[quote=ā9erken, post:20, topic:30797ā][quote=āupperdeck, post:13, topic:30797ā]Have always hated Rpi. If I lose by 100 to a good team, why should my ranking get better just because I played them? Shouldnāt that be a bad lose either way.
It is just another tool to minimize loses against strong opponents. (See ACC top 25 teams beating themselves). They had to have a way to make Good teams that lose to good teams to not be penalized as much. The RPI accomplishes that. For the bottom 3/4 of the RPI rankings, you can just throw out any real comparison.[/quote] RPI probably favors the power conferences the least out of all the ratings because it has a much greater home court correction than the other ratings do. This is especially true in the OOC, where most power conference schools play very few true road games. Itās not perfect, but I think the other ratings underrate the benefits of home court advantage. Itās difficult to accurately reflect how differently teams have to play when they are on the road (softer D to avoid fouls, close calls tend to go the other way, etc.), particularly when a lesser-known team is playing the well-known home team, and the better non-power teams rarely get very good teams to come play them on their own home court. It ends up being a little bit of a self-fulfilling prediction (the power conference team gets stronger numbers OOC because it plays at home more often with an undercorrection for homecourt advantage, itās more likely to win itās home games due to that advantage, and so the other power conference teams that beat them in conference play look stronger than they should). Itās no surprise that the selection committees have been using the rpi-SOS to evaluate teams but not individual rpi, because individual rpi takes home court into account and the SOS does not.
Thereās obviously limitations to it and some teams probably end up with misleading ratings due to quirks in how the schedule played out. The more sophisticated ratings could be much better than rpi IMO if they just had a slightly larger correction for home court.[/quote]
Actually, the RPI benefits Power Conference teams most of all because the home-court advantage is very limited.
Ratings like KenPom and Sagarin really only rate you based on your performance based on expectations, and that is very much effected by where games are played. The reason Power Conference benefit so much in the RPI is from home games.
This is due to snowballing of SOS in-conference because game weights for site location disappear in the SOS of teams.
Letās say Georgetown goes 10-5 (.667) in non-conference (7-1 at home, 2-2 Neutral, 1-2 on road), their adjusted win loss would be 7.6-4.6 (.622), which is marginally worse than true W-L. Now, when Villanova gets to play Georgetown, and Georgetownās record is added to Villanovaās strength of schedule, they get to add 10-5, NOT 7.6-4.6. Over 18 games that small boost really adds up and Villanovaās RPI is better for it.