AAC RPI

[tzznandrew]

Just looking at the conference, we don't have a lot at the top (though I think Cincy will rise) but the middle isn't that bad. If ECU and USF could get their act together (and UConn and Memphis too) this conference would be in good shape.

This is the ESPN RPI as of January 16.

RPI

16 Wichita State
36 Cincy
53 Houston
54 Temple
57 UCF
80 UConn
85 SMU
101 Memphis
112 Tulsa
123 Tulane
278 USF
293 ECU

KenPom is less than thrilling, though the top is more reasonable:

8 Cincy
11 WSU
38 SMU
39 Houston
78 UCF
94 Temple
128 UConn
138 Tulsa
139 Tulane
165 Memphis
313 USF
332 ECU

 

[Austin316]

Normally I don’t check it until the last ten games... that being said this year is a little different.

Can’t lose to Memphis.

This weekend will be HUUUUEEEEGE!

You can bet ur as5 Austin316 will be in attendance vs NOVA.

 

[Spokesbender]

My biggest takeaway is the huge difference for SMU's ranking with RPI vs KenPom

 

[businesslawyer]

Recognize that they are not even trying to accomplish the same thing. Ken Pom tries to say how good you are. It's an important betters tool. And that makes margins of victory and loss critical.

RPI, while incredibly imperfect, tries to measure what you've accomplished in a season based on wins, losses and strength of opponents and opponents opponents. It doesn't care about margins. No one is saying that, if you're making a future prediction, you don't care about margins. But what the reliance on these numbers is meant to do is say that the Committee doesn't care about margin of victory. A win is a win and a loss is a loss -- we just care how many and against what competition.

 

[Edward Sargent]

Just looking at the conference, we don't have a lot at the top (though I think Cincy will rise) but the middle isn't that bad. If ECU and USF could get their act together (and UConn and Memphis too) this conference would be in good shape.

This is the ESPN RPI as of January 16.

RPI

16 Wichita State
36 Cincy
53 Houston
54 Temple
57 UCF
80 UConn
85 SMU
101 Memphis
112 Tulsa
123 Tulane
278 USF
293 ECU

KenPom is less than thrilling, though the top is more reasonable:

8 Cincy
11 WSU
38 SMU
39 Houston
78 UCF
94 Temple
128 UConn
138 Tulsa
139 Tulane
165 Memphis
313 USF
332 ECU

Still looking at 2 tourney teams in all of that

 

[whaler11]

Recognize that they are not even trying to accomplish the same thing. Ken Pom tries to say how good you are. It's an important betters tool. And that makes margins of victory and loss critical.

RPI, while incredibly imperfect, tries to measure what you've accomplished in a season based on wins, losses and strength of opponents and opponents opponents. It doesn't care about margins. No one is saying that, if you're making a future prediction, you don't care about margins. But what the reliance on these numbers is meant to do is say that the Committee doesn't care about margin of victory. A win is a win and a loss is a loss -- we just care how many and against what competition.

They seem to be putting less value on RPI these days than in the past.

Ya gotta drag some of these folks into the 21st century.

Takes time.

 

[businesslawyer]

They seem to be putting less value on RPI these days than in the past.

Ya gotta drag some of these folks into the 21st century.

Takes time.

RPI is sort of silly theoretically compared to the more computer driven analytics, but I fully support (not just this year when it favors us) determining the field versus wins and loses and against whom, and not looking at margins and trying to determine who is better. If 2 teams play the exact same schedule, and Team A wins by 3, 2 and 5, and Team B wins by 25 and 18 and loses by 1, if you go to Vegas and get even money you might want to bet on Team B if A and B play on a neutral court, but it is Team A that deserves to be in the tourney ahead of B.

 

[whaler11]

RPI is sort of silly theoretically compared to the more computer driven analytics, but I fully support (not just this year when it favors us) determining the field versus wins and loses and against whom, and not looking at margins and trying to determine who is better. If 2 teams play the exact same schedule, and Team A wins by 3, 2 and 5, and Team B wins by 25 and 18 and loses by 1, if you go to Vegas and get even money you might want to bet on Team B if A and B play on a neutral court, but it is Team A that deserves to be in the tourney ahead of B.

A metric that doesn’t concern itself with margin of victory is fine.

RPI is a bit blunt for even that, and it’s pretty easy to game. Still, I don’t mind it being one of a bunch of systems used as long it’s not the dominant one like it used to be.

FWIW kenpom isn’t using margin of victory as any kind of input. Being efficient can manifest itself in the scores but it’s not a raw input.

At 68 teams it’s not like anyone who is actually deserving gets left out anymore. It’s just who gets screwed seeding wise.

 

[businesslawyer]

A metric that doesn’t concern itself with margin of victory is fine.

RPI is a bit blunt for even that, and it’s pretty easy to game. Still, I don’t mind it being one of a bunch of systems used as long it’s not the dominant one like it used to be.

FWIW kenpom isn’t using margin of victory as any kind of input. Being efficient can manifest itself in the scores but it’s not a raw input.

At 68 teams it’s not like anyone who is actually deserving gets left out anymore. It’s just who gets screwed seeding wise.

Again, I'm not defending RPI itself. But when you measure a team possession by possession, as opposed to be whether you won or lost, you are measuring margin -- you're just doing it in a different manner.

 

[SubbaBub]

It's still about Top 100, Top 50, T0p 25 wins and Sub 150 losses. You need so many of the first three without many of the last one. The rest is just arranging deck chairs. The old BE was a Top 100 opportunity machine. Even the BET opening round games frequently met that bar. Almost every win in conference was added to the resume and very few losses (losses that actually occured) came close to being bad losses in the eyes of the committee.

In the AAC, these opportunities are front loaded into the OOC schedule. The irony is that this is one of the better years for the AAC in terms of RPI opportunities.