Division 1 Ratings, Bracketology, Scheduling, Etc.

Discussion in 'Women's College' started by cpthomas, Sep 10, 2019.

  1. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Great and timely question.

    Brown's Winning Percentage is excellent -- 0.8500. That's the #15 ranked Winning Percentage and it accounts for 50% of the effective weight of their rating.

    The average of their Opponents' Winning Percentages is 0.6315. That's the #23 ranked OWP. It accounts for 40% of the effective weight of their rating.

    The RPI does take into account your Opponents' Opponents' Winning Percentages, but that only accounts for 10% of the effective weight of the RPI. Brown's OOWP rank is #73.

    You put those three together and compare them to the numbers of other teams and Brown comes in at #10.

    If you think about those last two as Brown's strength of schedule, which has an effective weight of 50% of their overall RPI, their Opponents' Winning Percentage accounts for 80% of the effective weight of their strength of schedule and the teams against whom they accomplished that winning percentage account for only 20%. This is a major flaw in the RPI, since it rewards you for playing teams with really good records almost regardless of who they played to achieve those records.

    What's interesting this year is that a bunch of the Ivies have had really good non-conference records. So, as Brown plays its Ivy League season, its Opponents' Winning Percentages are going to continue to be really good no matter who its Ivy League compatriots achieved their winning percentages against. (Although Brown does play a couple of non-conference opponents later who currently have poor records.) If Brown wins out in the League, it thus looks like their rank will continue to be really high.

    Brown has played some good opponents -- Hofstra, Texas A&M, and Providence aren't slouches. But overall it's benefitted from wisely or luckily -- based on the way the RPI formula works -- having scheduled opponents most of whom have turned out to have good records. From a scheduling perspective, the need to schedule this way is something teams are paying more and more attention to.
     
  2. Lovethegsme

    Lovethegsme New Member

    Nov 8, 2018
    Brown is a worthy NCAA tournament team. They went to Texas A&M and drew. Beat a top 25 Hofstra. Beat Providence. Ivy League teams are in a tough spot as they start a week later than everyone and not many top Power 5 schools in Northeast. Brown and Harvard would be fine in the tournament.
     
    hykos1045 repped this.
  3. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I've just published a full set of reports and simulations following Sunday's (October 6's) games.

    2019 RPI Ranks 10.14.2019 - These are teams' actual RPI ratings and ranks, and supplemental information, using the actual results of games played through Sunday, October 13. It includes the teams that are in the seed and at large selection "bubbles," at this stage of the season, based on past history. The report also includes teams' strength of schedule ranks and conferences' average RPI ranks. There also an explanation of why there are some minor differences between my ratings and ranks and the NCAA's. (Their data base is missing a game that the schools forgot to input into the NCAA's system.)

    2019 Simulated RPI Ranks 10.14.2019 - These are my simulated team RPI ratings and ranks using the actual results of games played through Sunday, October 13, and simulated results for the remainder of the season including simulated conference tournaments. The simulated results are based on teams' current RPI ratings.

    2019 Simulated Conference Standings and Conference Tournament Results 10.14.2019 - These are my simulated regular season conference standings and conference tournament brackets and results, using the actual results of games played through Sunday, October 6, and simulated results for the remainder of the season. Again, the simulated results are based on teams' current RPI ratings.

    2019 Simulated NCAA Tournament Bracket 10.14.2019 - This is my simulated NCAA Tournament bracket using the actual results of games played through last Sunday, October 13, and simulated results for the remainder of the season including simulated conference tournaments.

    In addition, at the RPI for Division I Womens Soccer website, there are two new Excel workbooks available for downloading for those who want more detailed information:

    2019 Website Factor Workbook 10.14.2019 - This is a relatively small three-worksheet Excel workbook that has the details underlying my simulated NCAA Tournament bracket for this week. It's an exhibit at the bottom of the NCAA Tournament: Predicting the Bracket, Track Your Team page at the RPI for Division I Women's Soccer website. If you're interested in the details, then before going to the workbook, read the text of the linked page, as it explains how to use the workbook.

    2019 RPI Report 10.14.2019 - This also is a relatively small three-worksheet Excel workbook that is similar to the 2019 RPI Ranks 10.14.2019, linked above, but it has considerably more detail. It's an exhibit at the bottom of the RPI: This Year's Reports page at the RPI for Division I Women's Soccer website.
     
  4. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I've just published a full set of reports and simulations following Sunday's (October 20's) games.

    2019 RPI Ranks 10.21.2019 - These are teams' actual RPI ratings and ranks, and supplemental information, using the actual results of games played through Sunday, October 20. It includes the teams that are in the seed and at large selection "bubbles," at this stage of the season, based on past history. The report also includes teams' strength of schedule ranks and conferences' average RPI ranks.

    2019 Simulated RPI Ranks 10.21.2019 - These are my simulated team RPI ratings and ranks using the actual results of games played through Sunday, October 20, and simulated results for the remainder of the season including simulated conference tournaments. The simulated results are based on teams' current RPI ratings.

    2019 Simulated Conference Standings and Conference Tournament Results 10.20.2019 - These are my simulated regular season conference standings and conference tournament brackets and results, using the actual results of games played through Sunday, October20, and simulated results for the remainder of the season. Again, the simulated results are based on teams' current RPI ratings.

    2019 Simulated NCAA Tournament Bracket 10.20.2019 - This is my simulated NCAA Tournament bracket using the actual results of games played through last Sunday, October 20, and simulated results for the remainder of the season including simulated conference tournaments.

    In addition, at the RPI for Division I Womens Soccer website, there are two new Excel workbooks available for downloading for those who want more detailed information:

    2019 Website Factor Workbook 10.21.2019 - This is a relatively small three-worksheet Excel workbook that has the details underlying my simulated NCAA Tournament bracket for this week. It's an exhibit at the bottom of the NCAA Tournament: Predicting the Bracket, Track Your Team page at the RPI for Division I Women's Soccer website. If you're interested in the details, then before going to the workbook, read the text of the linked page, as it explains how to use the workbook.

    2019 RPI Report 10.21.2019 - This also is a relatively small three-worksheet Excel workbook that is similar to the 2019 RPI Ranks 10.21.2019, linked above, but it has considerably more detail. It's an exhibit at the bottom of the RPI: This Year's Reports page at the RPI for Division I Women's Soccer website.
     
  5. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I've just published a full set of reports and simulations following Sunday's (October 27's) games. We're getting progressively closer to the end of the season, and the simulations are getting progressively closer to where we'll end up once we get there.

    2019 RPI Ranks 10.28.2019 - These are teams' actual RPI ratings and ranks, and supplemental information, using the actual results of games played through Sunday, October 27. It includes the teams that are in the seed and at large selection "bubbles," at this stage of the season, based on past history. The report also includes teams' strength of schedule ranks and conferences' average RPI ranks.

    2019 Simulated RPI Ranks 10.28.2019 - These are my simulated team RPI ratings and ranks using the actual results of games played through Sunday, October 27, and simulated results for the remainder of the season including simulated conference tournaments. The simulated results are based on teams' current RPI ratings as adjusted for home field advantage. Where conference tournament brackets already are set, I use those brackets.

    2019 Simulated Conference Standings and Conference Tournament Results 10.28.2019 - These are my simulated regular season conference standings and conference tournament brackets and results, using the actual results of games played through Sunday, October27, and simulated results for the remainder of the season. Again, the simulated results are based on teams' current RPI ratings as adjusted for home field advantage. And, where conference tournament brackets already are set, I use those brackets.

    2019 Simulated NCAA Tournament Bracket 10.28.2019 - This is my simulated NCAA Tournament bracket using the actual results of games played through last Sunday, October 27, and simulated results for the remainder of the season including simulated conference tournaments.

    In addition, at the RPI for Division I Womens Soccer website, there are two new Excel workbooks available for downloading for those who want more detailed information:

    2019 Website Factor Workbook 10.28.2019 - This is a relatively small three-worksheet Excel workbook that has the details underlying my simulated NCAA Tournament bracket for this week. It's an exhibit at the bottom of the NCAA Tournament: Predicting the Bracket, Track Your Team page at the RPI for Division I Women's Soccer website. If you're interested in the details, then before going to the workbook, read the text of the linked page, as it explains how to use the workbook.

    2019 RPI Report 10.28.2019 - This also is a relatively small three-worksheet Excel workbook that is similar to the 2019 RPI Ranks 10.21.2019, linked above, but it has considerably more detail. It's an exhibit at the bottom of the RPI: This Year's Reports page at the RPI for Division I Women's Soccer website.
     
    Mills, EvanJ and socalsoccer23 repped this.
  6. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I've just published a full set of reports and simulations following Sunday's (November 3's) games. We're one week from the end of the regular season, including conference tournaments, and if teams' perform according to their current ratings as adjusted for home field advantage, these should be pretty close to where we end up. Of course, however, as last weekend's games show, teams may not perform in accord with their ratings. In fact, there may be some pretty big upsets, especially in conference tournaments, so we still could see some significant moving around of teams.

    2019 RPI Ranks 11.4.2019 - These are teams' actual RPI ratings and ranks, and supplemental information, using the actual results of games played through Sunday, November 3. It includes the teams that are in the seed and at large selection "bubbles," at this stage of the season, based on past history. The report also includes teams' strength of schedule ranks and conferences' average RPI ranks. All 3 of my, Chris Henderson's AllWhite Kit, and the NCAA's numbers match.

    2019 Simulated RPI Ranks 11.4.2019 - These are my simulated team RPI ratings and ranks using the actual results of games played through Sunday, November 11, and simulated results for the remainder of the season including simulated conference tournament results. The simulated results are based on teams' current RPI ratings as adjusted for home field advantage.

    2019 Simulated Conference Standings and Conference Tournament Results 11.4.2019 - These are my simulated conference tournament results, using the actual results of games played through Sunday, November 3, and simulated results for the remainder of the tournaments. Again, the simulated results are based on teams' current RPI ratings as adjusted for home field advantage.

    2019 Simulated NCAA Tournament Bracket 11.4.2019 - This is my simulated NCAA Tournament bracket using the actual results of games played through last Sunday, November 3, and simulated results for the remainder of the season including conference tournament games.

    In addition, at the RPI for Division I Womens Soccer website, there are two new Excel workbooks available for downloading for those who want more detailed information:

    2019 Website Factor Workbook 11.4.2019 - This is a relatively small three-worksheet Excel workbook that has the details underlying my simulated NCAA Tournament bracket for this week. It's an exhibit at the bottom of the NCAA Tournament: Predicting the Bracket, Track Your Team page at the RPI for Division I Women's Soccer website. If you're interested in the details, then before going to the workbook, read the text of the linked page, as it explains how to use the workbook.

    2019 RPI Report 11.4.2019 - This also is a relatively small three-worksheet Excel workbook that is similar to the 2019 RPI Ranks 11.4.2019, linked above, but it has considerably more detail. It's an exhibit at the bottom of the RPI: This Year's Reports page at the RPI for Division I Women's Soccer website.
     
    McSkillz, socalsoccer23 and EvanJ repped this.
  7. Crazyhorse

    Crazyhorse Member

    Dec 29, 2007
    Nat'l Team:
    United States
    Could someone fill me in on Arkansas' outstanding season?
     
  8. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    #33 cpthomas, Nov 6, 2019
    Last edited: Nov 6, 2019
    An interesting question this year is how the Committee is going to look at the Ivy League. There's a pretty good chance they'll have four teams in the Top 50 at the end of the season: Brown 10, Harvard 39, Yale 43, Columbia 49. This means that each of them will have played whatever non-conference Top 50 teams that were on their schedules plus three Ivy Top 50 teams. And the League as a whole likely will end up with an RPI rank of 5. That looks pretty darned good.

    But, consider this:

    Brown has two good non-conference results (wins or ties being good) against Top 50 teams: It has a home win against Hofstra (likely to finish 16) and an away tie against Texas A&M (likely to finish 24). These make it a legitimate high level team, maybe a little more highly ranked than they should be but not a lot.

    Harvard has no good non-conference results against Top 50 teams. In fact, its best non-conference result is a home win against likely 117 UMass.

    Yale has no good non-conference results against Top 50 teams. Its best non-conference result is a home win against likely 109 Fairfield.

    Columbia has no good non-conference results against Top 50 teams. Its best non-conference result is a home win over likely169 St Joseph's.

    In fact, the League as a whole has only three good results against Top 50 teams. Brown's two good results are two of them. Princeton has the other, a home tie against likely 7 Rutgers.
    If you look at this and think something doesn't look right, I agree.

    Whether by intention or by luck, the Ivy League seems to have "tricked" the RPI by scheduling non-conference opponents against which the League as a whole was able to achieve a very high non-conference winning percentage -- coming in only behind the Pac 12 and the ACC. The way the RPI currently is designed, the primary way conferences distinguish themselves from each other is by their non-conference winning percentage. And, it matters not very much against whom the conference achieved that non-conference winning percentage. Thus the Ivy League this year, mostly by having a top non-conference winning percentage, has gotten four of its teams into the Top 50 notwithstanding that three of those teams have played very weak non-conference schedules. (Another factor helping the Ivies is that their proportion of non-conference games is higher than that for most other conferences. This makes having a great overall non-conference winning percentage especially beneficial for them.)

    I, and others, will be watching closely to see if the Committee's decisions indicate the Committee has figured this out.

    Later, I'll do a detailed case study of the Ivy League this year to show how the RPI's current structure has allowed this to happen.
     
    EvanJ and ping repped this.
  9. EvanJ

    EvanJ Member+

    Manchester United
    United States
    Mar 30, 2004
    Club:
    Manchester United FC
    Nat'l Team:
    United States
    39, 43, and 49 are bubble spots for Men's and Women's Basketball where a mid-major could be left out without it being a big deal. It's different for the Ivy League when they're ranked 5th, but unless the Committee specifically talks about it there won't be a way of knowing if they considered Ivy League teams to be worse than the RPIs in general.
     
  10. hykos1045

    hykos1045 Member

    May 10, 2010
    Club:
    Philadelphia Independence
    Nat'l Team:
    United States
    Still remembering women's basketball 13-ranked Princeton getting a 8 seed, with a 31-1 record. Almost will be surprised if Ivy does get two bids tomorrow. RPI is seen as one of the qualifications only, but is not used as a tiebreaker.
     
  11. Lovethegsme

    Lovethegsme New Member

    Nov 8, 2018
    Brown won the league with a 9 RPI.
    Harvard cane in 2nd and has the head to head vs Yale but Yale has a better RPI. Going to be very interesting. Georgetown is on bubble as well. Power 5 schools will get the nod as the always do which is not fair.
     
  12. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I've just posted my Final NCAA Tournament Projected Bracket, together with a brief explanation of the method I use and a few comments. It will be very interesting to see what the Committee does this year.
     
    Crazyhorse and Mills repped this.
  13. stubifier

    stubifier Member

    Real Salt Lake
    United States
    Jan 19, 2018
    If BYU gets a lower seed than Kansas, my faith in the RPI as such a dominant metric in the selection committee's decision-making process will have been significantly eroded. BYU won in Lawrence, and then didn't lose a game all season. If the numbers say they only deserve a 3 or 4 seed, fine. The committee has some difficult choices to make. But if they end up below Kansas, it will feel like a farce.
     
  14. Crazyhorse

    Crazyhorse Member

    Dec 29, 2007
    Nat'l Team:
    United States
    I just looked at it good stuff! Thank you. I am a bit perplexed at the other Chris' projections this year....South Carolina as a one seed :eek::eek::eek::eek:
     
  15. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I have them as the top #2 seed, so they're definitely a possible #1.
     
    Crazyhorse repped this.
  16. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Chris Henderson says about the same thing as you. I haven't looked closely at why my system puts Kansas ahead of BYU, but I guess eventually I'll take a close look at what it says about both of them. My system looks at a whole bunch of metrics, and when it considers head-to-head results it doesn't look at just one game, because of the A beat B, B beat C, C beat A problem, which says you can't just look at A beat B.
     
  17. stubifier

    stubifier Member

    Real Salt Lake
    United States
    Jan 19, 2018
    Which I think is completely fair, as most teams end up in one of those head-to-head "chains." But does the question apply to an undefeated team? There clearly has to be some sort of limit here, as an undefeated team with a weak OOC schedule and conference shouldn't get a #1 seed. But BYU also beat some really good teams, even if they also played a softer schedule in general.
     
  18. Nacional Tijuana

    Nacional Tijuana St. Louis City

    St. Louis City SC
    May 6, 2003
    San Diego, Calif.
    Club:
    Seattle Sounders
    Nat'l Team:
    United States
  19. McSkillz

    McSkillz Member+

    ANGEL CITY FC, UCLA BRUINS
    United States
    Nov 22, 2014
    Los Angeles
    Nat'l Team:
    United States
    Wha wha to the wah wah, anyone else watching this?
     
  20. Mills

    Mills Member

    Aug 23, 2019
    #45 Mills, Nov 11, 2019
    Last edited: Nov 11, 2019
  21. Mills

    Mills Member

    Aug 23, 2019
  22. McSkillz

    McSkillz Member+

    ANGEL CITY FC, UCLA BRUINS
    United States
    Nov 22, 2014
    Los Angeles
    Nat'l Team:
    United States
    Damn, I hate our place in the bracket.....
     
  23. Crazyhorse

    Crazyhorse Member

    Dec 29, 2007
    Nat'l Team:
    United States
    Besides Stanford, UCLA is the one other team I would rather FSU avoided. Ya'll ran us off the field earlier in the year, not to mention you all should of been higher seeded than FSU.
     
  24. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    While I'll do, and post, some detailed analyses of some of the Committee's decisions, there's a decent rationale (whether I would have made the same decision or not) for almost all of the Committee's decisions. So long as that's the case, I can't criticize the decisions.

    The ones I need to spend some time on, to see what the rationales might have been, are:

    #32 Florida Atlantic not getting an at large selection

    #10 Brown not getting a seed (although this doesn't completely surprise me and, I'm guessing, will turn out legit)

    #22 Penn State getting a seed (although this is very close to being a case I wouldn't feel the need to check out)

    Overall, given some of the difficult issues it had to deal with this year, I think the Committee did a really good job.

    PS - I had 14 of the 16 seeds, including all of the #1s in the right order. I had Brown as a #3 (with some "buts" on my mind) and Duke at #4, neither of which got seeds. I didn't have Texas Tech or Penn State at #4 although they were close, particularly Texas Tech.

    On the at larges, like Chris Henderson I had all but 2 right, although my two were different than his. I had Florida Atlantic and Tennessee (the latter with reservations due to its #11 SEC regular season finishing position) and didn't have Iowa and Utah although they were close.
     
  25. Mills

    Mills Member

    Aug 23, 2019
    I'm pretty certain that the NCAA's travel restrictions for round 1 was the major reason for their # 16 seed.
    Also, to a lesser extent, PSU is on a 9 game winning streak, including winning the Big10 championship.
    To be honest, though, I'd prefer if PSU had not been seeded.
     

Share This Page