RPI Futures 11/16

[bluepenquin]

I have added the # of RPI top 25 wins in the 1st parenthesis. This is part of the RPI adjustments. In addition, it appears that 14 teams will get a scheduling bonus - all 12 teams in the Pac12 + Texas.  Since I don't know how they quantify these bonus adjustments, RPI Futures excludes them.  Not a lot of change from last week, despite the upsets this weekend.

1. Stanford (12) (29-1, 19-1)
2. Texas (9) (24-1, 15-1)
3. Florida State (8) (28-2, 16-2)
4. Washington (9) (28-3, 17-3)
5. Wisconsin (5) (27-3, 18-2)
6. North Carolina (5) (25-3, 16-2)
7. Florida (4) (24-4, 18-0)
8. Penn State (3) (30-3, 18-2)
9. Oregon (2) (22-8, 12-8)
10. Illinois (4) (24-7, 16-4)
11. Colorado State (2) (29-2, 17-1)
12. Nebraska (1) (20-9, 14-6)
13. Kansas (1) (21-9, 9-7)
14. Kentucky (1) (26-5, 15-3)
15. Arizona (3) (23-9, 12-8)
16. Kansas State (3) (23-7, 10-6)
17. Iowa State (4) (17-10, 9-7)
18. BYU (0) (24-5, 15-3)
19. UCLA (3) (20-11, 11-9)
20. Oklahoma (2) (20-10, 10-6)
21. Miami-FL (0) (21-8, 14-4)
22. Marquette (1) (24-7, 14-4)
23. USC (1) (16-14, 9-11)
24. Texas A&M (2) (20-9, 12-6)
25. Long Beach State (0) (26-4, 16-0)
26. Duke (3) (21-8, 13-5)
27. UCF (0) (25-7, 18-2)
28. Alabama (1) (25-7, 13-5)
29. Creighton (2) (24-7, 17-1)
30. Lipscomb (1) (20-7, 13-1)
31. Ohio (0) (24-4, 16-0)
32. Ohio State (2) (21-11, 12-8)
33. Colorado (4) (19-13, 11-9)
34. Loyola Marymount (2) (24-6, 12-6)
35. Illinois State (0) (24-5, 18-0)
36. Arkansas-Little Rock (1) (26-4, 20-0)
37. Western Kentucky (0) (27-5, 15-1)
38. Oregon State (1) (19-12, 9-11)
39. Hawaii (0) (20-7, 12-4)
40. SMU (1) (25-7, 14-6)
41. Arizona State (2) (19-13, 9-11)
42. LSU (2) (18-9, 13-5)
43. Cal State Northridge (0) (19-9, 10-6)
44. Santa Clara (1) (22-9, 12-6)
45. Utah (3) (17-14, 7-13)
46. Pittsburgh (0) (24-7, 12-6)
47. Purdue (0) (23-9, 13-7)
48. San Diego (1) (18-12, 11-7)
49. Temple (0) (25-7, 16-4)
50. Pacific (0) (23-8, 12-6)
51. Xavier (0) (19-11, 12-6)
52. UNLV (1) (26-7, 14-4)
53. Tulsa (0) (21-10, 13-7)
54. Michigan State (0) (17-14, 10-10)
55. Baylor (2) (15-16, 5-11)
56. Michigan (1) (13-17, 8-12)
57. Seton Hall (1) (26-7, 14-4)
58. Virginia Tech (1) (16-15, 9-9)
59. George Washington (0) (21-7, 11-3)
60. Minnesota (1) (18-13, 8-12)
61. Harvard (0) (19-4, 12-2)
62. Wyoming (0) (23-8, 12-6)
63. Northern Colorado (0) (20-9, 12-4)
64. Ole Miss (0) (22-10, 8-10)
65. Virginia (0) (17-14, 10-8)
66. New Mexico (0) (20-11, 12-6)
67. Northwestern (1) (17-14, 7-13)
68. Arkansas State (0) (20-9, 15-5)
69. Wichita State (0) (19-9, 13-5)
70. Dayton (0) (26-6, 13-1)
71. Texas-San Antonio (1) (19-8, 15-1)
72. VCU (0) (16-12, 9-5)
73. LIU Brooklyn (0) (23-6, 14-0)
74. UMKC (0) (24-5, 12-2)
75. Towson (0) (27-4, 13-3)
76. American (0) (24-6, 15-1)
77. Coastal Carolina (0) (22-6, 11-3)
78. Furman (0) (21-6, 14-2)
79. Florida Gulf Coast (0) (19-10, 11-3)
80. Yale (0) (17-7, 12-2)
81. College of Charleston (0) (22-9, 14-2)
82. Butler (1) (20-10, 12-6)
83. Texas-Arlington (0) (24-8, 14-6)
84. Arkansas (1) (15-16, 9-9)
85. Louisville (0) (16-14, 8-10)
86. West Virginia (0) (15-15, 5-11)
87. Memphis (0) (23-11, 13-7)
88. Northern Iowa (0) (20-10, 14-4)
89. Missouri State (1) (19-11, 12-6)
90. Gonzaga (0) (17-12, 9-9)
91. Wisconsin-Milwaukee (0) (16-13, 10-4)
92. Iowa (0) (15-16, 7-13)
93. NC State (0) (16-14, 5-13)
94. Denver (0) (24-7, 13-3)
95. Rice (1) (21-9, 11-5)
96. South Dakota (0) (18-11, 12-4)
97. UNC Wilmington (0) (20-10, 10-6)
98. Northern Illinois (0) (21-9, 14-2)
99. Hofstra (0) (26-5, 13-3)
100. Radford (0) (22-7, 12-2)
101. Missouri (0) (17-16, 8-10)
102. Clemson (0) (19-13, 7-11)
103. Northern Kentucky (0) (18-10, 7-7)
104. TCU (4) (16-16, 5-11)
105. Oakland (0) (20-9, 12-2)
106. North Texas (0) (21-9, 11-5)
107. Texas State (0) (21-11, 16-4)
108. Appalachian State (0) (20-10, 10-10)
109. Valparaiso (0) (24-7, 9-5)
110. Texas Tech (1) (16-12, 4-12)
111. South Carolina (0) (17-14, 7-11)
112. Murray State (0) (23-5, 14-2)
113. Auburn (0) (13-18, 5-13)
114. Georgia (0) (15-16, 6-12)
115. Indiana (0) (16-16, 7-13)
116. Washington State (0) (11-21, 2-18)
117. New Mexico State (0) (17-10, 10-4)
118. Saint Louis (0) (20-9, 9-5)
119. Cal (1) (10-20, 2-18)
120. Miami-OH (0) (19-9, 12-4)
121. Idaho State (0) (21-8, 15-1)
122. Central Arkansas (0) (18-11, 13-3)
123. Louisiana-Lafayette (0) (20-12, 9-11)
124. Florida Atlantic (0) (18-11, 9-7)
125. Boston College (0) (12-17, 7-11)

[trollhunter]

Thanks Blue.

Some top teams may be getting penalties for bad losses depending on final rpi of opponents they lost to.

Those with losses to rpi 100+ teams are:

Kansas
Iowa State x2
USC
LSU
Oklahoma
Miami (FL)
Lipscomb x2
Cal Northridge x2
Santa Clara x2
SMU
San Diego x2
Utah
Purdue
Illinois State
Virginia x4
Tulsa x3
Xavier
Temple x3
Minnesota
Wyoming x3

[trollhunter]

Blue,

While you are tracking top 50 wins, can you also see who has scheduled the majority of their matches versus top 50 teams? By comparing final rpi with adjusted rpi of affected teams, we should be able to derive the actual numerical amount of the Bonus and Penalty.


Factor IV: Bonus: Scheduling majority of matches against teams ranked 1-50 in the RPI and for wins against teams ranked 1-25 in RPI.

Penalty: Scheduling majority of matches against teams 167 and above in RPI and for losses against teams 167 and above and non-Division I teams.

[BeachbytheBay]

Nov 16, 2014 20:40:53 GMT -5 trollhunter said:

Blue,

While you are tracking top 50 wins, can you also see who has scheduled the majority of their matches versus top 50 teams? By comparing final rpi with adjusted rpi of affected teams, we should be able to derive the actual numerical amount of the Bonus and Penalty.


Factor IV: Bonus: Scheduling majority of matches against teams ranked 1-50 in the RPI and for wins against teams ranked 1-25 in RPI.

Penalty: Scheduling majority of matches against teams 167 and above in RPI and for losses against teams 167 and above and non-Division I teams.

using RPI for ranking/ordering is one thing, hopefully the committee uses their discretion for evaluation of 'quality' Wins and Losses with something other than RPI (or uses both RPI and another set of data) supplemented by Pablo or Massey data - RPI breaks down in evaluating what is really a top 100 team

hopefully the committee is circumspect on determining what a bad loss is, as many teams out west will be over 100 RPI that are actually top 100 teams by a more meaningful measure (Pablo/Massey) of a team's strength (St. Marys, USF, Cal, Wash State, UCSB, Davis, UCIrvine) - so while losses to those teams might seem 'bad' if one uses RPI with tunnel vision, they in fact aren't that bad of a loss - the assignment of what is a bad loss (especially on conference road matches) should always be done with a healthy dose of skepticsm.


a prime example is LB State - going by RPI, LB has only had 8 top 50 matches and is 4-4 in those, with no other top 100 RPI matches - so LB is 4-4 top 50 RPI & 4-4 top 100 RPI, but going by Massey, LB is 4-4 top 50, and 12-4 top 100, which illustrates the skewing of perception and quality of wins/losses which happens with a 'strict' use of RPI and no other data - not saying RPI should not be used only pointing out RPI should not be used in a vacuum

[gnu2vball]

Question: Are the RPI Top 25 wins with respect to the RPI Top 25 when the victory took place or with respect to the RPI Top 25 now?

[mikegarrison]

Nov 16, 2014 21:31:50 GMT -5 gnu2vball said:

Question: Are the RPI Top 25 wins with respect to the RPI Top 25 when the victory took place or with respect to the RPI Top 25 now?

Neither. They will be with respect to RPI in the last week of the season.

[gnu2vball]

Aha! Thanks.

[bluepenquin]

Nov 16, 2014 20:40:53 GMT -5 trollhunter said:

Blue,

While you are tracking top 50 wins, can you also see who has scheduled the majority of their matches versus top 50 teams? By comparing final rpi with adjusted rpi of affected teams, we should be able to derive the actual numerical amount of the Bonus and Penalty.


Factor IV: Bonus: Scheduling majority of matches against teams ranked 1-50 in the RPI and for wins against teams ranked 1-25 in RPI.

Penalty: Scheduling majority of matches against teams 167 and above in RPI and for losses against teams 167 and above and non-Division I teams.

I don't have an easy way to do this - just saw this nugget for the 1st time this week. 

As mentioned at the top of this post - I believe that Scheduling Bonus will only go to all 12 Pac 12 teams plus Texas.  No other team will qualify.  None of the teams in the top 50 from the top conferences will get the Scheduling Penalty.  The one's I manually checked from smaller conferences were good for avoiding this penalty (Lipscomb, Western Kentucky, Ohio, ect).  I think a team like Harvard may get this penalty, but I didn't bother to check.

I do plan on adding losses to teams with an RPI of 167 or worse - such as LSU today.  I  don't think there will be many teams in the top 50(?), but I didn't look too close.  I probably will not go past the top 75-100 teams if I continue to use a manual process.

I am not really in a position to figure out what the actual bonus and penalties are.  To do this, I would need to compare week by week - but my program is only set up to calculate the season ending RPI.  I may try to back into this at the end of the season - but I figure that RK calculated this by comparing several weeks of data instead of just the final RPI.

How the heck do they program the wins in the top 25?  I did it by not making adjustments (which I wouldn't know how to do anyway).  I have Long Beach at #25 before doing adjustments.  Then after adjustments, what if Duke passes Long Beach?  But then we would have to re do the wins against the top 25 that can produce a different #25.  Seems like this would be and programming loop?

[trollhunter]

Blue,

Thanks for looking at Scheduling Bonus already, I missed that in your first post.

There will be a few bubble teams with penalties for 167+ losses for us to keep an eye on.

You can look at ncaastats.figstats.net/volleyball-rpi.cgi or RKPI to get the basic RPI numbers. Then by looking at what teams are listed in different order in NCAA RPI, you can see how much the basic number must have changed to have teams in different order.

If you look at rpiarchive.ncaa.org it appears that they calculate an "original" rpi for each week, then create an "adjusted" rpi list by applying the bonuses and penalties to the "original" rpi list. So the top 25 win bonus looks at "original" rpi list, and the top 50 scheduling bonus looks at the same "original" list.

We could probably look at some bottom PAC12 teams with no top 25 wins, but that get Schedule Bonus, and see how far they are moved up on NCAA RPI list to see approximately what that bonus is.

And then find a team with just one top 25 win, but without schedule bonus (or any penalties) and compare their basic RPI ranking with NCAA list to find the range of the top25 win bonus.

Same for penalties. Does that make sense? Let's pick a couple teams and try it this week.

Hmmm. Although we will have to find a team that is bracketed by teams that aren't receiving bonus or penalty also.

[trollhunter]

Ok once NCAA rpi list comes out this week, we need to find which of the following teams (single top 25 win) is transposed on basic rpi list with a team that is receiving no bonus or penalty. Then we can do simple subtraction to get approx. bonus for a top 25 win.

Single Top 25 win list (with no other penalty/bonus)

Marquette
Duke
Lipscomb
Santa Clara
Loyola Marymount
Ohio
SMU
Illinois State
UNLV
Rice

[trollhunter]

If Washington State ncaa rpi rank is transposed with a team receiving no penalty/bonus, we can also derive the approximate range of the Schedule Bonus. Woooo.

[ay2013]

Well a lot of the bonus can be extrapolated from the wins versus top 25 teams, which you already point out.

The primary criteria for selection is rather clear:

(In No Particular Order)

Won-lost record
Strength of schedule
Eligibility and availability of student-athletes for NCAA championships
Rating Percentage Index
Head-to-head competition
Results versus common opponents
Significant wins and losses

Given that, I see this bracket

LOUISVILLE
1 Stanford
16 UCLA/Nebraska/Kansas/Kentucky
9 Illinois
8 Florida

MINNEAPOLIS
5 Wisconsin
12 Kansas State
13 Arizona
4 FSU

SEATTLE
3 Washington
14 UCLA/Nebraska/Kansas/Kentucky*
11 Oregon
6 North Carolina

AMES
7 Penn State
10 Colorado State
15 Iowa State*
2 Texas

The asterisks is to note that I think that Iowa State should have the #14 seed over the others, but Iowa State is a regional host and can't be in a regional with Washington.

The 15/16 spot could go to a number of teams, it depends on what the Committee values most. UCLA has the most number of significant wins among the three, Nebraska would have the highest overall SOS among the three, Kentucky has the best Win/Loss ratio among the three and they are really close to Louisville (ugh), Kansas because they are in the RPI range.

I just think that given the primary criteria, if the committee actually followed it, we really shouldn't be seeing anything too different (given this data set). Perhaps a spot or two shift, such as Washington and FSU (3/4) switching, or Wisconsin and North Carolina (5/6) switching or PSU and Florida switching (7/8). Given the primary criteria I think the top 8 seeds could easily fall with RPI and we get (1)Stanford/(8)PSU, (2)Texas/(7)Florida, (3)FSU/(6)North Carolina, (4)Washington/(5)Wisconsin. This of course would be DISASTROUS bracket, but the primary criteria can easily justify it, and it's not like the committee didn't give us something similar last year (Stanford/PSU regional, All Pac-12 Regional, All SEC Regional)

[ay2013]

Also, Utah, now with 3 top 25 wins, has separated themselves from all other bubble teams. Assuming USC finds another win, I think it's a pretty good bet that the Pac-12 will get a record 10 teams into the tournament.

As for the Big 10, I think that Purdue is still probably in, but it will be close....if some of those higher RPI teams from one bid leagues lose their conference tournament, Purdue will find themselves out. Big 10 will probably get 6 in, but I would not be surprised in the slightest if they only got 5.

[volleyfan24]

Nov 17, 2014 1:48:51 GMT -5 ay2013 said:

Also, Utah, now with 3 top 25 wins, has separated themselves from all other bubble teams. Assuming USC finds another win, I think it's a pretty good bet that the Pac-12 will get a record 10 teams into the tournament.

As for the Big 10, I think that Purdue is still probably in, but it will be close....if some of those higher RPI teams from one bid leagues lose their conference tournament, Purdue will find themselves out. Big 10 will probably get 6 in, but I would not be surprised in the slightest if they only got 5.

I love your bracket. I really hope it pans out this way. Ten is like getting the entire conference in yikes really speaks to the depth and how good the teams are in the conference I think this will serve all teams well come tournament time. 

[ay2013]

Nov 17, 2014 3:33:24 GMT -5 volleyfan24 said:

Nov 17, 2014 1:48:51 GMT -5 ay2013 said:

Also, Utah, now with 3 top 25 wins, has separated themselves from all other bubble teams. Assuming USC finds another win, I think it's a pretty good bet that the Pac-12 will get a record 10 teams into the tournament.

As for the Big 10, I think that Purdue is still probably in, but it will be close....if some of those higher RPI teams from one bid leagues lose their conference tournament, Purdue will find themselves out. Big 10 will probably get 6 in, but I would not be surprised in the slightest if they only got 5.

I love your bracket. I really hope it pans out this way. Ten is like getting the entire conference in yikes really speaks to the depth and how good the teams are in the conference I think this will serve all teams well come tournament time. 

I think the more interesting question is where these non-seeded Pac-12 teams will be placed. Colorado is almost surely going to CSU. It remains to be seen what subregional USC, UCLA (assuming they don't get a seed), Utah, Arizona State, and Oregon State will be in, as these teams will have to travel so their subregional could be anywhere. If these teams can land in the Kansas State/Kentucky/Iowa State subregional, I think all the Pac-12 teams can beat those teams. Will they is a different story, but I like their chances. If there is a Pac-12 team in the Kansas State, Kentucky, and Iowa State subregional, in addition to Colorado State (which we know Colorado will be in), and given 5 likely seeds, I'm going to go on record and predict that not only will the Pac-12 set a record for most bids to the tournament, but they will also set a record with 8 teams to make the sweet 16.