RebKell's Junkie Boards
Board Junkies Forums
 
Log in Register FAQ Memberlist Search RebKell's Junkie Boards Forum Index

2018-2019 Bracketology
Goto page Previous  1, 2
 
Post new topic   Reply to topic    RebKell's Junkie Boards Forum Index » NCAA Women's Basketball - General Discussion
View previous topic :: View next topic  
Author Message
myrtle



Joined: 02 May 2008
Posts: 23641



Back to top
PostPosted: 01/08/19 10:17 pm    ::: Reply Reply with quote

The SEC RPI is a bit of the chicken and the egg. The SEC is given a high starting number just because they are expected to be good. Thus, just by playing each other their RPI stays high. For instance, I haven't seen anything that SoCaro has done that deserves them to either be a tournament team or be ranked. Ditto Georgia. Ditto Auburn.



_________________
“Try to be a rainbow in someone's cloud.”
― Maya Angelou
PUmatty



Joined: 10 Nov 2004
Posts: 14484
Location: Chicago


Back to top
PostPosted: 01/09/19 11:31 am    ::: Reply Reply with quote

myrtle wrote:
The SEC RPI is a bit of the chicken and the egg. The SEC is given a high starting number just because they are expected to be good. Thus, just by playing each other their RPI stays high. For instance, I haven't seen anything that SoCaro has done that deserves them to either be a tournament team or be ranked. Ditto Georgia. Ditto Auburn.


That is not how RPI works.

Rankings work like that, but RPI is based solely on the year's wins and losses.


myrtle



Joined: 02 May 2008
Posts: 23641



Back to top
PostPosted: 01/09/19 11:52 am    ::: Reply Reply with quote

My bad. I was wrong in saying that. I should have said the RPI can be and is gamed, especially by many of the P5 teams...which is why it's being dropped as a parameter for determining tourny teams on the men's side. They've more or less said it is a worthless number. So if it's being used as the reason to include teams like LSU who has already lost to several non-tournament teams, then it's a problem.



_________________
“Try to be a rainbow in someone's cloud.”
― Maya Angelou
ucbart



Joined: 21 Nov 2004
Posts: 1368
Location: New York


Back to top
PostPosted: 01/09/19 12:01 pm    ::: Reply Reply with quote

cthskzfn wrote:
Coyotes wrote:
Gonzaga's hosting in Minneapolis?

Also, he thinks that Bradley will beat Drake for the MVC?


CT in Albany? He's nuts. Laughing


I also don't see how Syracuse won't be in Albany. If they go through the ACC and only lose two games, to say ND and Louisville, Q will have a legit gripe about not being a 3 in Albany.


PUmatty



Joined: 10 Nov 2004
Posts: 14484
Location: Chicago


Back to top
PostPosted: 01/09/19 12:03 pm    ::: Reply Reply with quote

myrtle wrote:
My bad. I was wrong in saying that. I should have said the RPI can be and is gamed, especially by many of the P5 teams...which is why it's being dropped as a parameter for determining tourny teams on the men's side. They've more or less said it is a worthless number. So if it's being used as the reason to include teams like LSU who has already lost to several non-tournament teams, then it's a problem.


Absolutely.


Matt5762



Joined: 27 Feb 2005
Posts: 502



Back to top
PostPosted: 01/09/19 12:57 pm    ::: Reply Reply with quote

myrtle wrote:
My bad. I was wrong in saying that. I should have said the RPI can be and is gamed, especially by many of the P5 teams...which is why it's being dropped as a parameter for determining tourny teams on the men's side. They've more or less said it is a worthless number. So if it's being used as the reason to include teams like LSU who has already lost to several non-tournament teams, then it's a problem.


But this is probably a bad example. The SEC as a whole is ranked very similarly (and in a few cases notably better) in a more logical ranking system like Massey...

RPI-Massey
15-3 Miss State
24-16 Kentucky
23-19 Tennessee
34-27 S Carolina
31-30 Missouri
39-40 Auburn
43-43 Texas A&M
52-49 LSU
54-55 Arkansas
119-64 Georgia
105-113 Alabama
172-163 Vandy
233-221 Florida
228-226 Ole Miss


myrtle



Joined: 02 May 2008
Posts: 23641



Back to top
PostPosted: 01/09/19 3:48 pm    ::: Reply Reply with quote

hmmm. I haven't looked at how Massey gets his number...but right off the bat, I have to question MissSt being #3. They've played four top 25 teams and beaten three of them (that's good) but lost to the only top 10 team. How does that make them #3? Otherwise their schedule has been cupcake after cupcake. Any system that gives too much credit for beating cupcakes is suspect.



_________________
“Try to be a rainbow in someone's cloud.”
― Maya Angelou
calbearman76



Joined: 02 Nov 2009
Posts: 3221
Location: Carson City


Back to top
PostPosted: 01/09/19 5:23 pm    ::: Reply Reply with quote

myrtle wrote:
hmmm. I haven't looked at how Massey gets his number...but right off the bat, I have to question MissSt being #3. They've played four top 25 teams and beaten three of them (that's good) but lost to the only top 10 team. How does that make them #3? Otherwise their schedule has been cupcake after cupcake. Any system that gives too much credit for beating cupcakes is suspect.



Massey's ratings are based much more on point differential.. If Mississippi St beats Stephen F Austin by 50 points and Texas only beats them by 40, Mississippi St is assumed to be better. It is not a perfect system by any means, particularly when the schedule is very weak, but it does have some validity. I believe it is a good guideline (certainly better than the RPI) but it has its deficiencies as well. MSU looks like the best team in the SEC, but whether they are 3 or 10 in the nation is hard to know at this point. Now that they are into the SEC schedule we will get a better understanding. If they are truly a top 3 team they should be able to get through the SEC with no more than 1 loss.


linkster



Joined: 27 Jul 2012
Posts: 3693



Back to top
PostPosted: 01/09/19 5:35 pm    ::: Reply Reply with quote

myrtle wrote:
hmmm. I haven't looked at how Massey gets his number...but right off the bat, I have to question MissSt being #3. They've played four top 25 teams and beaten three of them (that's good) but lost to the only top 10 team. How does that make them #3? Otherwise their schedule has been cupcake after cupcake. Any system that gives too much credit for beating cupcakes is suspect.


Here's a link to Massey that explains their methodology.
https://www.masseyratings.com/theory/massey.htm

I agree about cupcake wins but massey (unlike other ranking systems) actually overweights games against top teams. The other factor that systems like RPI ignore is scoring margins. Beating a weak opponent by a narrow margin can work against a team's Massey ranking.


Shades



Joined: 10 Jul 2006
Posts: 49620



Back to top
PostPosted: 01/12/19 6:37 pm    ::: Reply Reply with quote




_________________
Nnekalonians 1:14 - Thou shalt not accept that which is not earned


Last edited by Shades on 01/12/19 10:20 pm; edited 1 time in total
CompSci87



Joined: 15 Mar 2009
Posts: 693
Location: Palo Alto, CA


Back to top
PostPosted: 01/12/19 6:43 pm    ::: Reply Reply with quote

Thanks for the Belibi post. Wrong thread, though!


Phil



Joined: 22 Oct 2011
Posts: 985



Back to top
PostPosted: 01/13/19 12:04 pm    ::: Reply Reply with quote

calbearman76 is correct to point out that Massey ratings a better than the RPI. There are several factors but the single biggest one is that RPI ignores point differential. Massey does incorporate, not just the win but the size of the win. If I may be a little bit picky, it's technically not the point differential but more likely the ratio. Winning by 10 points 90 – 80, is not the same as winning by 10 points 50 – 40. Although there explanation doesn't give enough information about the math to say for certain, my guess is that winning 100 – 80 is about the same as winning 50 – 40.

Basketball scores tend to follow in a narrower range than football scores, so the difference between ratios and point differentials is modest and not terribly important, although the fact that the differential matters and is ignored in RPI is a big deal.

[Minor wording correction]




Last edited by Phil on 01/17/19 12:21 pm; edited 2 times in total
calbearman76



Joined: 02 Nov 2009
Posts: 3221
Location: Carson City


Back to top
PostPosted: 01/13/19 4:12 pm    ::: Reply Reply with quote

Phil wrote:
calbearman76 is correct to point out that Massey ratings a better than the RPI. There are several factors but the single biggest one is that RPI ignores point differential. Massey does incorporate, not just the win but the size of the win. If I may be a little bit picky, it's technically not the point differential but more likely the ratio. Winning by 10 points 90 – 80, is not the same as winning by 10 points 50 – 40. Although there explanation doesn't give enough information about the math to say for certain, my guess is that winning 100 – 80 is about the same as winning 50 – 40.

Basketball scores tend to follow in a narrower range than football scores, so the difference between ratios and point differentials is modest and not terribly important, although the fact that the differential matters and is ignored and RPI is a big deal.


There are so many ways to model outcomes and rank teams. I actually liked the Sagarin rankings most, but unfortunately he doesn't seem to be doing women's basketball this season. If I were to build my own system it would include a power rating component and a win component. The power rating would be fully points based, but would only include the first 35 minutes of each game. The last five minutes would not be considered because of the use of subs and fouling which can make a final score not reflect how close a game actually was. The second part would be a win model that would only consider wins and losses. This would be given more weight for competitive contests. Of course the difficulty in developing such a system is that there is no source for 35 minute scores, other than checking play-by-play of every game individually.

Of course I would like to see the NCAA use a better system than the RPI. There are all manner of factors that can be built in, including strength of schedule, play over the latter part of the season, and ideally those factors can be discussed and evaluated. The policymakers could provide the structure and then the math geeks could create the system.


Marquette Fan



Joined: 06 Mar 2005
Posts: 2334



Back to top
PostPosted: 01/14/19 7:58 pm    ::: Reply Reply with quote

Updated bracketology out this morning from Charlie Creme:

http://www.espn.com/womens-college-basketball/bracketology

The home court locations look correct on this projection. Marquette is up to a projected 3 seed in this one - wow! I'm hoping they can at least get a 4 seed this year - still expecting at least a couple head scratching losses so don't expect to see them get a 3. Butler is projected at winning the Big East right now with a 9 seed - with DePaul a 7 to make 3 BE teams projected in the field. Butler played a weak non-conference schedule but has my attention with their 5-0 start in BE play.


myrtle



Joined: 02 May 2008
Posts: 23641



Back to top
PostPosted: 01/14/19 8:40 pm    ::: Reply Reply with quote

If this actually became reality, UConn would have the easiest bracket by far. Of course that would just be normal.



_________________
“Try to be a rainbow in someone's cloud.”
― Maya Angelou
Phil



Joined: 22 Oct 2011
Posts: 985



Back to top
PostPosted: 01/17/19 12:20 pm    ::: Reply Reply with quote

The goal of the selection committee is to produce balanced brackets.

We all know (or at least think we know) that they often fall short of that ideal, but if they meet that ideal, then by definition, the bracket that has the top overall seed will have slightly weaker strength for the remaining 15 compared to the other three brackets. UConn has often been the top overall seed, so it is surprising that this would be a common result. It has nothing to do with UConn it has to do with the seeding principles.


ArtBest23



Joined: 02 Jul 2013
Posts: 12983



Back to top
PostPosted: 01/17/19 12:47 pm    ::: Reply Reply with quote

Phil wrote:
The goal of the selection committee is to produce balanced brackets.



Is it? I think that goal may be secondary to selling tickets and filling arenas.

Minimizing travel costs may also come before competitive balance or fairness.


Display posts from previous:   
Post new topic   Reply to topic    RebKell's Junkie Boards Forum Index » NCAA Women's Basketball - General Discussion All times are GMT - 5 Hours
Goto page Previous  1, 2
Page 2 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB 2.0.17 © 2001- 2004 phpBB Group
phpBB Template by Vjacheslav Trushkin