[81alum]

[quote author=81alum link=topic=11611.msg111541#msg111541 date=1422015623]
UCLA is RPI 58
U$€ is RPI 63

Ugh.  In other words, losing to either of them would hurt a lot, but winning won't help all that much. 

While the latter statement may be completely true about people's subjective evaluation of the team, it's not really quite correct about the RPI.

The RPI is a simple formula that merely adds in your winning percentage (25%)* plus your opponents' winning percentage (50%)**, plus your opponents' opponents' winning percentage (25%).**

Thus winning or losing matters, period. And your opponents winning against everyone else helps. But there is no kind of power rating evaluation of individual games that makes wins over the better teams more helpful than wins over poorer teams.
Another way to put it is that an unexpected win plus an unexpected loss is a wash, subject to home/away issues.

It is repeatedly stated by commentators that losing to poor teams is particularly damaging. That is true only sort of. It is true that in a loss to a bad team both the first two parts of the formula take a hit. However, the second part takes the same hit regardless of the outcome of that game and the first isn't hurt any more than in any other loss.


* There is a home/away adjustment for a team's own winning percentage. A home win is worth 0.6 wins, a home loss 1.4 losses, an away win is worth 1.4 wins, and an away loss is worth 0.6 losses.
** Opponents' win/loss percentage isn't adjusted for home/away but is adjusted by taking out results against the team for whom you are calculating the RPI.
[/quote]

So if you calculated our RPI by dropping our win over Connecticut, would our RPI wind up the same as if you calculated it dropping our win over Santa Barbara?

[81alum]

[quote author=81alum link=topic=11611.msg111541#msg111541 date=1422015623]
UCLA is RPI 58
U$€ is RPI 63

Ugh.  In other words, losing to either of them would hurt a lot, but winning won't help all that much. 

While the latter statement may be completely true about people's subjective evaluation of the team, it's not really quite correct about the RPI.

The RPI is a simple formula that merely adds in your winning percentage (25%)* plus your opponents' winning percentage (50%)**, plus your opponents' opponents' winning percentage (25%).**

Thus winning or losing matters, period. And your opponents winning against everyone else helps. But there is no kind of power rating evaluation of individual games that makes wins over the better teams more helpful than wins over poorer teams.
Another way to put it is that an unexpected win plus an unexpected loss is a wash, subject to home/away issues.

It is repeatedly stated by commentators that losing to poor teams is particularly damaging. That is true only sort of. It is true that in a loss to a bad team both the first two parts of the formula take a hit. However, the second part takes the same hit regardless of the outcome of that game and the first isn't hurt any more than in any other loss.


* There is a home/away adjustment for a team's own winning percentage. A home win is worth 0.6 wins, a home loss 1.4 losses, an away win is worth 1.4 wins, and an away loss is worth 0.6 losses.
** Opponents' win/loss percentage isn't adjusted for home/away but is adjusted by taking out results against the team for whom you are calculating the RPI.
[/quote]

So if you calculated our RPI by dropping our win over Connecticut, would our RPI wind up the same as if you calculated it dropping our win over Santa Barbara?