RPI for 2008

I was kind of grousing to myself about possibly losing some ground beacause the conference is off a bit this year, but your analysis gives me great hope.

UPSF

Thanks for all your analysis. I'd wager that there is no one in the country who understands this better than you do. Most folks have no idea about RPI. We do, for two reasons. One is that we're nuts and somewhat crazy about this team, this game and this program. And the second is that we've gotten seriously disadvantaged in the past as a result of not having a top 4 RPI score. So we really care about this in a way that others don't need to. Most other top teams don't have to worry that if they finish #5 or #6 in the country that they will be packed off to Omaha to play the first two games.

I'll believe it when i see it when UNC is #6 and has to go to Atlanta or Birmingham.

This is a copy of my post on the RPI thread on the BigSoccer website:

Next Monday, October 6, is the first day the NCAA is supposed to publish its first RPI rankings for the 2008 season. These will be the RPI rankings including bonus and penalty adjustments for high quality wins/ties and poor losses/ties. To access the NCAA RPI rankings, use the following link: http://www.ncaa.com/soccer-womens/default.aspx?id=396. This will take you to the NCAA Women's Soccer page. In the upper left, under the heading "Women's Soccer," click on "RPI." This should take you to the RPI rankings. Right now, last year's end-of-season rankings are posted, so check the date on the ranking list to be sure you have this year's rankings. Although the NCAA says it will post the rankings on Monday, in past years their postings have been late, so don't be surprised if you have to wait a day or two past Monday.

Since the NCAA is supposed to post its RPI rankings on Monday, I won't be posting mine. Instead, I'll be doing two tasks. First, if the NCAA does what it did last year, it will be posting the teams' win-loss-tie and away-neutral-home records together with their rankings. I'll be comparing those data with mine and then, if there are differences, figuring out any data errors I have and also any data errors the NCAA has so I can correct my errors or advise them of theirs. (They're responsive to well-documented corrections.) Second, once I've gone through the data-error-identification process, I'll be comparing my rankings with theirs to see if I can get my bonus and penalty amounts closer to theirs. (Their bonus/penalty amounts are secret, so I have to go through this process to try to figure out what their amounts are. There are 24 different amounts, so it's not easy.)

The NCAA publishes only its rankings, and not the rating numbers from which it derives the rankings. So, towards the end of the week, I'll hopefully be able to post the rating numbers so people can see how they arrived at the rankings.

Early next week, the NCAA Division 1 women's soccer regional advisory committees will be having their first meetings, to review the RPI data and do preliminary rankings of teams within their regions. On Thursday, the Women's Soccer Committee will be meeting to review the RPI data and the regional advisory committees' rankings. These meetings represent the beginning of the process that ultimately will lead to ultimate at large selections and seeding of teams at the end of the pre-Tournament season.

If you want to follow this process yourself, here is what I suggest you do. It's in steps. (1) Download the NCAA's RPI rankings. (2) Identify the top 10-15 teams from each of the NCAA's six Division 1 Women's Soccer regions. I think I'm going to do the top 12, myself. Initially rank each region's teams in the order of their RPI rankings. (The regions to which the NCAA assigns teams are available from the NCAA.org website. Unfortunately, I can't access it right now, so I can't provide a link.) (3) For each region, look at the results so far of each team you've identified. Note any head-to-head results between teams on your list. (You can use the NSCAA/Adidas data source, although I should warn that its data are not 100% perfect.) (4) In addition, for each region note any results against common opponents of teams on your list. (5) For each region, after looking at the initial RPI-based rankings, consider whether head-to-head results or results against common opponents would justify a deviation from the RPI-based rankings. Based on this consideration, and on no other factors, adopt a region ranking list. (There are a couple of other criteria that will be considered at the end of the season, but they are time-based, so you won't be able to consider them now.) (6) When you have completed this process, you will have matched the work of the six regional advisory committees. (7) Now, you have to merge the six region ranking lists. The criteria are the same: RPI rankings, head-to-head results, and results against common opponents. I'm not sure exactly how to go about this. The difficulties will come if a region's list does not follow the RPI rankings. This is where the Women's Soccer Committee really does its work. You'll have to figure out how to do a merger in those situations.

I'm hoping some of you will go through this process, at least for a region or two and possibly for all six regions and for the national merger, and will post your results. I think that will help all of us understand the process better. My guess is that if enough of us work on this, come the end of the season we will be able to come very close to guessing in advance what the Women's Soccer Committee will do for the Tournament. Should be fun, if nothing else!

Albyn Jones' SoccerRatings now are available covering games through September 28. Use the following link to access them: http://soccerratings.com/index.php?title=Women%27s_NCAA_Division_I_Soccer_Ratings.

Well UCLA has beaten our conference opponents 1-0 Pepperdine and 3-0 USD. Does this make those games a do or die situation?

Auto Pilot wrote:Well UCLA has beaten our conference opponents 1-0 Pepperdine and 3-0 USD. Does this make those games a do or die situation?

Hard to say at this point. The three main NCAA criteria are RPI, head-to-head results against teams under consideration, and results of teams under consideration against common opponents. Let's assume we and UCLA win out. First, the NCAA will look at the RPI. Let's also assume that we have a higher RPI than UCLA. Then, in the NCAA's initial ranking, we'll be ahead of UCLA in the seeding process. Then, the NCAA will look at head-to-head results. Clearly, UCLA is ahead of us there. At the same time, they'll look at results against common opponents, in this case Pepperdine and USD. One of the things that is not clear is whether the NCAA considers only who won and lost or also the score differential. If the NCAA considers only who won or lost, then both teams would have beaten Pepperdine and USD, so results against common opponents would not be a factor. If the NCAA considers score differential, then it might matter what the Pilots' scores were against these teams. The NCAA staff has explicitly rejected the consideration of score differential in setting up the RPI formula, saying it never will consider score differential. It seems logical that would carry over to considering results against common opponents, but I don't know whether the NCAA is committed to being logical on this. If the NCAA doesn't consider score differential, then it would have to compare the size of the RPI differential between the Pilots and UCLA to the fact that the Pilots lost to UCLA.

If the NCAA feels unsure after considering these three criteria, then it will consider both results over the last eight games and results against teams in the Tournament bracket (other than conference champion automatic qualifiers ranked below #75 in the RPI). The last eight games criteria probably would not be helpful. So, it could come down to comparing the Pilots' and UCLA's results against other teams in the bracket.

Does that answer your question?

Yes and I am beginning to think the NCAA is more secretive than the former KGB. I am glad we have our own intelligence.

The Pilots' opponents' (both already played and to-be-played) for the weekend, which will contribute to Element 2 of the Pilots' RPI (opponents' winning percentage against teams other than the Pilots) were 8-3-3 on Sunday and 9-3-1 on Friday/Saturday, for totals of 17-6-4. Add to that UW's record of 9-1-0 (excluding the result of Friday's game, since it doesn't count for purposes of the Pilots' RPI), and we get 26-7-4 being added to the Pilots' Element 2 numbers. This is not quite right, since the RPI computes each opponent's RPI and then averages all those RPIs to get Element 3. Still, you get the idea: this was a good weekend for Element 2 of the Pilots' RPI. (It's a little late, so if that isn't coherent, sorry.)

I could have put this on the new "Reflections and Preparation" thread, but it fits here perhaps a little better.

I've been thinking over the weekend about the dynamics of the RPI, particularly in relation to conference play. Here are some thought (which may be more than any of you want to know):

1. Obviously, the Pilots' focus has to be on winning every conference game. This will give them as high a winning percentage (RPI Element 1) as they can achieve over the remaining games. That, in turn, will give the Pilots the best chance at one of the four #1 seeds and the assurance of home games for rounds 1 and 2 (and 3 and 4) of the tournament. This is the part of the process that is in their hands.

2. It is when we get to the opponents' strength of schedule (RPI Element 2) that the RPI dynamics get interesting. As I believe Stonehouse pointed out on another thread, once teams get to conference play, part of the dynamic is that there is a drag on teams' strength of schedule towards .5000. This is because we know that a team is going to play each other team in the conference (or, at least, in that teams' division in conferences that have two divisions). If conference teams A and B play each other, one is going to win and one is going to lose; or both are going to tie. In the win-loss scenario, the winning team's winning percentage will go up, but the losing team's winning percentage will go down. In the tie scenario, each team will be adding to its winning percentage a game that is treated as a .5000 game (a tie counts as half a win and half a loss). Roughly speaking this means that whether the game is a win/loss or a tie/tie, for each conference team not involved in that game, the game will be adding two games with a net value of .5000 to the team's opponents' winning percentage number. (I say "roughly speaking" because it's a little more complicated than that. Element 2 is the average of a team's opponents' winning percentages. Exactly how an intra-conference game affects the two teams' winning percentages depends on what each of their winning percentages is going into the game and also on how many games each has played. In general, however, the tendency will be to pull the average winning percentage towards .5000.)

3. However, the conferences are different sizes. The WCC has eight teams, so each team plays seven conference games. The Big 12, on the other hand, has eleven teams, with each team playing ten conference games. In addition, the Big 12 has an end-of-season tournament, which will add games for those teams that qualify for the tournament. (Some conferences have two divisions, with teams playing all other teams in their division plus some teams in the other division; plus a conference tournament.) For a conference in which teams play a large number of conference games, there will be more games for those teams dragging their RPIs towards .5000.

4. On the other hand, although there is the "dragging" effect, there also is another effect on strength of schedule at work once conference play starts. This is the effect of the non-conference games that teams already have played. So, if a team is in a conference whose teams on average have good records in their non-conference games, these non-conference records will have an upward pull on the team's RPI. Similarly, if a team is in a conference whose teams on average have poor records in their non-conference games, these records will have a downward pull on the team's RPI. In addition, for a team that advances through a conference's end-of-season tournament, its opponents as the team progresses through the tournament typically will be the ones that have had success in their non-conference games and thus have relatively high winning percentages, which will have an upward pull on the team's RPI.

5. In addition, when evaluating the different conferences' strengths for purposes of understanding how conference play will affect a team's RPI, you can't really look at the conferences' average RPIs. Using the WCC as an example, if we want to look at how conference games will affect Portland's RPI, we have to take Portland's winning percentage out of the conference numbers -- because Portland won't be playing itself, but rather will be playing only the other seven teams. So, when looking to see how conference play will affect the strength of schedule RPI elements of two teams from two different conferences, what you have to look at is the average winning percentages of the teams in the two conferences (or the two conferences' average RPIs as a surrogate) after first deleting the winning percentages (or RPIs) of the two teams you're comparing. Assuming you're comparing teams as potential #1 seeds, this means that the average winning percentages (or RPIs) of the conferences (after deleting the records of the teams you're comparing) are going to be lower than the average winning percentages (or RPIs) for the entire conferences. In this particular circumstance, if a team is in a smaller conference, then if you delete its winning percentage (or RPI) from the calculation, you will get a bigger proportional drop in the remaining teams' average winning percentage (or RPI) than if the team were from a larger conference.

6. How the dynamics of all this work out, in terms of whether it's better to be in a small conference or a large conference and in terms of whether its better to have a conference tournament or not to have one, is not clear to me. How it works out may vary from situation to situation and from year to year.

Here are the RPIs (as adjusted by "estimated" bonus/penalty amounts based on my best guesses as to what these secret amounts are) for the top 55 RPI-ranked teams for games through October 5. FYI, the gap between Portland's RPI and the next team (North Carolina) is very large for RPI purposes. Last year, the NCAA's at large selections came from the top 47 RPI-ranked teams. The top 16 RPI-ranked teams were the 16 seeded teams (close to, but not exactly in accord with, the teams' RPI rankings), with the top 4 RPI-ranked teams receiving #1 seeds.

Portland U 0.7493
North Carolina U 0.7156
Notre Dame 0.6995
Stanford 0.6991
Texas A&M 0.6979
Duke 0.6779
Boston College 0.6707
Florida U 0.6696
Florida State 0.6692
Rutgers 0.6671
UCLA 0.6584
Illinois U 0.6575
Oklahoma State 0.6569
Texas U 0.6543
Virginia U 0.6527
Colorado U 0.6499
USC 0.6359
Washington U 0.6349
Penn State 0.6345
California U 0.6343
Missouri U 0.6338
Wake Forest 0.6225
Long Beach State 0.6225
UCF 0.6202
TCU 0.6197
South Carolina U 0.6183
Princeton 0.6130
Georgetown 0.6102
Old Dominion 0.6100
James Madison 0.6076
William and Mary 0.6053
Wisconsin Milwaukee 0.6048
Virginia Tech 0.6041
Arizona State 0.6029
Dayton 0.6027
Colorado College 0.6013
Michigan State 0.5998
Richmond 0.5995
Rice 0.5994
Ohio State 0.5992
Denver 0.5991
Kansas U 0.5980
West Virginia U 0.5978
Villanova 0.5938
Purdue 0.5928
San Diego U 0.5897
East Carolina 0.5882
Brown 0.5858
Georgia U 0.5858
Minnesota U 0.5852
Vanderbilt 0.5847
BYU 0.5834
Loyola Marymount 0.5814
Oregon U 0.5812
Saint Louis 0.5803

UPSF

The margin for UP over the 5th team on this list is .05. That same variance is the margin of #5 over #18! That appears to match what you say in your note about how unusual this is. You described it as "very large".

Implies that if UP wins out in WCC chances for #1 seed might be good?

UPSF - great work, as usual.

Question: You mentioned before, I think, that the NCAA was releasing its initial RPI standings on Monday, Oct. 6th. I assume these are the first NCAA numbers, and not your database?

Just for fun, but not to put you on the spot because your RPI numbers are becoming more and more valuable to more and more of us, can you compare your top ten with the NCAA's top ten? I know yours is close, if not right on.

(I'm sure most of us would say we trust your numbers more than the NCAA's. I'm also aware the NCAA puts in some other criteria that you are not privy to. Correct?)

FANatic wrote:Question: You mentioned before, I think, that the NCAA was releasing its initial RPI standings on Monday, Oct. 6th. I assume these are the first NCAA numbers, and not your database?

Just for fun, but not to put you on the spot because RPI numbers are becoming more and more valuable to more and more of us, can you compare your top ten with the NCAA's top ten? I know yours is close, if not right on.

(I'm sure most of us would say we trust your numbers more than the NCAA's. I'm also aware the NCAA puts in some other criteria that you are not privy to. Correct?)

Actually, the numbers I gave you are my numbers, not the NCAA's. As typically has occurred in the past, the NCAA's published information is that it would post its numbers yesterday, but they haven't published them yet. Could be today or tomorrow or ? Once we've got the NCAA's rankings, we can do a comparison. For the teams on the list, my rankings should be very close to theirs. They only will post the rankings, by the way, not the actual RPI ratings.

The only part of the NCAA formula to which we are not privy is the bonuses for good wins/ties and the penalties for poor losses/ties. For those, the only way to figure them out is through a process of trial and error, inputting guesses into the computer program and seeing how close my ratings come to theirs. I'm pretty close with the bonuses, as my ratings matched theirs for about the top 80 teams last year. I'm not as close on the penalties, but they typically have minimal relevance to the teams under consideration for NCAA at large selections and seeds. One of the barriers to figuring out the bonuses and penalties is the potential for data errors either by me or by the NCAA. So, part of my process is vetting my numbers and theirs to be sure the NCAA and I are entering the same data.

Martin Jol wrote:The margin for UP over the 5th team on this list is .05. That same variance is the margin of #5 over #18! That appears to match what you say in your note about how unusual this is. You described it as "very large".

Implies that if UP wins out in WCC chances for #1 seed might be good?

I suspect, but am not sure, that the Pilots' margin will be big enough for a #1 seed if they win out. Since the dynamics of the RPI are complicated when it is dealing with strength of schedule, I'm not sure. Plus, the Pilots' strength of schedule is dependent on how the non-conference teams we played fare. So far, they've been faring very well, so let's hope that continues. We all can follow how the RPI evolves as we go through the season.

Last year, the difference between the top three teams at the end of the regular season was .0181. The difference between the top team and the #5 team was .0484. (These numbers are based on the NCAA's data last year, which as mentioned previously included an error that unfortunately gave Stanford a higher rating than it should have had.)

A complicating factor will be UCLA's win over the Pilots. In addition to considering the RPI, the NCAA considers head-to-head results and results against common opponents. So, if the relative rankings were to stay the same, there's an argument that UCLA should receive a seed at least equal to the Pilots even if they are outside the top four in the RPI. We can worry about that later, as we see who beats whom in the Pac 10.

Hello Pilot Nation! This is my first-ever post so it's going to be brief. For the past 2 months (when I landed on Pilot Nation out of curiosity) I have read with increasing interest and awe the incredible postings from all of you out there. But I got hooked big-time by the depth of knowledge, humor, goodwill, and obvious love of the Pilots by such posters as Purple Geezer, FANatic, and Stonehouse. I know it's dangerous to name names but those 3 deserve a huge pat-on-the-back in my book. But at the top of my list - bar none - is UPSF! His work on RPI's - and helping others to understand and appreciate how those ratings work - is nothing short of amazing. We are incredibly lucky to have him. But what I admire most is his appreciation and respect for everything the Pilots women's soccer program stands for - especially our late, great coach Clive Charles. Thanks for listening. Hope to add more as this exciting season moves on.

Shadrach,

welcome, and wonderful first post. You are absolutely correct we have a wealth of info here, and we are all very lucky to benefit from said knowledge. I am constantly in awe of the info, and the different perspectives that each bring, Stoney as an Alum, UPSF and his knowledge of the RPI of course Geez and his webspiders, and FANatic for his constant attention to details others miss. Hopefully we can continue to grow, and share.

Ha ha... you know, in thinking of what UPSF contributes in terms of the RPI stuff and what I contribute (lots of rah-rahism), I'm reminded of this old joke:

"The difference between Oregon and Oregon State is the difference between culture and agriculture."

(No offense to any Beavers out there... I am definitely unafilliated when it comes to the state schools.)

Anyway, somehow, someway this seems to make sense... substitute "The difference between Oregon and Oregon state" with "The difference between UPSF's RPI posts and my fan posts" and the punchline still works.

Welcome, Shadrach!

Welcome Shadrach. We have fun!

The NCAA now has published its RPI rankings. Use the following link to get them: http://web1.ncaa.org/app_data/weeklyrpi/2008WSOrpi1.html

My rankings and theirs are very close. Now I get to try to see if we have data inconsistencies; and then to see if I can make adjustments to my bonus/penalty amounts to match their results exactly. Good thing I like puzzles.

Wow some interesting ranking in there... Glad theirs have the Pilots #1 as well

In response to a question, I believe from FANatic, about how my RPI rankings compare to the NCAA's, here are my top 30 (first column) compared to the NCAA's top 30 (second column). We don't exactly match up on home-away-neutral data due to what I believe are a couple of reporting errors when teams entered their data into the NCAA's system, but I doubt our data differences make any difference in the comparisons. I suspect that the differences are due to my not quite having the correct bonus/penalty amounts. I'll be working on trying to tweak the bonus/penalty formula I use to see if I can exactly match the NCAA, but I'm pretty close. When I get farther down the list, where penalties for bad losses and ties start to creep in, I'm off more since I haven't spent as much time on the penalties than on the bonuses.

1. Portland U Portland U
2. North Carolina U North Carolina U
3. Stanford U Stanford U
4. Texas A&M Texas A&M
5. Notre Dame Notre Dame
6. Duke Duke
7. Florida U Florida U
8. Boston College Boston College
9. Florida State Florida State
10. Rutgers Rutgers
11. UCLA UCLA
12. Oklahoma State Oklahoma State
13. Illinois U Texas U
14. Texas U Illinois U
15. Virginia U Virginia U
16. Colorado U Colorado U
17. USC USC
18. Penn State Penn State
19. Washington U Washington U
20. California U California U
21. Wake Forest Wake Forest
22. Missouri U Missouri U
23. TCU South Carolina U
24. Long Beach State TCU
25. South Carolina U Long Beach State
26. UCF UCF
27. Old Dominion Old Dominion
28. Princeton Princeton
29. Georgetown Georgetown
30. William and Mary William and Mary

Here are some data comparing regions and conferences.

The region comparisons are based only in inter-regional games. Further, the system I use at this point in the season is able to calculate RPIs only for 208 teams, when the game data is limited to inter-regional games. Therefore the average team RPI by region is a number you have to consider with that in mind. As the season progresses and we get more inter-regional games, more teams will come into the picture.

Average RPI (Adjusted) By Region

Central .5153
Great Lakes .4779
Mid Atlantic .4844
Northeast .4924
Southeast .5014
West .5472

Total Win-Loss-Ties By Region

Central 93-100-15
Great Lakes 108-99-28
Mid Atlantic 108-119-33
Northeast 78-96-27
Southeast 125-118-29
West 82-62-22

Average Inter-Regional Games Per Team, By Region

Central 3.65
Great Lakes 3.92
Mid Atlantic 5.42
Northeast 4.90
Southeast 4.46
West 3.25

The Conference comparisons are based only on "out-of-conference" games. Since almost all the remaining games are intra-conference games, these numbers should change very little through the balance of the season.

Average RPI by Conference

Pac 10 .6104
ACC .6089
West Coast .5696
Big 10 .5664
Big East .5625
SEC .5610
Big 12 .5782
Ivy .5518
Big West .5419
Mountain West .5364
Colonial .5334
Conf USA .5287
Atlantic 10 .5205
Southern .4909
Horizon .4830
Mid American .4782
Sun Belt .4761
Atlantic Sun .4637
Metro Atlantic .4596
Missouri Valley .4562
America East .4551
Big Sky .4495
Patriot .4486
Big South .4483
Western Athletic .4471
Northeast .4444
Independent .4324
Summit .4295
Southland .4281
Ohio Valley .4231
United .3912
Southwestern .3318

Pepperdine won today v Cal State Bakersfield (weak opposition) and San Diego won against Long Beach State (strong opposition). Way to go WCC!

Brown came from two goals down to tie Bryant, 3-3. Bryant, fyi, is a new entrant to Division 1. Makes won wonder what was going on when Brown beat Penn State 1-0 and tied UCLA 0-0. Or, what has changed for Brown since then.

I found an error in how I computed the conferences' average RPIs. Here is the corrected list:

1 Pac 10 0.6103
2 ACC 0.6098
3 Big 12 0.5774
4 West Coast 0.5708
5 Big 10 0.5670
6 Big East 0.5647
7 SEC 0.5613
8 Ivy 0.5494
9 Big West 0.5413
10 Mtn West 0.5352
11 Colonial 0.5326
12 Conf USA 0.5298
13 Atlantic 10 0.5142
14 Southern 0.4916
15 Horizon 0.4830
16 Mid American 0.4781
17 Sun Belt 0.4682
18 Atlantic Sun 0.4631
19 Metro Atlantic 0.4571
20 America East 0.4570
21 Big South 0.4569
22 Missouri Valley 0.4563
23 Patriot 0.4510
24 Big Sky 0.4494
25 WAC 0.4462
26 Northeast 0.4459
27 Independent 0.4352
28 Southland 0.4265
29 Summit 0.4180
30 Ohio Valley 0.4127
31 United 0.3905
32 Southwestern 0.3303

Albyn Jones' SoccerRatings now are out for games through October 5. He has Portland at #3, close behind Notre Dame and North Carolina. Use the following link to see the complete ratings. (The West region, according to him, has 20 out of the top 66 teams.) http://soccerratings.com/index.php?title=Women%27s_NCAA_Division_I_Soccer_Ratings

Here are the RPIs after adding in this past weekend's games. You're getting all 318 teams, for what it's worth.

As you will see, the Pilots have dropped down to #3. Despite their win over Gonzaga, which increased their winning percentage, they also got to add Gonzaga's 5-6-1 record to their strength of schedule (not good); and unlike the week before, this week their prior opponents did not do well, having a record of 14-13-4 (also not good, since the Pilots' opponents' prior to this week had a winning percentage around 72%).

I've realized that every time one of the Pilots' Pac 10 opponents plays another of their Pac 10 opponents, it's like a WCC game in the sense that one opponent will win and another will lose, thus pulling the Pilots' strength of schedule towards .5000 (not good). There will be six games like that, of which two have been played so far (UW v USC and UW v UCLA).

The other factor affecting the Pilots' rank is that most of the other top teams played and won two games this past weekend, whereas the Pilots only played and won one; and many of the other top teams' weekend opponents had better records than Gonzaga, and their prior opponents may have done better than the Pilots' prior opponents over the last week.

There's still a pretty large gap between the #4 team and the #5 team, so that's good. However, as I've learned from this past week, at this stage in the season there can be pretty large shifts in teams' RPIs. The deeper into the season we get, the lesser the shift resulting from each game. (I.e., the Gonzaga game represents 1/12 of the Pilots' schedule so far and thus has a 1/12 impact on its RPI. By the end of the season, the last game will contribute only 1/18 of the impact.)

The other thing to consider is that some of the top contenders are from the same conferences and will be knocking each other off as conference play continues. So, winning out is critical for the Pilots.

1 North Carolina U 0.731093
2 Notre Dame 0.723069
3 Portland U 0.721687
4 Stanford 0.715122
5 Duke 0.687230
6 UCLA 0.686750
7 Texas A&M 0.683450
8 Boston College 0.674524
9 Florida State 0.674375
10 Oklahoma State 0.670747
11 Florida U 0.661791
12 Colorado U 0.658319
13 USC 0.650935
14 Texas U 0.643174
15 Virginia U 0.640469
16 Rutgers 0.636965
17 UCF 0.634571
18 Wake Forest 0.633809
19 South Carolina U 0.630458
20 Penn State 0.629112
21 California U 0.627618
22 Illinois U 0.625977
23 Michigan State 0.623641
24 BYU 0.615113
25 Virginia Tech 0.613762
26 Washington U 0.613223
27 Minnesota U 0.609224
28 William and Mary 0.607329
29 Long Beach State 0.606522
30 San Diego U 0.606359
31 Georgia U 0.604913
32 Old Dominion 0.603965
33 Missouri U 0.603483
34 Princeton 0.600944
35 Wisconsin Milwaukee 0.600240
36 Purdue 0.595093
37 East Carolina 0.594856
38 Marquette 0.592604
39 James Madison 0.592512
40 Georgetown 0.591643
41 Arizona U 0.589426
42 UNC Greensboro 0.589271
43 Oregon U 0.588229
44 Colorado College 0.587152
45 Charlotte 0.585805
46 Kansas U 0.585678
47 TCU 0.585260
48 Denver 0.585090
49 Washington State 0.584024
50 LSU 0.581956
51 Ohio State 0.581627
52 Miami FL 0.581210
53 New Mexico U 0.581193
54 Loyola Marymount 0.580562
55 UNC Wilmington 0.579816
56 Brown 0.578475
57 West Virginia U 0.577761
58 Arizona State 0.575532
59 Villanova 0.573284
60 Vanderbilt 0.573234
61 Pennsylvania U 0.572801
62 Santa Clara 0.572755
63 Saint Louis 0.572299
64 Auburn 0.571267
65 Hofstra 0.570978
66 Columbia 0.569682
67 Boston U 0.569522
68 Dayton 0.569309
69 Harvard 0.568409
70 Richmond 0.566025
71 Fairfield 0.564580
72 Memphis 0.564342
73 Northwestern U 0.559393
74 College of Charleston 0.558851
75 Rice 0.558784
76 Coastal Carolina 0.557585
77 UC Irvine 0.557091
78 UC Santa Barbara 0.556808
79 St Johns 0.556705
80 Radford 0.555958
81 Louisville 0.555785
82 Mississippi U 0.555164
83 Davidson 0.555084
84 Nebraska U 0.553586
85 Western Kentucky 0.551947
86 NC State 0.549723
87 Western Carolina 0.548612
88 Virginia Commonwealth 0.548479
89 Utah U 0.547949
90 Tennessee U 0.547763
91 Connecticut U 0.547190
92 Clemson 0.545851
93 Fordham 0.545246
94 Creighton 0.544620
95 U of Pacific 0.543594
96 UNLV 0.542038
97 Georgia State 0.540765
98 Belmont 0.540201
99 Yale 0.537483
100 St Marys 0.536922
101 Michigan U 0.535671
102 Kennesaw State 0.535520
103 Iowa U 0.535287
104 McNeese State 0.533983
105 Texas Tech 0.532318
106 Alabama U 0.532280
107 Seton Hall 0.531738
108 SMU 0.531209
109 Northern Arizona 0.529141
110 Dartmouth 0.528802
111 Evansville 0.528225
112 South Florida 0.526358
113 Central Michigan 0.525123
114 Furman 0.524407
115 UTEP 0.523972
116 Detroit 0.523894
117 St Josephs 0.523112
118 Siena 0.521677
119 Cincinnati 0.521525
120 Cal Poly 0.520045
121 CS Northridge 0.519425
122 Kentucky U 0.519140
123 UC Davis 0.518810
124 Pepperdine 0.517812
125 CS Fullerton 0.517689
126 Long Island 0.516962
127 Canisius 0.515604
128 St Bonaventure 0.515239
129 Toledo 0.514962
130 Navy 0.514678
131 Maryland U 0.513631
132 Loyola Chicago 0.513033
133 Wisconsin U 0.512340
134 Valparaiso 0.511465
135 Butler 0.511376
136 Northern Illinois 0.511127
137 Bowling Green 0.510524
138 South Dakota State 0.509956
139 San Diego State 0.509136
140 Samford 0.507084
141 Mercer 0.506639
142 Baylor 0.503724
143 Loyola MD 0.503390
144 Syracuse 0.500645
145 Army 0.500569
146 Ball State 0.500161
147 Central Connecticut 0.500035
148 Akron 0.499633
149 Wright State 0.499618
150 George Washington 0.499105
151 Arkansas U 0.498476
152 St Francis 0.497833
153 Duquesne 0.497630
154 Utah State 0.496377
155 Oregon State 0.495571
156 North Texas 0.495091
157 Northeastern 0.494802
158 Providence 0.494378
159 Monmouth 0.493503
160 Illinois State 0.492826
161 Marshall 0.492137
162 UAB 0.491842
163 Pittsburgh 0.491261
164 Hawaii U 0.491180
165 Niagara 0.491098
166 Eastern Michigan 0.490956
167 Missouri State 0.489938
168 Bucknell 0.489816
169 Quinnipiac 0.488352
170 Stony Brook 0.487531
171 Indiana U 0.487213
172 Florida Gulf Coast 0.487118
173 Boise State 0.486140
174 Gonzaga 0.485678
175 La Salle 0.485246
176 Oklahoma U 0.485151
177 Seattle 0.483234
178 Tulsa 0.482673
179 Iowa State 0.482542
180 Longwood 0.482152
181 Western Michigan 0.481630
182 Middle Tennessee 0.481441
183 High Point 0.479646
184 East Tennessee State 0.477686
185 Delaware U 0.477460
186 Drake 0.477431
187 Elon 0.477355
188 UC Riverside 0.476991
189 Murray State 0.476991
190 Jacksonville U 0.476851
191 Stephen F Austin 0.476312
192 Weber State 0.475892
193 Texas State 0.474343
194 UT Martin 0.472511
195 Florida Atlantic 0.472058
196 Rhode Island U 0.470573
197 Stetson 0.467149
198 Wyoming U 0.466895
199 Massachusetts U 0.464409
200 Hartford 0.464218
201 Oral Roberts 0.463354
202 Fresno State 0.462420
203 San Francisco 0.461548
204 Louisiana Tech 0.461181
205 George Mason 0.460433
206 Montana U 0.460388
207 Mississippi State 0.459917
208 Buffalo 0.456923
209 Northwestern State 0.455489
210 N Florida 0.452592
211 Southern Mississippi 0.452458
212 Cleveland State 0.450968
213 De Paul 0.450577
214 Utah Valley State 0.449510
215 Sam Houston State 0.449408
216 Colgate 0.448792
217 Kent 0.448377
218 Portland State 0.447905
219 Sacramento State 0.445145
220 Arkansas Little Rock 0.444619
221 Charleston Southern 0.444553
222 Towson 0.443020
223 Georgia Southern 0.442076
224 Houston 0.441976
225 Lehigh 0.441573
226 Campbell 0.438200
227 Manhattan 0.438124
228 Air Force 0.437625
229 Wofford 0.435536
230 Appalachian State 0.435146
231 Troy 0.434927
232 Southern Illinois 0.434604
233 IUPU Indianapolis 0.432144
234 Florida International 0.431916
235 Drexel 0.431883
236 American 0.431642
237 UTSA 0.430594
238 Cal State Bakersfield 0.430542
239 Xavier 0.429700
240 Ohio U 0.428366
241 UNC Asheville 0.428223
242 Houston Baptist 0.427404
243 South Alabama 0.426806
244 Maine U 0.426190
245 Robert Morris 0.423452
246 San Jose State 0.422367
247 SE Louisiana 0.421609
248 New Hampshire U 0.419695
249 Louisiana Lafayette 0.418676
250 Fairleigh Dickinson 0.417802
251 Idaho State 0.415616
252 UW Green Bay 0.415009
253 Liberty 0.414905
254 Miami OH 0.414742
255 Francis Marion 0.414732
256 Jacksonville State 0.414149
257 Oakland 0.414113
258 Louisiana Monroe 0.413240
259 Binghamton 0.412619
260 Alabama A&M 0.411797
261 VMI 0.409749
262 Southern Utah 0.409278
263 Cornell 0.408116
264 UT Chattanooga 0.406686
265 Marist 0.405899
266 Bryant 0.404144
267 Presbyterian 0.404118
268 Eastern Washington 0.403808
269 SE Missouri 0.403516
270 Temple 0.403330
271 Holy Cross 0.400100
272 Nevada U 0.399096
273 Eastern Kentucky 0.397200
274 North Dakota State 0.395330
275 Winthrop 0.393605
276 Youngstown State 0.388556
277 Arkansas State 0.388362
278 Morehead 0.386731
279 Tennessee Tech 0.385399
280 Northern Colorado 0.384434
281 UMBC 0.382654
282 Gardner Webb 0.382492
283 The Citadel 0.382303
284 Western Illinois 0.381071
285 Jackson State MS 0.379436
286 USC Upstate 0.379076
287 Vermont U 0.377231
288 Iona 0.376538
289 Albany 0.375255
290 Idaho U 0.370205
291 Sacred Heart 0.365843
292 Austin Peay 0.364676
293 Central Arkansas 0.362262
294 Lamar 0.359327
295 Howard 0.358504
296 Lipscomb 0.357917
297 Centenary 0.355020
298 Indiana State 0.349391
299 Grambling 0.349368
300 Eastern Illinois 0.348783
301 IPFW 0.340672
302 Wagner 0.339404
303 Delaware State 0.336923
304 Lafayette 0.336023
305 Mount St Mary 0.336023
306 Northern Iowa 0.331828
307 Arkansas Pine Bluff 0.328203
308 Mississippi Valley 0.327061
309 Rider 0.317757
310 NJIT 0.312241
311 St Peters 0.303433
312 Nicholls State 0.302662
313 Prairie View A&M 0.292232
314 SC State 0.279728
315 Southern U 0.258011
316 Texas Southern 0.222355
317 Alabama State 0.213891
318 Alcorn State 0.154799