Oklahoma downs Baylor

[ndmb]

RPI is relevant as long as it is relevant to the committee. The poster above overstates Baylor's SOS - currently 21 on the site I use and RPI 3.

While UConn does not show well in RPI, that is obviously conference driven. The same website has UConn's OOC strength of schedule as 2. Part of the issue is that the poor teams drive down factor 2 and because they play each other they appear 25+ times in factor 3. RPI as a concept has never bothered me, but it is very, very limited and should be scrapped in favor of something else by the NCAA - however, I strongly feel that margin of victory should not be a factor.

In general, this is a bad year in one respect - the decent teams have played an inordinate number of worse-than-usual little sisters. I don't know if it is travel restrictions (which I think is part of it) in a world of shrinking budgets, but bleh. I worried about RU playing 4 horrible teams and BeK, the RU board stat guy, pointed out that over half the top 50 had played at least that many stinkers.

To me, not accounting for margin of victory is one of the major holes in RPI. Is your concern about spreading playing time to end of the bench players and teams running the score up in blowouts? If so, would you feel differently if the benefit from MOV was capped at 20 or 30 points?

 

[bballnut90]

Baylor should be out of the running for a #1 with no big wins and now two losses.

My seeding rankings for the top 8 are as follows:
1. UCONN
2. Notre Dame-lots of good wins, on pace to win the ACC
3. South Carolina-1 loss team right now, should be able to lose to UT and slide into a #1 seed without any trouble
4. Tennessee-lots of good wins but 4 losses. If they can beat USC in the SEC Tournament I think they hold onto the last #1. If not, they drop to the top or second best #2 seed
5. Maryland-If they win out it'll be a toss up between them and Tenn. If they lose they're down to a 2.
6. Oregon State-Solidly a 2 seed, will be a #1 seed if they win out and Tennessee/Maryland both lose before the tournament
7. Baylor-Solidly a 2, they could probably drop another game and they'll be a 2 still but don't have a chance to be a 1.
8. Florida State-strong resume, could go up to be a stronger two if they win out and lose to ND in the ACC Championship, or they can go down to a 3 seed if they bow out early in the ACC Tournament

 

[EricLA]

I think this Baylor loss and the TN loss propel MD into the lead position for the final #1 seed

Totally agree. I would expect that if/when Tennessee loses in the SEC tourney, IF Maryland runs the table, it will be a no brainer for the 4 top seeds. UCONN, ND, SC and Maryland...

One thing that stood out to me was Maryland's record against the top 50. Much better than I would have thought, and certainly better than Tennessee's or Baylor's...

Plus how perfect would it be to have Tenn as the 2 seed in UCONN's bracket?? Tenn, Baylor, Oregon State and L'ville or FSU as the 2 seeds...

 

[KnightBridgeAZ]

To me, not accounting for margin of victory is one of the major holes in RPI. Is your concern about spreading playing time to end of the bench players and teams running the score up in blowouts? If so, would you feel differently if the benefit from MOV was capped at 20 or 30 points?

While I am not concerned per se with teams running up the score, I suppose that could be an issue. While I don't agree with playing "stupid", if the other team can't stop you you still should be trying to score. Maybe run down the shot clock, require 20 passes or whatever, but you don't give up possessions just for the heck of it in a normal game.

More importantly, a rating system with point differential figured in will always favor teams that win by a large margin and hurt teams that always win the close one. Which isn't really fair - RU for example traditionally played teams close, truly fairly intentionally. Hold down the score and get one more point than the opponent. Yet in those same seasons, we did sometimes top UConn, which has always combined productive scoring and solid defense into an often nation leading differential. I think it distorts the system. The object of the game is to win, not to do it in a certain way.

 

[UcMiami]

Baylor should be out of the running for a #1 with no big wins and now two losses.

My seeding rankings for the top 8 are as follows:
1. UCONN
2. Notre Dame-lots of good wins, on pace to win the ACC
3. South Carolina-1 loss team right now, should be able to lose to UT and slide into a #1 seed without any trouble
4. Tennessee-lots of good wins but 4 losses. If they can beat USC in the SEC Tournament I think they hold onto the last #1. If not, they drop to the top or second best #2 seed
5. Maryland-If they win out it'll be a toss up between them and Tenn. If they lose they're down to a 2.
6. Oregon State-Solidly a 2 seed, will be a #1 seed if they win out and Tennessee/Maryland both lose before the tournament
7. Baylor-Solidly a 2, they could probably drop another game and they'll be a 2 still but don't have a chance to be a 1.
8. Florida State-strong resume, could go up to be a stronger two if they win out and lose to ND in the ACC Championship, or they can go down to a 3 seed if they bow out early in the ACC Tournament

Not a bad list though I would flip MD and TN. TN wins are looking less impressive as teams in the SEC keep losing and their losses except to SC don't look very good.

I actually think SC is benefitting from being 'the last undefeated power 5 team' as their schedule hasn't been all that challenging either. I just am having a hard time buying a bunch of these SEC teams being 'strong' and their OOC was weak - Syracuse, Duke, and the sisters of mercy, until the Uconn game.

 

[DobbsRover2]

RPI is relevant as long as it is relevant to the committee. The poster above overstates Baylor's SOS - currently 21 on the site I use and RPI 3.

While UConn does not show well in RPI, that is obviously conference driven. The same website has UConn's OOC strength of schedule as 2. Part of the issue is that the poor teams drive down factor 2 and because they play each other they appear 25+ times in factor 3. RPI as a concept has never bothered me, but it is very, very limited and should be scrapped in favor of something else by the NCAA - however, I strongly feel that margin of victory should not be a factor.

In general, this is a bad year in one respect - the decent teams have played an inordinate number of worse-than-usual little sisters. I don't know if it is travel restrictions (which I think is part of it) in a world of shrinking budgets, but bleh. I worried about RU playing 4 horrible teams and BeK, the RU board stat guy, pointed out that over half the top 50 had played at least that many stinkers.

Relevant how? Are you on the committee and have some inside information? Please do share more of what's going on there.

The committee has always stated that that brain-dead RPI is just one of "many resources/tools" that "may or may not be utilized by each committee member" of the many statistical tools given to them to employ. From inside stories of the past for both women and men, the RPI tool was long since shunted aside as a major factor in anything except maybe those last group of at-large choices. But the pundits keep popping up here proclaiming that the committee considers it relevant in a major way beyond using it as one of the many easily understandable indexes to explain a decision abut a team when it is useful as a explanation, otherwise when it doesn't they'll run to "good road wins" or a conference tourney championship or one of the many other stats that can explain why they did something.

So why should anyone be spouting RPI nonsense as a major factor in a discussion about #1 seeds? Did the top four RPI teams last year (ND, UConn, Stanford, Duke) all get #1 seeds? Don't think so, as I believe that South Carolina and Tennessee got the third and fourth assignments behind the two obvious undefeated teams.

And as for "UConn does not show well in RPI, that is obviously conference driven," in a less idiotic SOS system such as Massey, UConn was no. 1 in OOC SOS and will finish at #2 in SOS overall this year, instead of something like in the high 30s or 40s in brain-dead RPI. So why shrug off the idiocy of the system as like "oh of course that's the truth and reality of UConn's schedule"? You may see it that way, but Massey plainly sees it differently, and that's where I'm putting my money for any Tournament bets.

 

[triaddukefan]

Great to see Oklahoma win this game.

 

[DobbsRover2]

Massey has the following predicted "most likely" scores for the top five candidates for the fourth #1 seed, plus Princeton, which has no shot at a seed better than a #5 but is rated highly in the system.

Maryland: 75-72 over Baylor, 72-69 over Princeton, 69-66 over Oregon State, 71-67 over Tennessee, 74-70 over Florida State
Baylor: 71-70 over Princton, 68-67 over OSU, 70-68 over UTenn, 72-69 over FSU
Princeton: 65-64 over OSU, 66-65 over UTenn, 69-67 over FSU
Oregon State: 64-63 over UTenn, 66-64 over FSU
Tennessee: 68-66 over FSU
Florida: not favored in any matchup

 

[j66kicker]

Boomer Sooner!

 

[bballnut90]

Not a bad list though I would flip MD and TN. TN wins are looking less impressive as teams in the SEC keep losing and their losses except to SC don't look very good.

I actually think SC is benefitting from being 'the last undefeated power 5 team' as their schedule hasn't been all that challenging either. I just am having a hard time buying a bunch of these SEC teams being 'strong' and their OOC was weak - Syracuse, Duke, and the sisters of mercy, until the Uconn game.

Tennessee's loss to Texas now looks bad, but at the time Texas was a much better team and was looking like a serious Final Four threat early in the season. On the flip side, their loss to UT-Chat doesn't look as bad in hindsight, as Chattanooga has been phenomenal this year and is ranked #18 in the nation. Not a "good" loss, but arguably better than losing to Oklahoma, Washington State or Washington like Baylor/Maryland/Oregon State.

Tennessee still has a big win over Oregon State, wins against A&M, Mississippi State, and out of conference wins against Stanford and Rutgers which aren't hugely impressive, but those are better wins than Baylor has against anybody.

The ACC is the best conference this year, and the SEC is the 2nd best conference. It's a relatively weak year overall for women's basketball. There isn't a lot of star power or top teams outside of Connecticut. Notre Dame is probably the 2nd best team in the nation once again and they are a shell of the squad they were last year. The 2015 graduating class is one of the worst to come through in a while, and at quick glance, 2016 isn't strong outside of Stewart/Loyd/Jefferson/Mitchell. The 2017 and 2018 classes look to be pretty solid, but even some of the stars from those classes (ex. Alexis Jones, Diamond DeShields) are out this season. Even if the SEC isn't very strong this year outside of USC/Tennessee, no other conference besides the ACC is either, so the SEC will get their share of tournament bids and most likely place a good number of teams in the 2nd round/Sweet 16 even if they only get 1 or 0 representatives in the Final Four.

 

[VFLfan]

Totally agree. I would expect that if/when Tennessee loses in the SEC tourney, IF Maryland runs the table, it will be a no brainer for the 4 top seeds. UCONN, ND, SC and Maryland

Plus how perfect would it be to have Tenn as the 2 seed in UCONN's bracket?? Tenn, Baylor, Oregon State and L'ville or FSU as the 2 seeds...

Eric...while I agree with you about Maryland getting the last #1 seed if that happens, I do not think the committee will put UTenn in the same bracket as UCONN...for that to happen, UTenn would have to be the worst of the #2 seeds and with them still being in contention for the last #1 seed...how can they be the lowest of the #2s?

And also Miami, you say besides UTenn's loss to USC, their other losses look bad? So an 11 pt loss to ND on their home floor is bad? Or a loss to Texas when they were undefeated and still had their best player playing on their home floor is bad? Granted, ain't no way I can justify losing to Chatt (and I live in Chatt too!) even while they are now up to #18 in one poll...THAT should have never happened! I just don't think our losses are that "bad" as you put it. I'd love if you could provide me your insights as to why you think so...and I truly ask with all due respect. That's why I'm here...to hear other's opinions!

 

[DobbsRover2]

Indeed, UTenn certainly does not have any of the poor losses that they've gotten in recent years, since after an ehh loss to Arkansas State, Chattanooga has run off 21 wins in a row, including one over the team that beat UConn. If the Vols can handle GA tonight, then Vandy and whoever their first SECT opponent is, they should be set for a #2 seed, and if MD-Baylor-OSU have any problems and UTenn at least gets to the SECT championship game, a #1 seed could still be out there. But the committee in past years hasve usually given strong weight to teams that win both of their conference championships even if they were in a weak conference such as the PAC for Stanford for many years, so if MD, Baylor, and OSU get a double it would be highly unlikely for UTenn to be at all close to a #1 seed without the upset of USCar.

 

[Vowelguy]

Eric...while I agree with you about Maryland getting the last #1 seed if that happens, I do not think the committee will put UTenn in the same bracket as UCONN...for that to happen, UTenn would have to be the worst of the #2 seeds and with them still being in contention for the last #1 seed...how can they be the lowest of the #2s?

That's not how it works. The women's tournament does not use the S-curve.
The #5 team (ie, the highest 2 seed) gets placed in the regional closest to home, subject to other rules. If TN were the #5 and SC were in G-boro, then TN would get sent to Albany or OKC -- it is roughly equidstant from both.

There are other issues that come into play -- making the brackets balanced, and other conflicts that arise that could cause the 2 seeds to be shifted again. But that's more or less how it works. It is possible to have the overall #1 and overall #5 in the same region.

If MD is a #1 and TN is not, I'd say there's a very good chance the Vols end up in Albany.

 

[VFLfan]

That's not how it works. The women's tournament does not use the S-curve.
The #5 team (ie, the highest 2 seed) gets placed in the regional closest to home, subject to other rules. If TN were the #5 and SC were in G-boro, then TN would get sent to Albany or OKC -- it is roughly equidstant from both.

There are other issues that come into play -- making the brackets balanced, and other conflicts that arise that could cause the 2 seeds to be shifted again. But that's more or less how it works. It is possible to have the overall #1 and overall #5 in the same region.

If MD is a #1 and TN is not, I'd say there's a very good chance the Vols end up in Albany.

Gotcha! Thanks for the clarification. I get lost when it comes to all the rules of the tourney seeding.

 

[Zorro]

I would love to see a tape of that game, just to see Kim shredding gold lame on the sidelines.

 

[DobbsRover2]

Gotcha! Thanks for the clarification. I get lost when it comes to all the rules of the tourney seeding.

Not sure if the information you were just given is true, but from reading the Women's Tournament Seeding Principles and Procedures, you can make your own judgment. There are a huge number of considerations for trying to balance the brackets and all of the guidelines that need to be employed.

First off, if it was true, why did #2 seed Baylor end up being sent to Notre Dame's regional last year instead getting dumped in with UConn to the nearer Lincoln Neb. regional? One #2 seed Stanford was going to be in the Stanford regional, and Baylor would have to been on the S-curve ahead of Duke and WVU, but they did not go to the nearest regional. So other factors like possibly the S-curve were in play here to send Duke (likely no. 8 team) to Lincoln and Baylor to ND.

Also, it is simply not true as the above poster said that "The women's tournament does not use the S-curve." If he had read the guidelines, he would note that it is used all over the place as in "Once the s-curve is finalized, it remains unchanged while placing the teams into the championship bracket." It is the core of the seeding procedures, though factors obviously such as teams being in the same conference is an important criteria. But the committee does refer to the S-curve until the end of the procedures to make sure they are setting up fairly balanced regions, though they will never be equally divided by the perceived seeding strength.

In the event of a high #2 seed for UTenn with a #1 seed USCar in Greensboro, it is highly unlikely the Vols are also in Greensboro, as the one time the committee did that to the Vols, their lives were in danger (and it is generally considered a big no-no in the guidelines unless there is no other possible choice. Probably OSU will be in Spokane as a #2 seed (outside chance that the Beavers would get a #1 seed and that Vols would be there as a #2), and the yes the likelihood would then for the Vols to be out in OK City with ND (though for some reason I forgot Creme has them in Spokane), or in Albany with you know who.

And I can't believe that the comments for Creme's latest Bracketology can just turn into Geno bashing for Vol fans. Just mindless.

 

[KnightBridgeAZ]

Relevant how? Are you on the committee and have some inside information? Please do share more of what's going on there.

The committee has always stated that that brain-dead RPI is just one of "many resources/tools" that "may or may not be utilized by each committee member" of the many statistical tools given to them to employ. From inside stories of the past for both women and men, the RPI tool was long since shunted aside as a major factor in anything except maybe those last group of at-large choices. But the pundits keep popping up here proclaiming that the committee considers it relevant in a major way beyond using it as one of the many easily understandable indexes to explain a decision abut a team when it is useful as a explanation, otherwise when it doesn't they'll run to "good road wins" or a conference tourney championship or one of the many other stats that can explain why they did something.

So why should anyone be spouting RPI nonsense as a major factor in a discussion about #1 seeds? Did the top four RPI teams last year (ND, UConn, Stanford, Duke) all get #1 seeds? Don't think so, as I believe that South Carolina and Tennessee got the third and fourth assignments behind the two obvious undefeated teams.

And as for "UConn does not show well in RPI, that is obviously conference driven," in a less idiotic SOS system such as Massey, UConn was no. 1 in OOC SOS and will finish at #2 in SOS overall this year, instead of something like in the high 30s or 40s in brain-dead RPI. So why shrug off the idiocy of the system as like "oh of course that's the truth and reality of UConn's schedule"? You may see it that way, but Massey plainly sees it differently, and that's where I'm putting my money for any Tournament bets.

I agree with some of your factual statements - it is only one tool the committee uses, etc - although, since as I said, it is of extremely limited use at all, as I said they should discontinue using it. I never claimed it reflects accurately any schedule strengths - or rather, why does it include results against very weak teams that a good team should beat - a better comparison is between teams that are "in the same general level". I never said that was the truth of UConn's schedule - you could have played the top teams in every conference and based on the tough games you did play OOC - probably beaten most of them by 20 points or more. That is immaterial to saying that the committee, or some committee members, place too much weight on the RPI, therefore it is necessary to include it in the many more important seeding factors.

Also, like it or not, there is a section in your link that clearly states that the actual placement of teams is heavily influenced by geography - at least when they want it to be.

 

[cabbie191]

RPI is relevant as long as it is relevant to the committee. The poster above overstates Baylor's SOS - currently 21 on the site I use and RPI 3.

While UConn does not show well in RPI, that is obviously conference driven. The same website has UConn's OOC strength of schedule as 2. Part of the issue is that the poor teams drive down factor 2 and because they play each other they appear 25+ times in factor 3. RPI as a concept has never bothered me, but it is very, very limited and should be scrapped in favor of something else by the NCAA - however, I strongly feel that margin of victory should not be a factor.

In general, this is a bad year in one respect - the decent teams have played an inordinate number of worse-than-usual little sisters. I don't know if it is travel restrictions (which I think is part of it) in a world of shrinking budgets, but bleh. I worried about RU playing 4 horrible teams and BeK, the RU board stat guy, pointed out that over half the top 50 had played at least that many stinkers.

Do you have any sense what being in the Big ten has done for your team's overall travel compared to your days in the BE? Any compensation in who you play OOC?

 

[DobbsRover2]

I agree with some of your factual statements - it is only one tool the committee uses, etc - although, since as I said, it is of extremely limited use at all, as I said they should discontinue using it. I never claimed it reflects accurately any schedule strengths - or rather, why does it include results against very weak teams that a good team should beat - a better comparison is between teams that are "in the same general level". I never said that was the truth of UConn's schedule - you could have played the top teams in every conference and based on the tough games you did play OOC - probably beaten most of them by 20 points or more. That is immaterial to saying that the committee, or some committee members, place too much weight on the RPI, therefore it is necessary to include it in the many more important seeding factors.

Also, like it or not, there is a section in your link that clearly states that the actual placement of teams is heavily influenced by geography - at least when they want it to be.

Absolutely right, and that is a no-brainer, and it is exactly why I said that it is then a puzzle that Baylor did not go to Lincoln if that is the first and only factor. Geography is one of those many considerations that the committee uses, yeah we know that. What is simply not true is that the women's committee does not use the S-curve as a factor. If you did indeed read the guidelines, you will see that they do use it quite prominently.

 

[Beemer]

23-13 foul difference in favor of OU at home. I bet Mulkey was peeling off her clothes and running up and down the sidelines.

Aiiiiiiiiiii my eyes! My eyes!!!

 

[Beemer]

Great to see Oklahoma win this game.

Oklahoma has a good team and a good coach. I've seen them a few times on TV, they've been fun to watch.

 

[baylorprof]

Oh you mean we haz to be a TWO! Oh no! But that's OK just PLEASE don't throw us in dat OKC reegnul where our fans can travel Waco one weekend and OKC the next! We'll take our chances with OU on the OKC court 2 outta 3.

[Seriously, most of us feel if we can get to elite 8 that's not only our ceiling, but without a legit BIG a darn GOOD season.]

 

[Wbbfan1]

I'd take the PAC12 top to bottom over the SEC. After South Carolina and Lady Vols who are the good SEC teams. IMHO both Kentucky and A&M are overrated. I'd take Oregon State, Ariz State, Cal and Stanford to win the majority of the games if they were to play a round robin tournament with the 4 best SEC teams.

Tennessee's loss to Texas now looks bad, but at the time Texas was a much better team and was looking like a serious Final Four threat early in the season. On the flip side, their loss to UT-Chat doesn't look as bad in hindsight, as Chattanooga has been phenomenal this year and is ranked #18 in the nation. Not a "good" loss, but arguably better than losing to Oklahoma, Washington State or Washington like Baylor/Maryland/Oregon State.

Tennessee still has a big win over Oregon State, wins against A&M, Mississippi State, and out of conference wins against Stanford and Rutgers which aren't hugely impressive, but those are better wins than Baylor has against anybody.

The ACC is the best conference this year, and the SEC is the 2nd best conference. It's a relatively weak year overall for women's basketball. There isn't a lot of star power or top teams outside of Connecticut. Notre Dame is probably the 2nd best team in the nation once again and they are a shell of the squad they were last year. The 2015 graduating class is one of the worst to come through in a while, and at quick glance, 2016 isn't strong outside of Stewart/Loyd/Jefferson/Mitchell. The 2017 and 2018 classes look to be pretty solid, but even some of the stars from those classes (ex. Alexis Jones, Diamond DeShields) are out this season. Even if the SEC isn't very strong this year outside of USC/Tennessee, no other conference besides the ACC is either, so the SEC will get their share of tournament bids and most likely place a good number of teams in the 2nd round/Sweet 16 even if they only get 1 or 0 representatives in the Final Four.

 

[DobbsRover2]

Oh you mean we haz to be a TWO! Oh no! But that's OK just PLEASE don't throw us in dat OKC reegnul where our fans can travel Waco one weekend and OKC the next! We'll take our chances with OU on the OKC court 2 outta 3.

[Seriously, most of us feel if we can get to elite 8 that's not only our ceiling, but without a legit BIG a darn GOOD season.]

What do mean no legit big? Nina Davis is clearly at least 6'5" to be pulling down all those rebounds.

 

[Vowelguy]

The "s-curve" typically refers to the idea of seeding the teams such that 1 vs 8, 2 vs 7, 3 vs 6. So that you end up with
.. Geo1 Geo2 Geo3 Geo4
.. #1--->#2--->#3--->#4->|
|<-#8<---#7<---#6<---#5<-|
|->#9...#10...#11...#12
..#16...#15...#14...#13

If you draw a line in order of ranking, you create an "S" shape as in red above. This is what the men's committee uses.

The women's committee uses the phrase "s-curve", but they seem to use it to refer to their ranking of teams from 1 to 64. "The committee will create an s‐curve (i.e., rank of the teams 1 through 64)." To me, that's not what "s-curve" means.

The men's committee, to my understanding, does place the teams roughly according to the 1-64, 2-63, etc format. *Then*, it goes and moves things around for the sake of avoiding conference matchups and certain geography.
In contrast, the women's committee goes to geography first for placing the teams. So, when they get to the #5 team, they don't look to placement with the #4; instead they're looking for putting the #5 (the top 2 seed) closest to home. This is a shift in procedures that they started several years ago.
"The committee will attempt to assign each team to the most geographically compatible regional and first‐/second‐round site, by order of the s‐curve. When multiple teams are a similar distance from a site, the team seeded higher in the s‐curve will be assigned to the closest geographical proximity site."

My understanding, and the way that I read experts like Creme, is that geography is considered first. It is not the sole factor, though, as I noted in my previous post. They do need to balance the brackets, and avoid conference matchups (this year for the first time, 2 teams from the same conference can't be in the same region's top 4, unless >four teams get a top-4 seed).

Unlike the men's bracket, where it's reasonable to assume that the top 1 seed will play the bottom 2 seed, there is no such expectation on the women's side.