NCAA RFI Formula

[SoAZDad]

Oops. That's RPI (thanks Riseball).

Anybody know what it is? I poked around a little and it seems like it may not be public info. Seems like records (overall, road, neutral and home) are taken into account, with more points given for beating teams with higher power rankings, and the opponents of the opponents. I'd love to see the formula though.

Thanks.

 

[riseball]

I think you are referring to the Rating Performance Index (RPI) that NCAA uses to rank various sports based on strength of schedule. A Google search will cough up the formula.

 

[SoAZDad]

I think you are referring to the Rating Performance Index (RPI) that NCAA uses to rank various sports based on strength of schedule.

Correction noted, thanks. I found the Rating Percentage Index described on Wikipedia. They gave a formula for NCAA baseball, so I'll assume the formula for softball is identical or similar.

 

[riseball]

It is pretty straight forward but requires a pretty extensive database.

The current and commonly used formula for determining the RPI of a college basketball team at any given time is as follows.

RPI = (WP * 0.25) + (OWP * 0.50) + (OOWP * 0.25)

where WP is Winning Percentage, OWP is Opponents' Winning Percentage and OOWP is Opponents' Opponents' Winning Percentage.

The WP is calculated by taking a team's wins divided by the number of games it has played (i.e. wins plus losses).

For Division 1 NCAA Men's basketball, the WP factor of the RPI was updated in 2004 to account for differences in home, away, and neutral games. A home win now counts as 0.6 win, while a road win counts as 1.4 wins. Inversely, a home loss equals 1.4 losses, while a road loss counts as 0.6 loss. A neutral game counts as 1 win or 1 loss. This change was based on statistical data that consistently showed home teams in Division I basketball winning about two-thirds of the time.[2] Note that this location adjustment applies only to the WP factor and not the OWP and OOWP factors. Only games against Division 1 teams are included for all RPI factors. As an example, if a team loses to Syracuse at home, beats them away, and then loses to Cincinnati away, their record would be 1-2. Considering the weighted aspect of the WP, their winning percentage is 1.4 / (1.4 + 1.4 + 0.6) = 0.4117

The OWP is calculated by taking the average of the WP's for each of the team's opponents with the requirement that all games against the team in question are removed from the calculation. Continuing from the example above, assume Syracuse has played one other game and lost, while Cincinnati has played two other teams and won. The team in question has played Syracuse twice and therefore must be counted twice. Thus the OWP of the team is (0/1 + 0/1 + 2/2) / 3 (number of opponents - Syracuse, Syracuse, Cincinnati). OWP = 0.3333

The OOWP is calculated by taking the average of each Opponent's OWP. Note that the team in question is part of the team's OOWP. In fact, the most re-occurring opponent of your opponents is the team in question.

Continuing the example above, a team has played Syracuse twice and Cincinnati once. Syracuse has played one other game and lost, while Cincinnati has played two other games and won. Next, for simplicity, assume none of the unnamed teams has played any other games.

The OOWP is calculated as (Syracuse's OWP + Syracuse's OWP + Cincinnati's OWP ) / 3.

Syracuse has played and beat the team in question (which, excluding the games against Syracuse, only lost to Cincinnati), lost to the team in question (excluding Syracuse, only lost to Cincinnati), and lost one other game (excluding Syracuse, this team has no WP). Syracuse's OWP is (0/1 + 0/1) / 2 = 0.0000.

Cincinnati has played the team in question (excluding Cincinnati, they went 1-1 vs. Syracuse) and won versus two other opponents each of which have no WP when games versus Cincinnati are excluded. Cincinnati's OWP is (1/2) / 1 = 0.5000.

For the team in question, the OOWP is thus (0.0000 + 0.0000 + 0.5000) / 3 = 0.1667

For the team in question, the RPI can now be calculated:

RPI = (WP * 0.25) + (OWP * 0.50) + (OOWP * 0.25)

Plugging in numbers from the above example gives you

RPI = (0.4117 * 0.25) + (0.3333 * 0.50) + (0.1667 * 0.25) = 0.3113

Rating Percentage Index - Wikipedia, the free encyclopedia

 

[CoogansBluff]

The formula itself is simple, but it's almost impossible to calculate it for over 100 teams without a program.

Formula -

RPI = (Team's winning PCT x 0.25) + (opponent's winning PCT x 0.50) x opponent's opponent's winning PCT x 0.25)

The weakness of an RPI method is that it does not consider margin of victory. Systems that do use margin of victory provide a better gauge of the teams' actual strength, IMO.

 

[riseball]

The data is the biggest obstacle. With the right dataset the calculations including a margin of victory would be easy. Would be interesting to play with. God I am a frickin nerd.

 

[SoAZDad]

Note that NCAA changed the weighting in baseball for a win from 1.0 to 0.7 at home, 1.0 to 1.3 away, and did not change the neutral game value (1.0). The 0.3 seems somewhat arbitrary but it might be close...

 

[riseball]

Note that NCAA changed the weighting in baseball for a win from 1.0 to 0.7 at home, 1.0 to 1.3 away, and did not change the neutral game value (1.0). The 0.3 seems somewhat arbitrary but it might be close...

I imagine that was derived in part from a sample of home, away, and neutral records. But like you suggested at some point it is just a SWAG.