RPI 101

Posted by Cougar:

 

 Well, it’s back. I am using the RPI system (ratings percentage index) to rate each team in the KVKL. To me, this is the most objective method to rank teams one through 28. I first instituted this system last year and it had an impact (how big it was only lies in the committee members) in placing each team in the appropriate spot in the playoffs. This is a similar “RPI” system used by the NCAA for the men’s and women’s basketball teams. This system does not factor the margin of victory or loss (in each game) in the formula to rate each team. It ONLY oloks into wins and losses.

            Now that we have the preamble of the RPI system out of the way, I’m now going to explain how the RPI works. So get your pens (yes, I did mean to use the plural form!) ready and steno pads. Class is in session:

            The RPI formula is calculated by taking 25% of one team’s winning percentage + 50% of the team’s strength of schedule (average winning percentage of opponents) + 25% of the opponents’ strength of schedule (opponents’ opponents’ average winning percentage).

            The winning percentage of each team is rather simple. You take the number of wins and divide that by the total number of games played. That is the wining percentage component. And don’t make me explain why this component is important. That’s like me explaining why pigs can’t fly!

            To measure a team’s strength of schedule, one must take the total number of games won by the opponents divided by the total games played by each opponent WITHOUT THE RESULT OF THE GAME AGAINST THE TEAM.

Why is that? If a team beats an opponent and that loss that opponent suffered, from that team, is accounted in its record. That loss will lower the team’s opponents’ average winning percentage. So that win would hurt that team’s “strength of schedule” component.  At the same token, a team losing an opponent would put a win for the opponent. That would increase the opponents’ average winning percentage. So that team’s loss would help their “strength of schedule” dynamic. Either way, it would not make sense to include the team’s results against that opponent (or any other opponent in their schedule) in the strength of schedule factor.

            Why is strength of schedule important? Well, it validates how tough your opponents are. It can potentially break a tie between two (or more) teams who may share the same record. If a team has a less than stellar record but has a tougher opponents (by team record), then that team can fairly be placed in the rankings. On the other hand, if a team has a good record but played against opponents who are not as strong, then that team should be in a lowered, but appropriate placement in the rankings.

            The final component to the RPI formula is the opponents’ strength of schedule (opponents’ opponents’ average winning percentage). This is very similar to calculating the team’s strength of schedule, but now the opponents of the team’s opponents are tested here. Again the average winning percentage is computed. The game results of the team’s opponents against their opponents are excluded. There is one good thing for a team to help one’s cause.

For example: if Team A (aka your team) played Team B (the game result between the two is irrelevant), and Team B lost to Team C, but your team (Team A) defeated Team C. To measure the opponent’s opponents’ average winning percentage, the game result between Team B and Team C has to be left out (to see why, see the reasoning in the paragraph that starts with the phrase “To measure…”). But every other game result by Team C has to be accounted for. This includes Team A’s (your team) win over Team C. That is how the last component helps your own cause to have a better RPI. Not only does your team’s win over Team C count toward your own winning percentage (ala the first component of the RPI formula) but it counts again for the third and final component, your opponents’ opponents’ winning percentage. Of course the tables could be turned against you if your team (Team A) lost to Team C. If this were to happen, it would damper your RPI rating.

Why are the opponents’ opponents’ winning percentage important? One might think that just simply taking the team’s opponents’ average winning percentage is enough to validate a team’s rating. However, it’s important to check to see if each opponents’ schedule is tough or not (or in between). This validates the opponents’ rating as well. This will further distinguish one team from the other on who should be rated higher in the rankings. To me, the more VALID components used to rate each team, the better, more objective the ratings should be.

Once get the game results sort out, again take the formula into effect. Take your team’s winning percentage and multiply by 25% (.250). Take your strength of schedule (opponents’ winning percentage) and multiply by 50% (.500). Take your opponent’s strength of schedule (opponent’s opponents’ winning percentage) and multiply by 25% (.250). After you the multiplication of the components, add the three components and you get YOUR TEAM’S RPI.   

Do I think the RPI is perfect? No. The one flaw to the system is that the RPI formula accentuates the schedule a bit too much. Not every team in the KVKL has the same schedule. If each team were to play every opponent in the league, then the rankings would be much clearer. But that would mean each team would have to play 27 opponents. That would mean the regular season would end in the last week of November! The RPI system is a tool. It is not the absolute automatic cure. It is a tool that, if used well, can help the KVKL rate teams for the playoffs.

Class dismissed……..