3rd in Rpi

[Fishy]

Leave snarky to the people who are good at it.

Where am I wrong?

 

[Danzz]

I agree, RPI this early in the season is worthless.

Also, if you want a better rating system go to kenpom.com. UConn is currently ranked 28th using his analysis.

I'd even say that was generious. We play nobody outside the BE this eason.

 

[Gurleyman]

Normally, we would be around 20-30 in the RPI around now and have to rely on a tougher Big East schedule to get our RPI to go up. Now our RPI is in the top five, and we have the tougher Big East schedule in front of us to solidify it. Now if we lose a lot of Big East games (7 or 8, say), our RPI will go down, but we'll still be safely in the at-large pool, whereas if our RPI was 25 or 30 now, that type of conference season could put us in danger zone. And if we only lose 4-5 Big East games, we'll probably be a two or three seed (depending on what other teams do and who our wins are against). Our current RPI isn't definitive, but it's definitely nice to be where we're at.

 

[Waquoit]

Sure, in my first post I did point out that Ken Pom has them ranked 28th using his analysis, but that has nothing to do with the merits of their #3 spot in RPI. The RPI has it's formula and using the current data set, the formula says UConn is #3.

Do you think this UConn team today is closer to being 3rd best team in the country or the 28th?

 

[RuffRuff]

This Fishy/BU argument sounds a lot like why I go to counseling to figure out how to deal with my wife - arguing two different points.

Hard to argue with Fishy's take on this one - Uconn has capitalized on it's OOC schedule by earning a #3 RPI - it's not too early to qualify that as valuable. Unless you haven't been watching, OOC success is an enormous variable when total resumes are observed. Whether it was intentional or not, they've done a nice job of what Duke does often - no lethal early season matchups to accumulate the losses, but smart games against very beatable teams that aren't way down on the totem pole.

 

[arnie3]

What's taken into account by the RPI is the extraordinary amount of wins by non-BCS schools over BCS schools- last time I heard it was around 100 wins. Add to that, Long beach State, Wagner, and UNLV and you get a much better sample of how close the margin is between the mid-major and major schools. Even JC referred to it when he said that Fairfield could beat some BE schools. If not them, then Iona for sure. Here we are at the mid-point of the regular season and some of these teams have a place in the RPI top 50 that is usually reserved for the BCS schools. How good is Murray State? They may remain undefeated through the regular season. They beat a pretty good Southern Miss team, creamed Dayton and got by Memphis-in Memphis.
I think that the argument that some of these schools should not have such a high RPI is bogus. I would rather see some team like an Iona or Fairfield get a bid even if they don't win their conference than a VT or Miami that is less deserving. Harvard should have been in last year. Based on what Princeton did (a near upset of Kentucky), both were deserving!

 

[BlueMAB]

How could the RPI today be meaningless? The data that puts at #3 isn't erased with the start of conference play, our results and how they RPI calculates them will be built upon nearly half a season's games that left us at #3. We are in a great position right now. Nobody said it guarantees anything in March, or that it is more imporatant that what happens next.

Arguing otherwise is like saying Drummond dunks too much, or McCombs wouldn't average many yards per carry if you don't count the long runs.

 

[DogMania]

How could the RPI today be meaningless? The data that puts at #3 isn't erased with the start of conference play, our results and how they RPI calculates them will be built upon nearly half a season's games that left us at #3. We are in a great position right now. Nobody said it guarantees anything in March, or that it is more imporatant that what happens next.

Arguing otherwise is like saying Drummond dunks too much, or McCombs wouldn't average many yards per carry if you don't count the long runs.

Meaningless might be a tad strong, but incomplete or inconclusive might be better terms to use. Although we're at the halfway point of the hoops season, all the teams are just beginning conference play. If a team like UConn who's played 13 underdogs and beat all but one, begins to lose games to fellow BE teams that are likely more within their talent range, expect their RPI to drop. Now if they keep on winning a large percentage of them, their RPI should remain very high. Conversely, mid-major teams Zags, UNLV, Murray State, Long Beach State, etc. (I counted 8 mid-majors in the top 25 RPI) will likely take a bit of a RPI hit as they go through their conference schedule, especially if they lose a few of them.

So the RPI isn't completely irrelevant at this point, but simply lacks enough data to really give us a good picture of how the field stacks up against itself. Now if you think otherwise, do you really think Seton Hall is the 4th best team in the nation or anywhere close to that? Is Mich State is the 6th best out there? Are the Zags the 7th? UNLV 10th? So and and so on...

You have to remember although the RPI is based on a teams W-Ls and SOS, in addition to a team's W-Ls changing over the season, the SOS is ever in flux. What looked like a good SOS opponent early on doesn't exactly calculate as well over time if I understand how this all works. If some of 13 teams we played thus far begins to pile up some conference losses, their SOS boost won't be such a boost. Fortunately for us we still have 6 top 25 RPI teams on our schedule and we play two of them twice (SU & SH...hopefully SH continues to do well except against us). Those games should help our RPI to remain high. Hopefully the teams at the bottom of the BE won't hurt our RPI too badly. Without doing some in depth analysis, it's hard to tell if the bottom of our conference will hurt us any more than say the bottom of the Big-10, ACC, SEC, etc.

One way to look at this...at least I look at it this way...the current order of teams based on the RPI is not an indicator of how good each team is playing at the moment and how they stack up against the field. But the RPI will be a better indicator of how teams stack up against the field in about 20 games from now. I'm not saying that the final RPI is an exact scientific indicator to that, but until the season nears its completion where the most of data is in, along with the relationships between common opponents are fully compounded, the RPI picture in incomplete.

I took a snapshot of this weeks RPI with the intent of comparing it later on during the season. Let's see how close or far off it turns out.

Bottom line, keep on winning especially against the top BE teams and finish strong at the end of the regular season and conference tourney to keep the RPI high. As for seeding, in addition to teams overall RPI, W-Ls and poll rankings, the committee looks at how well you did against to top 10 RPI, top 25, etc. That's where the RPI has its importance.

 

[ctchamps]

Meaningless might be a tad strong, but incomplete or inconclusive might be better terms to use.

Exactly.

 

[caw]

Yeah, meaningless is an absurd overstatement. Non-predictive of future success or incomplete is perfectly accurate.

UConn currently sits at #8 in both the coaches and AP polls. That with a #3 RPI would be at worst a 3-4 seed. The thing holding UConn back from a top 1 or 2 seed would be obviously having the worst loss of the top 12 teams while not having more than 1 really good win (Harvard is a really good win, regardless of what others may say). I mean seriously UConn's loss was really, really bad compared to the rest of the top twelve. The other 11 teams have 12 losses combined and have lost to basically each other or another top 20 team: #12 Indiana x2, #2 UK, #17 UNLV, #6 OSU, #14 KU x2, #3 UNC, #5 Duke, #2 UK, #9 Georgetown, and #17 Michigan State.

I think all of the other 11 teams would have legit claim at a 1-3 seed right now ahead of UConn. Florida wouldn't with 3 losses (one to Rutgers), neither would Mississippi State or Kansas. Going lower shows teams with bad losses like UConn or muli-loss teams (to teams not in the top 20). I guess Murray State could be the final 3 seed ATM, pushing UConn to a four, but RPI might be the tie-breaker.

Seton Hall would make the tournament if the tournament was tomorrow. They are receiving votes in both polls as well as having an RPI of #4.

Of course this will all change as the season plays out and upsets occur. What probably won't change is the UCF will probably be the worst loss going forward of any team challenging for a 1-3 seed.

 

[auror]

I would argue that it SO CLOSE to meaningless to be able to be called so accurately, because not only does OUR upcoming win/loss record matter, but also the future win/loss records of our previous AND future OPPONENTS. We could win the rest of our games, but if all of our past & future opponents lose the rest of theirs, we will have a RPI.

Sure, the current games will remain on the record and will impact things somewhat, but each game we play dilutes the relative importance of the previous teams' games AND each game our opponents play dilutes the relative importance of the previous games each of them have played also. So you could say, hey, we're halfway through the season, we have half the data we're going to have! But that's not true at all. We have less than half the data for less than half the teams...for example say around 13 games for 13 teams so far = 169 games of data impacting UConn's current RPI. Season's end there will be around 32 games (repeated teams count each time, each instance weighted the same except road variable) each having played around 32 games = 1024 games. So at the end the current data will make up 17% of the final impact on UConn's RPI. And that doesn't even take into account that the RPI measure's the W/L of our opponent's opponents.

It is not like we built a foundation and now are building the house. We're about to start building the house, but the foundation is going to be rebuilt at the same time.

As Palantine said, we're 13 games in now...the RPI is not meaningless. What it means is that nearly halfway through the season, and effectively through the the out-of-conference portion of the year, UConn is where they need to be to secure a one or two seed.

Unfortunately, it is mathematically possible for UConn to win all of their remaining games and still have their RPI fall out of position that would put them in contention for a 1 or 2 seed. Therefore, it does not matter whatsoever right now that UConn is in position for a 1 or 2 seed. Making the RPI at this stage of the season meaningless.

 

[Gurleyman]

Unfortunately, it is mathematically possible for UConn to win all of their remaining games and still have their RPI fall out of position that would put them in contention for a 1 or 2 seed. Therefore, it does not matter whatsoever right now that UConn is in position for a 1 or 2 seed. Making the RPI at this stage of the season meaningless.

It's also mathematically possible for me to bang Scarlett Johansson, but that's just as likely as that scenario.

(I kid, of course - I think Scarlett likes me)

 

[BlueMAB]

Meaningless might be a tad strong, but incomplete or inconclusive might be better terms to use. Although we're at the halfway point of the hoops season, all the teams are just beginning conference play. If a team like UConn who's played 13 underdogs and beat all but one, begins to lose games to fellow BE teams that are likely more within their talent range, expect their RPI to drop. Now if they keep on winning a large percentage of them, their RPI should remain very high. Conversely, mid-major teams Zags, UNLV, Murray State, Long Beach State, etc. (I counted 8 mid-majors in the top 25 RPI) will likely take a bit of a RPI hit as they go through their conference schedule, especially if they lose a few of them.

So the RPI isn't completely irrelevant at this point, but simply lacks enough data to really give us a good picture of how the field stacks up against itself. Now if you think otherwise, do you really think Seton Hall is the 4th best team in the nation or anywhere close to that? Is Mich State is the 6th best out there? Are the Zags the 7th? UNLV 10th? So and and so on...

You have to remember although the RPI is based on a teams W-Ls and SOS, in addition to a team's W-Ls changing over the season, the SOS is ever in flux. What looked like a good SOS opponent early on doesn't exactly calculate as well over time if I understand how this all works. If some of 13 teams we played thus far begins to pile up some conference losses, their SOS boost won't be such a boost. Fortunately for us we still have 6 top 25 RPI teams on our schedule and we play two of them twice (SU & SH...hopefully SH continues to do well except against us). Those games should help our RPI to remain high. Hopefully the teams at the bottom of the BE won't hurt our RPI too badly. Without doing some in depth analysis, it's hard to tell if the bottom of our conference will hurt us any more than say the bottom of the Big-10, ACC, SEC, etc.

One way to look at this...at least I look at it this way...the current order of teams based on the RPI is not an indicator of how good each team is playing at the moment and how they stack up against the field. But the RPI will be a better indicator of how teams stack up against the field in about 20 games from now. I'm not saying that the final RPI is an exact scientific indicator to that, but until the season nears its completion where the most of data is in, along with the relationships between common opponents are fully compounded, the RPI picture in incomplete.

I took a snapshot of this weeks RPI with the intent of comparing it later on during the season. Let's see how close or far off it turns out.

Bottom line, keep on winning especially against the top BE teams and finish strong at the end of the regular season and conference tourney to keep the RPI high. As for seeding, in addition to teams overall RPI, W-Ls and poll rankings, the committee looks at how well you did against to top 10 RPI, top 25, etc. That's where the RPI has its importance.

Like Fishy said, the RPI today is not meaningless, the rest is just noise.

Nobody made the argument that we can take the second half of the season off because of the RPI this week.

 

[Mr. Wonderful]

RPI is so misunderstood.

You can attain a high RPI two ways - by playing the top 100 teams, or by avoiding playing the 100 worst teams. The latter is never discussed.