Teams' Trends

Discussion in 'Women's College' started by cpthomas, Nov 22, 2018.

  1. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I'm wondering about this one, but the numbers suggest a move up candidate may be Glad Bugariu at Texas RGV.

    upload_2018-12-12_20-10-41.png
     
  2. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    #52 cpthomas, Dec 12, 2018
    Last edited: Dec 12, 2018
    A note: I am not including Power 5 coaches among the ones I'm identifying, from the charts, as potential move up candidates, even if their teams' trends show significant improvement year over year. That's because I'm doubting they are likely to move, although it does happen occasionally. When I've finished going through all the teams and have published my workbook, those interested will be able to look at all the Power 5 teams and see which teams have been improving over time.
     
  3. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Tennessee:

    upload_2018-12-13_14-43-32.png
     
    L'orange repped this.
  4. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Another possible move up candidate, Chris Cissell at UMKC:

    upload_2018-12-13_15-18-42.png
     
  5. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    This chart for Vermont suggests that Kristi Lefebvre might be a candidate to move up a level. It also illustrates where an evaluator needs to be careful in considering a team's trend -- which I'll point out in some notes below the chart.

    upload_2018-12-14_12-1-44.png

    As the chart shows, from the year before Lefebvre arrived, Vermont's ARPI trend has improved at a rate of about 9.5 rank positions per year. If you consider Massey as most likely to be showing a team's true rank, however, the rate of improvement is smaller, about 6 to 7 positions per year. This still is good.

    The other thing to look at is the America East line in the "team information" area of the chart. The America East performance percentages all are in the 80s, which for Massey and the 5 Iteration ARPI version is very unusual. These percentages mean that all of the rating systems have trouble properly rating America East teams, significantly overrating them on average. So, although Vermont's improvement trend is as shown on the chart, the rank levels the chart shows most likely are making Vermont look better than it really is, from one end of the chart to the other.

    Why do the systems have trouble properly rating America East teams? There's a combination of reasons. First, the America East teams play in what I call the East regional playing pool, which is a group of conferences whose teams typically play a majority of their games against teams from that pool. As compared to other regional playing pools, the East pool plays a high percentage of its games within that pool -- 76.0%, second only to the West pool. For the West pool, there's a practical geographic reason for its high percentage of games played "in pool," but for the East pool it's more a matter of choice. And, looking specifically at the America East, it plays 90.5% of its games within the East pool. Further, of its East pool non-conference games, America East plays almost two-thirds against conferences that are in the bottom half of the conference rankings. Thus America East largely has isolated itself from the rest of the country and from opponents from strong conferences. When you have a situation like that, no rating system can properly rate its teams in relation to those from other conferences and, if the conference is weak as is the case with America East, all rating systems will tend to overrate its teams.
     
  6. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Here are two Virginia Tech charts, one starting from 2007 which shows Virginia Tech''s trend basically as being "no change" and one starting from 2010, the year before Chugger Adair arrived, and showing a slight decline of about 1.5 positions per year.

    upload_2018-12-14_14-38-33.png

    upload_2018-12-14_14-39-2.png
     
  7. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    VMI's Chris Bergmann bears watching for a possible move up, although it's a little early to tell:

    upload_2018-12-14_14-54-42.png
     
  8. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    And, last but not least, Wake Forest:

    upload_2018-12-14_15-7-36.png
     
  9. Lord Kril

    Lord Kril Member

    Pittsburgh Riverhounds
    Jul 3, 2018
    Some interesting things here. Thanks for doing the research.
     
  10. Carolina92

    Carolina92 Member

    Sep 26, 2008
    @cpthomas I think there are errors in the 2018 rankings for Va Tech and Wake Forest. Va Tech finished 22nd in the Massey rankings for 2018, while Wake Forest finished 23rd. Your charts show them both finishing much worse.
     
  11. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    Actually, my numbers and yours both are correct! Yours are Massey's ratings after he added the NCAA Tournament games into his data base. Mine are after completion of the regular season but before the NCAA Tournament. At the end of the regular season Massey had Virginia Tech at #44, and after adding in the NCAA Tournament results they moved all the way up to #22. Likewise Wake Forest moved up from #41 to #23.

    I always use numbers that do not include the NCAA Tournament games, for a specific reason. My interest is in the information that the Women's Soccer Committee uses when setting up the Tournament bracket. That includes the RPI and other games data. And, when I do comparisons of other systems to the RPI, I likewise use those systems' ratings before the NCAA Tournament games are entered into their data bases. (Thus, for Massey, I always have to remember to download his ratings before the Tournament begins.)

    Another reason for my doing this is that it provides more useful information to coaches for scheduling purposes. What they want to know is, What will our playing X team do for us in terms the information the Committee will be looking at when making its NCAA Tournament decisions? By using only the regular season data, I'm able to provide the most useful information in response to that question.
     
    Carolina92 repped this.
  12. cpthomas

    cpthomas BigSoccer Supporter

    Portland Thorns
    United States
    Jan 10, 2008
    Portland, Oregon
    Nat'l Team:
    United States
    I've finished with all of the teams. At the RPI for Division I Women's Soccer website, I've uploaded an Excel workbook with a Guide page, a Data page, and a page for each team. Each team page includes a table and a chart like the ones I've been posting here. I think the team pages, particularly the charts, present team rank and trend information in a manner that makes it easy to get a sense of what's going on with a team. And, although I originally did not start this project as a scheduling tool, I think it will be really helpful to teams as they work on their future non-conference schedules.

    For practical purposes, if you want to use the workbook, I think you'll need to download it as an Excel file. I've tried to open it directly on the website rather than downloading it, and it being a Google site it won't open the workbook as an Excel file and appears not able to convert it to a Google spreadsheet. It will open it, but not in a form that I find usable. So, again, I suggest you download it as an Excel file and simply use it on your computer.

    If you download it, it's set up now to open at the Guide page. I suggest you review that at the start and then have at it.

    The workbook is an attachment to the NCAA Tournament: Scheduling Towards the Tournament page at the website. At the bottom of the page, you'll see the attachment Team Histories Since 2007, which is the workbook. To download it, go to the right of that attachment and click on the download arrow.

    Please let me know if you have any questions or comments.
     
    L'orange repped this.

Share This Page