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

RPI benchmarks

 
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
calbearman76



Joined: 02 Nov 2009
Posts: 5152
Location: Carson City


Back to top
PostPosted: 01/18/18 3:43 am    ::: RPI benchmarks Reply Reply with quote

Trying to evaluate teams based on their RPI ratings is dangerous to say the least, but at this point in the season there are some statistics that can be gleaned to use as benchmarks. These are the current win-loss percentages by category:

Top 10 vs.

Top 10 8-8 50%
11-25 23-5 82%
26-50 33-1 97%
51-100 34-2 94%
101-200 41-0 100%
201+ 27-0 100%

(Note: Two of the sub-25 losses are #10 Rutgers against Purdue and Washington St. The other is #9 Texas against TCU. Rutgers is also the only top 10 team without a win over a top 25 team.)

11-25 vs.

Top 10 5-23 18%
11-25 12-12 50%
26-50 19-11 63%
51-100 52-5 91%
101-200 70-1 98%
200+ 57-1 98%

(Note: Of the 7 losses to sub-50 teams, #20 Villanova is responsible for 3 of them, Creighton, Butler and Penn)

26-50 vs.

Top 10 1-33 3%
11-25 11-19 37%
26-50 31-31 50%
51-100 56-24 70%
101-200 124-9 93%
200+ 105-1 99%

Looking at teams based on these categories (essentially the breakdowns in the nitty gritty report) gives a better sense of the quality of teams, but unfortunately they get skewed by teams that are mischaracterized by the straight RPI. Rutgers and Villanova are both teams that are much higher rated than their results would indicate. Other teams in the top 50 that look overrated based on their records include Oklahoma, West Virginia, Northern Colorado (all in the top 40 with more losses to sub-50 teams than wins over top 50 teams) and #42 Houston (no wins over top 100 teams and 2 losses vs sub-100 teams). Playing, and more importantly beating any of these teams will look better on the report than it should.


ClayK



Joined: 11 Oct 2005
Posts: 11102



Back to top
PostPosted: 01/18/18 10:41 am    ::: Reply Reply with quote

Good stuff ... and as the season progresses, the RPI (and other algorithmic measures) get more and more accurate.



_________________
Oṃ Tāre Tuttāre Ture Svāhā
calbearman76



Joined: 02 Nov 2009
Posts: 5152
Location: Carson City


Back to top
PostPosted: 02/06/18 2:04 am    ::: Reply Reply with quote

As of today there are 5 teams in the top 50 of the RPI that do not have a win over a top 50 team. There are another 4 teams that have only 1 win. On the other side of the ledger there are 8 teams in the top 50 with 1 loss against a team outside the top 100 and 7 more with more than 1 loss. Strangely, 6 of the 9 teams with less than two top 50 wins also have at least 1 sub 100 loss.

Here are all the teams that fit into each category:

RPI----School-----Top 50 -51-100 ---100+
--------------------Record--Losses ----Losses

12 ----Stanford --- 4-7 ---- 0 ---------- 1
13 --- Green Bay - 4-1 ---- 0 ----------1
22-----Buffalo----- 1-2 ---- 0 ----------2
27 - South Florida - 3-3 ---- 1 --------- 1
30 --Oklahoma --- 1-7 ---- 3 ---------- 2
31 -- Marquette --- 3-6 ---- 1 ---------- 1
32 Nor Colorad0 -- 2-0 ---- 2 -----------4
36 ------FGCU ----- 2-1 ---- 2 ---------- 1
37 --California ---- 0-8 ---- 0 ---------- 0
38 ----Syracuse --- 1-6 ---- 1 ---------- 0
39 ---Creighton --- 3-5 ---- 3 ---------- 1
42 -----Gonzaga --- 0-1 ---- 2 ---------- 1
44 ----Houston ---- 0-3 -----0 -----------3
45 --- Princeton ---- 0-2 ---- 1 --------- 1
46 West Kentucky - 1-3 ---- 2 ---------- 0
47 - Cent Florida -- 0-2 ---- 3 ---------- 2
48 --- Duquesne --- 2-2 ---- 0 ---------- 2
50 ------Miami ----- 2-4 ---- 1 --------- 2

As a top 25 team Buffalo's resume is particularly weak. The two top 50 teams (Arizona St and Central Michigan) that beat the Bulls are, at least for now being credited with a top 25 win that is undeserved. Oklahoma is in a similar position (1 top 50 win 2 sub-100 losses, but since they are in the 26-50 group the 7 teams that have beaten them are not being overly benefitted. Although Northern Colorado had two very good wins early (DePaul and LSU) their subsequent losses make them unwarranted of a top 50 ranking. But the only teams that are given false benefit are the teams they beat, who should hace a bad loss. Cal is an interesting case. All of their losses are to top 50 teams, and all are ranked in the current AP top 25, but they no longer have a top 50 win because USC, who they have beaten twice, has fallen to 51. If USC stays out of the top 50 the Bears will be a bubble team. Syracuse has a very similar resume to Cal, but they have one big win over Florida St and one not that bad loss to Virginia Tech.
Seven of the teams from 42-50 make this list. Of them Houston and Central Florida, two teams from the American Athletic Conference stand out. Neither team has beaten atop 50 team (CFla's best win is over #69 UCDavis, Houston's is over #96 George Mason). Central Florida has 2 sub100 losses, Houston has 3. CFla would not be in the top 50 except for their 80-44 loss to UConn, and they will move up close to the top 40 just by losing to them again on Wednesday. South Florida could be a major beneficiary of these ratings as they have already beaten CFla once and play both teams again.


pilight



Joined: 23 Sep 2004
Posts: 66772
Location: Where the action is


Back to top
PostPosted: 02/06/18 8:22 am    ::: Reply Reply with quote

Buffalo's high ranking is puzzling. They don't have a great SOS (#72) and while their record is good, it's no better than a number of other teams.

Quote:
CFla would not be in the top 50 except for their 80-44 loss to UConn, and they will move up close to the top 40 just by losing to them again on Wednesday.


This is the biggest flaw in RPI. Getting blown out by UConn doesn't prove anything about your team. Anyone could do that. RPI would be better if the opponent record (and opponent opponent record) of a team's wins counted more than those of their losses. I would count them twice as heavily. That would push down both the cupcake schedulers like Baylor and teams like Oklahoma that play plenty of good opponents but can't beat them.



_________________
Let us not deceive ourselves. Our educational institutions have proven to be no bastions of democracy.
purduefanatic



Joined: 10 Aug 2011
Posts: 2819
Location: Indiana


Back to top
PostPosted: 02/06/18 11:35 am    ::: Reply Reply with quote

pilight wrote:
Buffalo's high ranking is puzzling. They don't have a great SOS (#72) and while their record is good, it's no better than a number of other teams.

Quote:
CFla would not be in the top 50 except for their 80-44 loss to UConn, and they will move up close to the top 40 just by losing to them again on Wednesday.


This is the biggest flaw in RPI. Getting blown out by UConn doesn't prove anything about your team. Anyone could do that. RPI would be better if the opponent record (and opponent opponent record) of a team's wins counted more than those of their losses. I would count them twice as heavily. That would push down both the cupcake schedulers like Baylor and teams like Oklahoma that play plenty of good opponents but can't beat them.


Yes, I agree that Buffalo is kind of surprising to still be that high. The one that really surprises me is Northern Colorado. Yes, they beat #24 DePaul and #25 LSU in their first 3 games of the year. However, since then they have only played 2 Top 100 teams and lost them both (Q-Pac & Fordham, both on the road). In addition to those 2 losses, they have also lost at home to #112 Colorado St, at #140 Weber St, at #158 North Dakota and at home to #162 Idaho.

I have a hard time seeing how a team that has only played 4 Top 100 teams all season long and only split them can possibly ranked among the Top 32 teams in the nation. That simply makes less than zero sense.


Queenie



Joined: 18 Nov 2004
Posts: 18013
Location: Queens


Back to top
PostPosted: 02/06/18 8:41 pm    ::: Reply Reply with quote

Stupid question: how does opponents' RPI affect a team's RPI? Because if I recall correctly, St. John's has a higher than reasonable RPI because we scheduled aggressively in the non-conference season. Can that be inflating Buffalo?



_________________
All your Rebecca are belong to the Liberty.

(now with spelling variations)
calbearman76



Joined: 02 Nov 2009
Posts: 5152
Location: Carson City


Back to top
PostPosted: 02/07/18 12:48 am    ::: Reply Reply with quote

Actually the biggest inflator of an RPI is playing good teams from bad conferences. Generally teams that play more games from top conferences have relatively lower RPI rankings than teams that play similarly difficult schedules (as measured by Sagarin, Massey or other computer based programs) but fewer games from those conferences. An excellent schedule in terms of inflating an RPI this year would be:

Jackson St (RPI -307) 9-7
F Dickinson (266) 13-8
Morehead St (215) 15-8
Radford (163) 14-7
Robert Morris (148) 15-5
Weber St (140) 14-5
Bethune Cookman (130) 15-4
Stephen F Austin (126) 13-4
Rice (107) 17-3
Navy (84) 17-4
Princeton (45) 14-4
Houston (43) 18-6
Northern Colorado (31) 16-6
Buffalo (22) 17-4

Playing those teams would give you a 72% opponents' winning %. A top 25 team would be able to go at least 13-1 against that group, which (even if you impute a 40% opponents' opponents' win% would put you 3rd or 4th in the RPI. I included the last four teams because that way mot only would the team have a high RPI but they would also have 4 top 50 games, including 1 top 25 game.


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
Page 1 of 1

 
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