Carolina RPI/SOS

[houstondukie]

Can somebody help me understand how UNC has an RPI of 2 and SOS of 7 when their wins include:

RPI top 25: BYU, Clemson (2x)
RPI top 50: Miami, Kent St., Davidson

That's only 6 wins against 5 teams in the RPI top 50.

They are 1-1 against the ESPN/AP top 25 (Duke is 4-1)

I just don't get the RPI/SOS.

[Classof06]

Can somebody help me understand how UNC has an RPI of 2 and SOS of 7 when their wins include:

RPI top 25: BYU, Clemson (2x)
RPI top 50: Miami, Kent St., Davidson

That's only 6 wins against 5 teams in the RPI top 50.

They are 1-1 against the ESPN/AP top 25 (Duke is 4-1)

I just don't get the RPI/SOS.

Your guess is as good as mine. I too don't understand how Carolina has a higher SOS than Duke. They played way more OOC road games but Duke definitely played better teams. That is one of the things I don't understand about Carolina. I also don't understand why many people have penciled them in to win Saturday and I can't understand how people have already written them off as a #1 seed. It boggles my mind. The good part is that we get the chance to show those people how stupid they are come Saturday.

Joe Lunardi said the only reason he doesn't currently have Duke as a 1 seed is because of our "shaky losses," despite the fact that all 3 of our losses were on road/neutral courts and both of Carolina's losses were at home (one of them was to us). Lawson or not, that just doesn't make sense.

[SharkD]

It's because Ty Lawson is really, really ridiculously good looking. (But can't turn left.)

http://graysmatter.codivation.com/content/binary/zoolander_face.jpg

Or, that definitive win over Coppin State.

[coachbob]

Their schedule is harder because they had to play Duke while Duke got off easy-they played unc.

[Chitowndevil]

The problem with the RPI is, depending on how you calculate it, it weights hugely on opponents' winning percentage and even opponents' opponents' winning percentage. Duke's RPI SoS is hurt tremendously by NC Central, Princeton, Michigan, Albany, and Eastern Kentucky: the first two have a combined 6-43 record, and the last two have a combined 9 losses to teams with an RPI of 250 or worse and NO wins versus the top 150.

UNC had their share of cupcakes, sure, but Rutgers for example plays in the Big East and has wins over Villanova and Pitt, and even Nicholls St has a win over Stephen F Austin. This is part of why RPI SoS can be so strange--it can penalize you disproportionately for playing really bad teams.

And by the way, according to two ratings I like much better, Duke's schedule is actually stronger: Duke's schedule is ranked 21st in both the Pomeroy and Sagarin ratings, while UNC is 47th in the Pomeroy and 34th in the Sagarin.

[keithg]

RPI and RPI SOS are both calculated the same way:

It's 2/3 opponents' winning percentage and 1/3 opponents' opponents' winning percentage.

So the multiplication of these two ratios when associated with North Carolina's schedule is higher than the ratio when the same formula is applied to Duke's schedule.

Pretty simple (and somewhat flawed) really..but no great Conspiracy against Duke.

[Clipsfan]

RPI and RPI SOS are both calculated the same way:

It's 2/3 opponents' winning percentage and 1/3 opponents' opponents' winning percentage.

So the multiplication of these two ratios when associated with North Carolina's schedule is higher than the ratio when the same formula is applied to Duke's schedule.

Pretty simple (and somewhat flawed) really..but no great Conspiracy against Duke.

What's interesting about this discussion is that K has often been accused of manipulating the schedule so as to have a high RPI. What's happened is that many teams do this now. As was mentioned above, the key is to avoid the really bad teams and hope that you don't tire your team out by playing teams in the 150-200 range rather than the 250-300 range. Good teams rarely lose (unless they're UK) to the latter, but it can happen against the former, resulting in fewer chances to rest.

[yancem]

Your guess is as good as mine. I too don't understand how Carolina has a higher SOS than Duke. They played way more OOC road games but Duke definitely played better teams. That is one of the things I don't understand about Carolina. I also don't understand why many people have penciled them in to win Saturday and I can't understand how people have already written them off as a #1 seed. It boggles my mind. The good part is that we get the chance to show those people how stupid they are come Saturday.

Joe Lunardi said the only reason he doesn't currently have Duke as a 1 seed is because of our "shaky losses," despite the fact that all 3 of our losses were on road/neutral courts and both of Carolina's losses were at home (one of them was to us). Lawson or not, that just doesn't make sense.

I think that part of the reasons that UNC is getting more love than Duke is due to pre-season expectations. Its kinda like football, if you don't start out the season as a top tier team, its really hard to move up the ranking but so far. Conversely, if your a pre-season top 5, as long as you don't completely blow it you'll only drop so much. UNC was a top 3-5 pick pre-season and although they have had some scares, they've only lost 2 games (one of them without their second best player) Duke on the other hand was expected to have a solid year but not the kind of year we're having so some people don't know if we're fools gold or the real thing. I guess we will have to prove it on Saturday.