives teams more credit for wins against teams from higher classifications, and less credit for wins against teams from lower classifications. For example, a win over a team one classification higher would be worth 1.2 wins. A win over a team two classes higher could be worth 1.4 wins. A win over a team one class lower would count as 0.8 of a win. Or something along those lines.
These are just some of the factors that need to be considered — the major ones. There were also complaints about teams from the same league meeting in the regional round. While this is not ideal, you have to remember that the more you modify the bracket, the more problems you create.
It’s a lot to digest and consider. But sometimes, that’s what is required to get something right.
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 his Twitter handle @360TMart.
