Toggle sidebar

RPI up to ....

R

RU MAN

Heisman
Oct 29, 2001
23,610
10,193
113
Our ladies are doing really well. Not only should we be receiving top 25 votes, there is no reason why this iteration of women shouldn't be ranked in the top 25. They just need to keep focused and keep winning especially when we begin B1G play.
 
  • Like
Reactions: Bon1674
R

RUich

All-Conference
Aug 2, 2001
13,552
4,003
0
I kind of like it that they apparently are still "unknown" to the pollsters. It just seems that Rutgers teams have had some problems once they get a number attached to them.
 
  • Like
Reactions: DHajekRC84
R

RUClassof67

All-Conference
Jan 2, 2004
9,603
1,722
113
Update: Going into Sunday’s game vs FDU, we are at 14 with an SOS of 28. Unfortunately FDU is way up there and will impact both.
 
B

BeKnighted_rivals

All-Conference
Jan 15, 2003
16,269
1,360
0
RPI is not a great predictor, and even less of a great predictor this early in the season. That said, relatively small differences in the actual RPI can still lead to big differences in RPI ranking.
 
ScarletDave

ScarletDave

Heisman
Oct 7, 2010
34,595
15,348
85
RPI means nothing if your name is the Scarlet Knights ... our women's soccer team had an RPI of 11 on selection sunday this year (3-seed) and were given the 7-seed (#25-28) ... never count on anything. Let's go win the Big Ten !
 
  • Like
Reactions: ScarletteK80
K

Knight Time Fan_rivals

Senior
Jul 26, 2001
3,543
688
0
cubuffsdoug said:
Can someone explain to me all of this. The way I read it (example) is a team like Purdue is likely to beat Rutgers (again) this year. Purdue is not a good team this year, but RPI says otherwise.
Click to expand...
Like BeK says, RPI is not a great predictor, nor is it intended to be.

As a measure, the first 25% is your record - obvious factor
The second half - yes half - is your opponents records. That's why it is beneficial to play teams with good records. There is a benefit to playing the best team in a weak conference, if they are the sort that dominates their conference.
The last quarter is you opponent's opponent's record - hence, there is some benefit realized from playing a mediocre team from a strong conference - they probably have a respectable over-all record from their OOC play and then in-conference you benefit from the winning records of their conference mates.

Of course, then the NCAA applies "secret" factors at selection time, relating, primarily, to home / away / neutral contests. including points differential is considered a taboo, for obvious reasons.

That said, services that do power rankings - such as Massey - are better predictors. I don't follow any of the numerical rankings, except RPI as a predictor of NCAA seeding - where there is certainly some correlation. In my personal opinion - most teams are who they look like they are if you glance at their record and who they won / lost to, without factoring in a bunch of other stuff.
 
B

BeKnighted_rivals

All-Conference
Jan 15, 2003
16,269
1,360
0
To add to KTF's post, one funny thing about RPI is that it looks like opponents' record is more important than W/L, but W/L actually is the most important component. Basically, there's a lot more variation in W/L than in opponents' record, so even though it's just 25%, you can get anything from 1.00 (perfect record) to 0.00 (winless) in that category, while by the end of the year, opponents' records tend to be in a relatively narrow range.

For instance, in 2010 (hey, it's what I could find easily), the top SOS was .6454, while the bottom SOS was .3918, with a difference between top and bottom of .2536, while the top W/L was 33-0 and the worst was 0-33. (SOS actually is composed 2/3 of opponents' record and 1/3 of opponents' opponents' record, and account for 3/4 of the total RPI, but it's still useful for this purpose.) So the best you could get from W/L was 0.25 and the worst was 0.00, a difference of 0.25, while the best you could get from SOS was .4840 and the worst you could get was .2938, a difference of .1902. That means that W/L was a more important component in RPI than opponents' record and opponents' opponents' record combined.
 
  • Like
Reactions: Knight Time Fan_rivals
R

RU-Choppin-Ohio

Heisman
Jul 31, 2011
32,978
37,755
113
RU MAN said:
Our ladies are doing really well. Not only should we be receiving top 25 votes, there is no reason why this iteration of women shouldn't be ranked in the top 25. They just need to keep focused and keep winning especially when we begin B1G play.
Click to expand...

Based on the 1 top 25 vote received it looks like they have Rutgers ranked 5th in the Big 10
 
D

dmd78

Senior
Jul 25, 2001
41,275
482
0
sammybb said:
Everything above about not taking RPI too seriously is correct, but that being said...

RU's RPI is now up to #12, with a #20 SOS. http://realtimerpi.com/college_Women_basketball_rpi.html

Houston is currently ranked #32, so that would be another high quality win (as of right now).
Click to expand...

That site projects a 19-11 season for us. But while we have 2 quality wins, none of our next few opponents do, so there are definitely some winnable games. I’ll be surprised if we don’t get at least 20 wins before the regular season ends.
 
D

DHajekRC84

Heisman
Aug 9, 2001
30,709
19,818
0
this stuff is weird...# 8 Oregon 8-2, 0-2 top 25, 1-2 top 50 SOS 106

#18 RU 11-2 0-1 top 25, 2-1 top 50 SOS 34
 
S

ScarletteK80

All-Conference
Mar 4, 2003
15,727
4,401
58
1984 said:
this stuff is weird...# 8 Oregon 8-2, 0-2 top 25, 1-2 top 50 SOS 106

#18 RU 11-2 0-1 top 25, 2-1 top 50 SOS 34
Click to expand...
Like Dave said up the thread a bit..."RPI means nothing if your name is the Scarlet Knights." Sad, and paranoid, but true IMHO. That's why CVS, et al know why we have to just win, win, and win.
 
K

Knight Time Fan_rivals

Senior
Jul 26, 2001
3,543
688
0
1984 said:
this stuff is weird...# 8 Oregon 8-2, 0-2 top 25, 1-2 top 50 SOS 106

#18 RU 11-2 0-1 top 25, 2-1 top 50 SOS 34
Click to expand...
Right now, as BeK's post indicated, there is a significant difference in factor 1 (record) - .85 vs. .80

Also, they have played a few more (serious) Little Sisters than Rutgers.

That said, it would take a detail look to figure it out, something I can't convince myself I want to do.

As BeK also said, a lot of things change the larger the sample size. It is, in terms of RPI and SOS, still early.
 
D

DHajekRC84

Heisman
Aug 9, 2001
30,709
19,818
0
Knight Time Fan said:
Right now, as BeK's post indicated, there is a significant difference in factor 1 (record) - .85 vs. .80

Also, they have played a few more (serious) Little Sisters than Rutgers.

That said, it would take a detail look to figure it out, something I can't convince myself I want to do.

As BeK also said, a lot of things change the larger the sample size. It is, in terms of RPI and SOS, still early.
Click to expand...

oh, to all the replies..I get it. It is not a perfect science and early, etc. Nothing matters but winning and seeing where you are at the end. Then my favorite are those blind resume match-ups. .

I wouldn't argue we should be #8 ...I more was wondering why Oregon would be.
 
A

Abro1975

Heisman
Nov 21, 2009
24,499
12,979
0
RPI is an algorithm based ranking.

USA Today, ESPN etc, are perception based, which at this early part of the year, with mostly OOC schedules, is based on little empirical data.
 
E

ecojew

All-Conference
Feb 1, 2006
9,767
2,271
0
dmd78 said:
Post-Houston game, they now have us at RPI 9, SOS 12
Click to expand...

Does the RPI factor in margin of victory? RU has had relatively few close wins - most being in double figures. And the two losses were by a total of just 13 points, IIRC.

Thanks.
 
B

BeKnighted_rivals

All-Conference
Jan 15, 2003
16,269
1,360
0
RPI is purely based on wins and losses. Not considering margin of victory is one of RPI's weaknesses. (On the other hand, some of the systems that do count margin of victory have penalized RU in the past, when pace of game meant that the team did not run up big margins compared to teams that played faster.)
 
E

ea_1

Junior
Dec 15, 2013
1,068
359
0
https://www.masseyratings.com/cbw/ncaa-d1/ratings
Rutgers #25, SOS 32 (to date), SSF 31 (includes future opponents), projected season 21-9. Off #100, Def #12
http://www.rpiratings.com/womrate.php
Rutgers #19 SOS 36 and #10 weighted ranking for recent play (a trend indicator)

http://realtimerpi.com/college_Women_basketball_rpi.html
Rutgers #8 SOS 13

http://realtimerpi.com/ncaab/college_Women_basketball_power_rankings.html
Rutgers #84 SOS 9

Massey and Sagarin have been pretty consistent with RU gradually rising with each win. The Massey season projection has been stable for weeks, but if you use the game-by-game projections, RU would be 10-5-1 for the conference season. Sagarin, the second ranking, is score-based. I have to wonder if a data-entry error occurred in the RealTime power rankings. How on earth can a 12-2 team with a top-10 schedule strength be ranked 84?

My final thought about these rankings and ratings is that I have believed since early in the season if RU could break into the top 100 in offense and stay there, the team would be fine (because the defense is so good). This is the first week RU made the top 100 in the Massey ratings this season. I suspect that is because the terrible mismatches of the guaranteed games have lessened as schedules have tightened. Not sure if BeKnighted has the ranked teams' scores in a spreadsheet--I would be interested in seeing the median and mean margins of victory as the weeks progressed during OOC play.
 
You must log in or register to reply here.
Top Bottom