Men's Teams
Women's Teams
Conferences - Men Conferences - Women Full URPI Rankings
Men | Women

URPI #'s Updated
3/12/23 - 3:49 PM MDT

My Other Site:

Dedicated to the Music of
NFL Primetime

The Unbiased RPI, or URPI, is a system for mathematically ranking Division I college basketball teams. As the
name suggests, the URPI uses the same basic concepts as the Ratings Percentage Index (RPI) used officially by
the NCAA for ranking teams & determining NCAA Tournament seeds. Like the RPI, the URPI is fundamentally
based on winning percentages. The main differences between the two, however, are that the URPI changes
the category weights & adds a small margin-of-victory component so as to remove the official RPI's inherent
strength of schedule—and, therefore, major conference—bias.

Here's the basic formula:

(1/3)*ModWIN% + (4/9)*SOS + (2/9)*OppSOS + (1/1000)*MOV = URPI

ModWIN% = Modified Winning % - As with the RPI, the URPI adds weights to home & road games—road wins &
home losses are worth more, while home wins & road losses are worth less. However, the official RPI does
something really bizarre. It uses the weighted home/road results for calculating a team's own winning
percentage for their own RPI rating, but it doesn't use them when calculating winning percentages for
strength-of-schedule purposes. It seems strange that the NCAA would deliberately make things 10
times more complicated than they need to be, but that's their prerogative. The URPI uses the same weights
across the board, but the weights are smaller than the RPI's. For the RPI, road wins & home losses are worth
1.4, while home wins & road losses are worth 0.6. That seems a bit counterintuitive, since it makes road wins more than twice as valuable as home wins. The URPI instead uses a 1.2 vs. 0.8 split (neutral wins/losses are each worth 1). Another flaw in the RPI is how it handles non-DI games. Non-DI wins are totally ignored, which is fine (the URPI does the same), but the official RPI also ignores non-DI losses for some reason. The URPI counts a non-DI loss as 2 losses—regardless of location—in its Modified Winning Percentage formula. Here's the equation:

(0.8*HW + 1.2*RW + NW) / (0.8*HW + 1.2*RW + NW + 1.2*HL + 0.8*RL + NL + 2*XL) = ModWIN%

XL = non-DI loss. All other wins & losses are DI only.

SOS* = Strength of Schedule = Opponents' Winning % - The Strength of Schedule component is a simple
average of each opponent's Modified Winning Percentage (as mentioned above, the URPI uses the home/road
winning percentage weights for SOS unlike the official RPI). The percentages are added up & divided by the
number of DI games played (for the divisor, each game counts as 1). Non-DI games are totally ignored for
Strength of Schedule (even losses). If the team in question plays another team more than once, that
opponent's ModWIN% is obviously added in for each game played. This is the SOS formula:

(ModWIN%(DI Opp#1) + ModWIN%(DI Opp#2) + …) / (# of DI Games) = SOS

OppSOS = Opponents' Strength of Schedule = Opponents' Opponents' Winning % - Opponents' Strength of
Schedule uses the same concept as SOS, just with a different stat. Instead of averaging the winning
percentages of a team's DI opponents, OppSOS averages the strength of schedule numbers of that team's DI
opponents. Thinking in terms of spreadsheets, it does the same thing but to the next column over. Here's
what that looks like:

(SOS(DI Opp#1) + SOS(DI Opp#2) + …) / (# of DI Games) = OppSOS

MOV = Average DI Margin of Victory - This is the simple stat. It's just a team's average scoring margin
(points scored - points allowed) for each DI game on the schedule. Equation:

(Points Scored (DI only) - Points Allowed (DI only)) / (# of DI Games) = MOV

* - The "SoS" number/rank listed on the team & conference pages is not simply the Opponents' Winning % number, but is in fact the weighted average of SOS & OppSOS:

"SoS" = (2/3)*SOS + (1/3)*OppSOS