Updated May 2025
As discussed on the “RPI: Formula” page, Element 2 is part of the “Strength of Schedule” portion of the RPI. Specifically, Element 2 is the average of Team A’s opponents’ winning percentages against teams other than Team A. Element 3, the other “Strength of Schedule” contributor, is the average of Team A’s opponents’ opponents’ winning percentages (computed using the Element 2 winning percentages). In the overall RPI, the combined Element 2/Element 3 “Strength of Schedule” has an effective weight of approximately 50%. Of this 50%, Element 2 accounts for about 40% and Element 3 for about 10%. Thus Element 2 is a key part of the RPI and the main contributor to Team A’s “Strength of Schedule.”
A possible change to the RPI involves the computation of Element 2. The change would modify Element 2 so that it becomes the average of Team A’s opponents’ winning percentages including within each opponent’s winning percentage the results of its game (or games) against Team A. This would be a change from Element 2 currently being the average of Team A’s opponents’ winning percentages against teams other than Team A.
To show how the current Element 2 qualification represented by “against teams other than Team A” works, look at Boston College’s contribution to its opponents’ Element 2s during the 2010 season:
Boston College’s 2010 regular season winning percentage, including its games in the ACC Tournament, is 0.6750 based on the winning percentage formula in effect in 2010. The formula for determining this amount is (W + 1/2T)/(W + L + T), where W is wins, L is losses, and T is ties. When ranking Boston College based solely on its winning percentage, this places it tied with the teams ranked #47 through 52 in terms of the teams' winning percentages. (Note: The NCAA changed the winning percentage formula in 2024 to count ties as 1/3 of a win rather than 1/2.)
Boston College tied Stanford. That being the case, in computing Boston College’s contribution to Stanford’s winning percentage, the “against teams other than Team A” qualification requires deleting this tie from Boston College’s winning percentage. In formula terms, this changes the formula for computing Boston College’s Element 2 contribution for Stanford to (W + ½(T – 1))/(W + L + (T-1)). As a result, Boston College’s contribution to Stanford’s Element 2 is 0.6842. In rankings based solely on winning percentages, this treats Boston College as ranked #46 for Stanford’s strength of schedule purposes.
Boston College won against Hofstra. This changes the formula for computing Boston College’s Element 2 contribution to Hofstra to ((W-1) + 1/2T)/((W-1) + L + T). As a result, Boston College’s contribution to Hofstra’s Element 2 is 0.6579, thus treating Boston College as ranked #59-60 for Hofstra’s strength of schedule purposes.
Boston College lost to Virginia Tech. This changes the formula to (W + 1/2T)/(W + (L-1) + T). As a result, Boston College’s contribution to Virginia Tech’s Element 2 is 0.7105, thus treating Boston College as ranked #36-38 for Virginia Tech’s strength of schedule purposes.
Boston College won once and lost once against Virginia. This changes the formula to ((W-1) + 1/2T)/((W-1) + (L-1) + T). As a result, Boston College’s contribution to Virginia’s Element 2 is 0.6944, thus treating Boston College as ranked #44 for Virginia’s strength of schedule purposes.
Boston College lost twice to Maryland. This changes the formula to (W + 1/2T)/(W + (L-2) + T). As a result, Boston College’s contribution to Maryland’s Element 2 is 0.7500, thus treating Boston College as ranked #30 for Maryland’s strength of schedule purposes.
So, because of the “against teams other than Team A” qualification, Maryland is treated as though it played the #30 team in terms of winning percentage, Virginia Tech the #36 to 38 team, Virginia the #44 team, Stanford the #46 team, and Hofstra the #59 to 60 team. Yet, they all played the same opponent, Boston College, that really is tied for #47 to 52 in the winning percentage list.
I never have seen an NCAA explanation supporting the “against teams other than Team A” qualification. On the other hand, I have seen suggestions that the reason for the qualification is that, without the qualification, Team B’s strength of schedule will be punished by Team B’s beating Team A (i.e., Team A will have a poorer winning percentage); and, conversely, Team B’s strength of schedule will be rewarded by Team B’s losing to Team A (i.e., Team A will have a better winning percentage). The argument, apparently, is that it isn’t “right” to have Team B’s strength of schedule “hurt” when it wins and “helped” when it loses. However, if the "against other teams" qualification is eliminated, then all of Team A’s opponents will be treated as having played the same opponent for strength of schedule purposes regardless of the game result. And, of course, they in fact all will have played the same opponent.
As the Boston College example shows, there is a good argument that the NCAA should delete the "against teams other than Team A" provision. Notwithstanding that, there is no evidence I have seen of the NCAA considering deleting the provision.