Real Time RPI

[VCU Finance 2008]

but I could see us losing @W&M, @ODU, @Hofstra, and vsGMU in the tourney final (just for example)

Terrible example. Completely unrealistic. This underlined part NEVER happens.

Hey, they kept it close last year. For the first 90 seconds or so.

 

[fmrick]

no team that i know of has been left out of the field with 26 wins.

There have been teams that have been left out with 26 wins. In fact, we need to look no further than Creighton last year with an RPI of 40 and 26 wins. That is the most comparable situation to where we are. However, last year's field was much stronger than this year's is shaping up to be and Creighton was blown away by Illinois St. in the MVC tournament. I like that the Pac-10 is down this year and the Big10 is a little softer than I expected. The best thing for us, IMO, is that we have a couple of bad (read: 2-3 bid conferences) BCS conferences and one or two 'super conferences' that have a bunch of at-large worthy teams but who beat each other up so much that, perhaps, an at large team is 2 games below .500 in conference. This would free up a bid, or, at the worst, have fewer 'locks' and a bunch of teams around where we are in the RPI with many fewer wins and more exposed flaws.

Other teams that have had 26 wins and not made it (according to Gary Parrish):
IUPUI and Robert Morris in 2008
Akron in 2007

However, the Creighton example is most analogous to where we could be. They had wins over New Mexico, Dayton, St. Joe's, and DePaul in their out of conference slate in addition to one bad loss at Arkansas Little Rock and at Nebraska.

While Creighton might seem to be a good example, they are much different then us at this point. At the end of their OOC play last year they were 10-2, RPI 61, SOS 230. Ours will be much better. Today we have moved to RPI 11, SOS 27, after WMU beat College of Charleston.

We will be a good at-large candidate this year, assuming we keep winning. My personal opinion is that we could lose two or three more in conference and be a lock. As long as those losses are on the road, against the top teams in the conference, and we get and win a good Bracket Buster game (we should), and beat ECU. My guess is that would put us in the 30's RPI, with a few solid OOC wins (OU, ur, URI, Nevada, and yes WMU, who is now 6-5 with an RPI of 69). Our SOS will help us greatly.

Just Win.

 

[BaNgMyPrOgRaM]

WMU was a loss. I guess you're saying that it was a 'quality' road loss.

-------------------
GO RAMS!
Beat Pirates!!

 

[xjohnx]

VCU up to #9 today.

Also CAA is up to #9th rated conference, just outside of #8.

 

[mrA10]

VCU up to #9 today.

Also CAA is up to #9th rated conference, just outside of #8.

From 12 to 9 because WMU beat Charleston last night. Imagine where would be if we beat WMU - we will never know

 

[mrA10]

Pomeroy uses a very different ranking system:

ODU - 8
VCU - 69
W&M - 77

http://kenpom.com/rate.php

I think realtimerpi's rankings are better

 

[fmrick]

Pomeroy uses a very different ranking system:

ODU - 8
VCU - 69
W&M - 77

http://kenpom.com/rate.php

I think realtimerpi's rankings are better

True, but the NCAA uses RPI. So it is THE one that matters.

Pomeroy is used to set betting lines.

 

[districtballer]

What's really great about this is how well the league is doing. That will help the league get an at-large if we keep it up.

This is just about the time when the numbers start to firm up a bit (10 games). It really doesn't fluctuate as much from here on out. We have put ourselves in pretty good shape going into CAA play as a team as well as a conference. If anyone told me we would be doing this well after Eric and Coach Grant left, I wouldn't believe it. We are ahead of schedule and just got two frontcourt pieces in Saintil and Skeen back.

If this is a rebuild year for us, we are doing great things as a program.

 

[Pavarotti]

 

[Rambunctious]

What's really great about this is how well the league is doing. That will help the league get an at-large if we keep it up.

This is just about the time when the numbers start to firm up a bit (10 games). It really doesn't fluctuate as much from here on out. We have put ourselves in pretty good shape going into CAA play as a team as well as a conference. If anyone told me we would be doing this well after Eric and Coach Grant left, I wouldn't believe it. We are ahead of schedule and just got two frontcourt pieces in Saintil and Skeen back.

If this is a rebuild year for us, we are doing great things as a program.

I think I see your point about this firming up, but I guess I see another side to it as well. On one hand we have played fully 1/3rd of the games for the year so statistically that would seem to be enough to constitute the numbers being firm. However, most of the teams remaining on our schedule (minus W&M and ODU and a possible BB opponent) are no where near the quality of RPI opponent that we've seen early in the season. Think Delaware twice, UNCW twice, GSU twice, etc. Meanwhile a lot of the BCS teams that are lower RPI right now because they've played OOC cupcakes will enhance their RPI as they play in-conference every week. The net of it all is even if we go 15-3 in conference we will probably drop in RPI some. I would think we could end up in the high 20s to mid-30s though - - and that is an outstanding place to be with a good resume at the end of the season.

 

[fmrick]

But the ranking is not static or in a vacuum. We don't just move because of what we do, we move because of what others around us do. The winning percentage of some of those teams in the BCS conferences will change because it is hard to win all the time in the big conferences. That is our advantage. We will take a hit when we get into the meat of conference play.

However, our conference RPI SOS is very strong. We are about 7-9 in overall SOS. The teams that make up that score are the ones we play for the rest of the year. Because the OOC play is pretty much over, the ranking of that score should not change much overall. This is where the numbers get complicated. See, in any conference the overall IN conference winning percentage is .5000. It is that way for EVERY conference. Each game has exactly one winner and one loser. So, once all the OOC play is finished, then every conferences will regress toward .5000. As long as you start above .5000, you can't go any lower than .5000 as a conference. So we should stay pretty much ranked in the 6-10 range RPI for the rest of the season as a conference.

BB is the only big deal changer for the conference at this point.

 

[VCU4LIFE]

WMU was a loss. I guess you're saying that it was a 'quality' road loss.

-------------------
GO RAMS!
Beat Pirates!!

No, he's saying that every team we play conitnues to impact our RPI regardless of whether we beat them or not.

 

[fmrick]

That is a great point that most often forget. We get their impact to our SOS if we beat them, or if we lose to them, either way it is the same. The impact from the win or loss is to the portion of RPI that comes from our winning percentage, which is only 25% of the total score. Our opponents winning percentage is 50% of the score. Their opponents winning percentage is the last 25%.

So for our SOS to be kicking at 27th at this point both our opponents and our opponents opponents must be doing pretty well.

 

[artRAMinMN]

That is a great point that most often forget. We get their impact to our SOS if we beat them, or if we lose to them, either way it is the same. The impact from the win or loss is to the portion of RPI that comes from our winning percentage, which is only 25% of the total score. Our opponents winning percentage is 50% of the score. Their opponents winning percentage is the last 25%.

So for our SOS to be kicking at 27th at this point both our opponents and our opponents opponents must be doing pretty well.

I'm glad to see this being explained well.

As far as the RPI goes, our wins over Richmond, Oklahoma, Rhode Island, or Nevada, each count less than each of our wins over Hampton and Tulane. Why would this be, when the first group of teams seems more impressive than the second? Because as far as its impact on your winning percentage, it doesn't matter who you beat, it matters where you beat them. Home wins get a lower value than road wins, regardless of how good or bad the teams beaten were. The quality of opponent does come into play when their winning percentage is factored in, but it's wholly separate... if we play a team with a good/high winning percentage, it helps us whether we win or lose... if we win, that helps us, but independently from any perceived quality of the win (aside from whether it was home, away, or neutral).

This failure to address quality wins/bad losses is the RPI's greatest weakness imo. It seems to me that there should be a factor that upgrades/downgrades the quality of a win or loss according to how the opponent in question performs in other games... perhaps not even over an entire season, but for some number of games before and after (to mitigate and account for injuries, a creampuff stretch of schedule, etc).

Because of the disconnect, RPI will always be discountable once human committees start subjectively judging and self-anointed experts (some probably with a legitimate claim) will trust their eyeball know-it-when-I-see-it more than numbers... they'll use RPI to justify their decisions when it suits them, and ignore it when it's contrary.

 

[Sometimes]

But, on the same token, the simplicity that might be one of the biggest faults with the RPI is also it's strength. It is a simple formula, and it is applied in the exact same manner to every D1 team. At some level you need some form of metric in order to provide a quantitative manner in which to rank teams. The RPI provides a simple, and fair, way of doing so.

This is not to say that I truly believe that whichever team is ranked #9 at the end of the season is truly the 9th best team in the league, but it is an important compliment to the "eye-test" coaches and AP polls.