Bloated RPIs

[Jamar Howard 4 President]

Ut oh. There goes the JH4P bat signal again.

The insanity here is too much for me to take. I'm going to sit back, let it go, eat some ice cream, and watch Creighton lose.

[ShockerFever]

The insanity here is too much for me to take. I'm going to sit back, let it go, eat some ice cream, and watch Creighton lose.

You must like insanity then.

Any chance you could sit back and let it go for the rest of the year?

[Jamar Howard 4 President]

JH4P posts and mentions Creighton. Fever responds 1 min later. Not surprised. :-)

[ShockerFever]

JH4P posts and mentions Creighton. Fever responds 1 min later. Not surprised. :-)

Fever knifes JH4P in half. JH4P responds 1 min later. Not surprised.

This is fun.

[ABC]

Seems like a bit of good RPI within BCS conferences are self-perpetuating. Kind of like Ivy League schools hiring other Ivy League graduates as professors.

Now, Georgia Tech with a RPI of 197 is a counter to that.

[im4wsu]

What annoys me about RPI is that a win against the #300 team is valued the same as a win against the #1 team. No weighting for whether the game was an expected win or loss. The "spread" is technically not designed to predict outcomes of games, but it does give some insight as to how teams do compared to expectations. What id doesn't reflect is the difference between a #200 team play all games against 150-250 teams ascompared to a #10 team playing games against #1-100 teams. Anyway, using the top 15 in the current NCAA RPI and the records from StatSheet.com as of 10 pm, here is a summary of the ATS records:

	RPI Rank	Record	Record ATS	Avg Spread	ATS Percent
Wichita St.	4	34-0	25-6	-11.9	80.6%
Villanova	5	28-4	21-8-1	-8.2	71.7%
Florida	2	31-2	18-9-2	-8.5	65.5%
Virginia	12	27-6	18-10-1	-6	63.8%
Arizona	1	30-4	19-12-1	-11.8	60.9%
San Diego St.	13	29-4	18-12	-7.2	60.0%
VCU	14	26-7	18-12	-8.1	60.0%
Creighton	8	26-7	18-13	-9.5	58.1%
Michigan	9	25-7	16-12-1	-5.7	56.9%
Duke	7	26-7	18-14-1	-11.1	56.1%
Syracuse	11	27-5	13-11-4	-8.1	53.6%
Kansas	3	24-9	16-16-1	-8.7	50.0%
Wisconsin	6	26-7	16-16-1	-7.7	50.0%
Iowa St.	10	26-7	15-15-1	-5.7	50.0%
Cincinnati	15	27-6	13-14	-5.6	48.1%

[Shox151]

i'm curious has anyone ever actually taken the time to calculate out what the rpi on this multiplicative idea? what do the numbers look like? how different from the current standings are they?

[Flip1381]

Sorry ciaomichael, misunderstood.
Shox151, that would take quite awhile to calculate, I think.

[Flip1381]

No, you wouldn't be penalized 3 times or any times for losing - but you wouldn't get any credit either:

If you win, you pick up: (2/3) x Opponents Win %
Plus (1/3) x Opponents opponents Win%

If you lose, you pick up: Zero

So, you have to win a game to get any positive credit whatsoever.

However, you get much more credit by beating a team with a good winning percentage, and even more credit by beating a team whose opponents have a good winning percentage. There's still plenty of weight in there for strength of schedule, but you must win to realize it.

So if you go by the calculation you have above, if a team loses all their games but 1 to a really good team that has only lost 1 game, their ranking would be really high, right?