Why the RPI sucks?

[MyTurn2Raise]

Alright, some of you have read me rail on the RPI forever. The question is why.

Fortunately, I can now present a simple answer to that question.
The Colton Index founder was posting his stuff at insideillini today. Turns out he has a video/slideshow showing the flaws of RPI very well. Colton is definitely an improvement, but still falls short of KenPom IMO. If you want, read up on it.

[Green Kool Aid]

MT2R,

A general question about the tournament...

Do you think a 5 point loss to a team your team was a 20 point underdog underdog to help or hurt this team's case for a tournament spot??

If this is too vague, I can expand. How much does the tournament committee factor in luck versus "just getting the job done?"

[CardSharpCook]

wow, what a miserable presentation. "the RPI 84th best team is ACTUALLY the 228th best team." .... um, ok. Is it? any proof whatsoever? So the team you arbitrarily called the 228th best team is the RPI 84th best. Ok. that's wonderful.

[CardSharpCook]

its funny because the guy has absolutely no clue.

[MyTurn2Raise]

[ QUOTE ]
its funny because the guy has absolutely no clue.

[/ QUOTE ]


this is the most ironic post I've seen in a long time

[MyTurn2Raise]

[ QUOTE ]
MT2R,

A general question about the tournament...

Do you think a 5 point loss to a team your team was a 20 point underdog underdog to help or hurt this team's case for a tournament spot??

If this is too vague, I can expand. How much does the tournament committee factor in luck versus "just getting the job done?"

[/ QUOTE ]


ahhhhh.... the good loss debate
well, it doesn't affect RPI at all. RPI doesn't include margin of victory or defeat.
As far as the rest of the criteria. It depends if the committee makes a point of it during discussion. If it happened to be a game a committee member watched, they will likely view it as a positive. Else, I imagine it isn't brought up at all.

[Austiger]

I can't watch the video at work, but the RPI is about as good as any other rating system. (I'm assuming you agree that they need a rating system.) They all have flaws. I don't follow KenPom, but at first glance I see that LSU is ahead of Vandy despite losing 5 more games in the weaker division of the SEC. The RPI has Vandy 37 and LSU 91 which seems about right. I know we could go back and forth as to which is better, but the point is you're never going to have one that's perfect and since the NCAA doesn't want to use MOV, which I can respect, the RPI is about as good as the next.

I'm assuming KenPom uses MOV so it's probably more accurate overall in terms of valuing teams for sports betting. But I think it's a bad idea to use MOV in the selection process because it encourages teams to keep their starters in for the whole game, embarrass the opponent by running the score up, etc.

[iggymcfly]

What I don't like about KenPom is that it's too dependent on margin of victory to the point that it hardly pays attention to wins and losses. If you're using it as a predictive tool, fine, but when you're selecting a bracket, who actually won and lost the games should be a pretty important factor. I like the Sagarin ratings ("rating" portion) as a good middle ground.

[NajdorfDefense]

It doesn't suck, it's one tool to use:
RPI
SOS
W-L
Margin of Victory,
all of the above for Non-Conf,
road v home record,
record v top 25, top 100, etc.

[MyTurn2Raise]

[ QUOTE ]
What I don't like about KenPom is that it's too dependent on margin of victory to the point that it hardly pays attention to wins and losses. If you're using it as a predictive tool, fine, but when you're selecting a bracket, who actually won and lost the games should be a pretty important factor. I like the Sagarin ratings ("rating" portion) as a good middle ground.

[/ QUOTE ]

not when the committee expressly stated that it wants the 34 'best' at-large teams, not who put together the more deserving resume