Teams in the Sweet Sixteen: RPI vs Pomeroy

storminspank

storminspank

Justin VanLaere
Conference RPI Rank + Number of Sweet Sixteens in parenthesis:
#1 MWC (0)
#2 B1G (4)
#3 BE (3)
#4 ACC (2)
#5 B12 (1)
#6 P12 (2)
#7 A10 (1)
#8 SEC (1)
#9 MVC (1)
#26 ASun (1)


Conference Pomeroy Rank + Number of Sweet Sixteens in parenthesis:
#1 B1G (4)
#2 BE (3)
#3 ACC (2)
#4 P12 (2)
#5 MWC (0)
#6 B12 (1)
#7 SEC (1)
#8 A10 (1)
#9 MVC (1)
#24 ASun (1)




9 of the Top 15 RPI teams are in the Sweet Sixteen (per realtimerpi)
11 of the Top 15 Pomeroy Rated teams are in the Sweet Sixteen.
 
What is really skewing the RPI are the tricks the MWC use to inflate their RPI. The biggest one is the DII and JuCo schools they schedule for warm up games that do not show up on the RPI. If they were forced to substitute those games against teams from the low majors then the RPI would fall right in place.

But the Pomeroy ranking system is definitely better than the RPI, it not only take into account the SOS (which is the main component of the RPI) but it also factors in individual games and such. While the RPI is easy to manipulate Pomeroy is a little more difficult.

We can debate RPI all day long and some here really hate it (understandably so) but one thing we all have to remember is with the RPI we know exactly what it looks at so the coaches know exactly what they are doing when they schedule games against bad low majors. Fran himself even said he was counting on a couple of upsets in the conference schedule that did not happen. Going forward I am sure he learned his lesson and will use the RPI to his advantage and schedule non conference better. You can get away with scheduling a couple of warm up games against bad low majors, but not SEVEN of them.
 
It's a joint process. The AD makes the calls, but the coach has input. Barta isn't the kind of guy to overrule.
 
ssckelley said:
You can get away with scheduling a couple of warm up games against bad low majors, but not SEVEN of them.
Click to expand...

Nobody has answered this so I will ask again.

Is it that easy of a task to tell 10 months before the next season starts who is going to be in the 125-199 RPI range and who will be 200+?

If it is, someone should volunteer that info to GarBar so he can schedule games against fairly easy competition without killing the RPI number.
 
SS, I see your point. But allow me to play a scenario out here.

We keep the OOC schedule relatively easy (perhaps a slight bit tougher-but mostly comparable to this years), knowing we'll get some of those wins in the B1G that we just missed this year (MichSt, Wiscy, Minny, Indiana, etc). Not only that, but we don't lose to a Purdue or Nebby.

I don't see that being the direction, but our overall win total is better and we still manage some of the marquee wins from the the conference schedule. Do you still need to beef up the OOC?

Thoughts?
 
Section136 said:
Nobody has answered this so I will ask again.

Is it that easy of a task to tell 10 months before the next season starts who is going to be in the 125-199 RPI range and who will be 200+?

If it is, someone should volunteer that info to GarBar so he can schedule games against fairly easy competition without killing the RPI number.
Click to expand...

I would think if you had someone dedicated to this task it would not be that hard to figure out.

First I would look at how the team finished the previous year in their conference. Then I would see who they have returning. I guess you could look at other teams in the conference as well to see how those teams look to see how a team might fare. I mean you can use KenPom and other stats to figure out things about teams.

You make it sound hard but I would not think it would be very hard to predict how teams should do if you had someone dedicated to this task.
 
Do you think the NCAA screwed Oregon with Oregon's seeding in the basketball tournament partially caused by a starting point guard out many games?

Do we want the RPI to reflect the entire season or the last, say, 10 games?

How does one explain Florida Gulf Coast? Or Butler of previous years?
 
Last edited:
HomerChampless said:
Do you think the NCAA screwed Oregon with Oregon's seeding in the basketball tournament partially caused by a starting point guard out many games?

Do we want the RPI to reflect the entire season or the last, say, 10 games?

How does one explain Florida Gulf Coast? Or Butler of previous years?
Click to expand...

You ask several different questions, which aren't all related.

Was Oregon underseeded? Probably. That does happen.

Not sure what the last 10 games has to do with it. According to the last 10 games theory, Gonzaga should still be playing since they hadn't lost in forever before losing to Wichita State.

As to how one explains FGC or Butler, the answer is that is tournament basketball. Tools like Pomeroy or Sagarin ratings can give us a good idea of a team's relative value compared to another team. But they aren't going to be 100% predictive of how a team will perform against an opponent. That's why they play the games. In 2010 (Butler's 33-5 season), Pomeroy had Butler as the 12th-best team in the country. Looking at the number it shouldn't have been totally shocking that they advanced to the final game. Florida Gulf Coast is currently in the Top 100 of Pomeroy and RPI. So while it was certainly an upset for them to defeat Georgetown and San Diego State, it's not as if you have a team ranked 150 or worse getting to the Sweet 16. FCG was a decent team who played as well as it could for 2 straight games. Can they duplicate the performance again this week? That's the fun answer.

No ranking system is going to predict the outcome of a tournament 100%. Too much can happen in a single game But taken as a whole, it's fairly evident the Mountain West was significantly overvalued by RPI. Others have pointed out why the MWC was rated so high, with its member teams playing games against non-Division I teams rather than low-level Division I opponents.

Hopefully the RPI will lessen in importance, but I wouldn't be on that. It is a way for the mid-majors to rank more highly than they should and gives the Selection Committee the cover to include teams like Middle Tennessee and Boise State as at-large teams.
 
Section136 said:
Nobody has answered this so I will ask again.

Is it that easy of a task to tell 10 months before the next season starts who is going to be in the 125-199 RPI range and who will be 200+?

If it is, someone should volunteer that info to GarBar so he can schedule games against fairly easy competition without killing the RPI number.
Click to expand...

Remember the RPI rank of your opponents has nothing to do with your own RPI, it is all about their win percentage along with their opponents win percentage. I can look at the teams rosters and tell you who should at least be decent in their own conference, so a coach should be able to do the same. Plus some teams are historically bad year over year. It is not as hard as you think and other teams are already doing this.
 
ssckelley said:
Remember the RPI rank of your opponents has nothing to do with your own RPI, it is all about their win percentage along with their opponents win percentage. I can look at the teams rosters and tell you who should at least be decent in their own conference, so a coach should be able to do the same. Plus some teams are historically bad year over year. It is not as hard as you think and other teams are already doing this.
Click to expand...

Yep and this goes for some conferences too, so if you're really concerned about not playing the bottom of the barrel D-1 teams, you can just avoid those conferences all together.
 
hawkfan33 said:
Yep and this goes for some conferences too, so if you're really concerned about not playing the bottom of the barrel D-1 teams, you can just avoid those conferences all together.
Click to expand...

You can't because those low majors are the ones looking for pay checks and will play here without returning the favor. Just avoid playing the teams projected to finish at the bottom of those conferences. Gardner Webb and WKU does not hurt you but S Carolina St and Howard do.
 
Eda said:
SS, I see your point. But allow me to play a scenario out here.

We keep the OOC schedule relatively easy (perhaps a slight bit tougher-but mostly comparable to this years), knowing we'll get some of those wins in the B1G that we just missed this year (MichSt, Wiscy, Minny, Indiana, etc). Not only that, but we don't lose to a Purdue or Nebby.

I don't see that being the direction, but our overall win total is better and we still manage some of the marquee wins from the the conference schedule. Do you still need to beef up the OOC?

Thoughts?
Click to expand...

My point is that the RPI is an antiquated, tired, useless metric used to help the committee select teams for the NCAA Tourney.

But to answer you question, the schedule will get beefed up. How much? Probably not as much as some expect.
 
storminspank said:
My point is that the RPI is an antiquated, tired, useless metric used to help the committee select teams for the NCAA Tourney.

But to answer you question, the schedule will get beefed up. How much? Probably not as much as some expect.
Click to expand...

It really doesn't need to be overly tweaked. If that Virginia Tech game this year was Kansas (solid example team from the mid west) would the Hawks have made the tournament with how dumb RPI is?
 
Eda said:
SS, I see your point. But allow me to play a scenario out here.

We keep the OOC schedule relatively easy (perhaps a slight bit tougher-but mostly comparable to this years), knowing we'll get some of those wins in the B1G that we just missed this year (MichSt, Wiscy, Minny, Indiana, etc). Not only that, but we don't lose to a Purdue or Nebby.

I don't see that being the direction, but our overall win total is better and we still manage some of the marquee wins from the the conference schedule. Do you still need to beef up the OOC?

Thoughts?
Click to expand...

Sorry, I missed this. These are the games Iowa played that killed their RPI and ultimately kept them out of the NCAA tournament:

South Carolina State - (6-24)
Texas A&M CC - (6-23)
Howard - (7-24)
Coppin State - (8-24)
South Dakota - (10-20)
Central Michigan - (11-20)
@ Virginia Tech - (13-19)

Combined W/L - (61-154)
Combined W % - 28%

Your opponents win percentage counts as half of your RPI score and that 28% win percentage of those 7 teams killed Iowa's RPI. Iowa does not really have to play a much tougher schedule all they got to do is avoid the really bad teams of the low majors. Of those teams Iowa played 3 teams from the MEAC if instead of playing Howard, Coppin State, and S Carolina St they play Savannah State (19-15), Delaware St (15-18), and North Carolina A&T (19-17) the opponent win percentage would have jumped 13% between those 7 teams (41%). Iowa would have won the same amount of games, probably would have ended up with an RPI in the 50s, and in the NCAA tournament.

People can gripe about the RPI all they want but at least with the RPI you know exactly what you are dealing with and how to schedule around it. Iowa needs to start using it to their advantage.
 
Plus every rating system has inherent biases. I've said the before but Kenpom has Pitt and wisky in the top 10. Anybody want to legitimately argue they deserved a 2 or 3 seed?
 
BryceC said:
Plus every rating system has inherent biases. I've said the before but Kenpom has Pitt and wisky in the top 10. Anybody want to legitimately argue they deserved a 2 or 3 seed?
Click to expand...

Because they rape teams and fouls actually get called when tourney time rolls around.
 
Clearly, the MWC gamed the RPI system very effectively.

Now, I think they had some good teams,and New Mexico was one of them that deserved to make the tourny,along with San Diego St. and Col. St. Boise and UNLV should have been relegated to to the NIT.

I think you could make a case for Virginia to get one of those slots...and maybe Iowa?...or Southern Miss.
 
You must log in or register to reply here.

Latest posts

Share this page

Top