1. Forum
  2. Usenet
  3. REC.SPORT.FOOTBALL.COLLEG

  • Re: Jeff Sagarin denounces Seattle Times

    From JGibson@21:1/5 to Randolph M. Jones on Wed Apr 12 13:12:45 2023
    On Monday, December 7, 1998 at 3:00:00 AM UTC-5, Randolph M. Jones wrote:
    Daryl D. Spillmann wrote:
    No, I think my description is pretty close. The only difference is what you pointed out. Jeff Sagarin develops a set of criteria for what he thinks determines the strngth of a footabll team. However, those criteria are his, and
    his alone. To his credit, he does not deviate from these rules as the season
    progresses. However, the rankings he develops are still just his opnion.

    There's another factor that's being left out of this discussion. It's true that
    Sagarin's system sets up a set of fixed rules by which to evaluate teams during
    the course of a season, and that makes the system somewhat objective (although it's being pointed out here that creating the rules in the first place is
    a subjective process).
    HOWEVER, Sagarin releases a set of rankings before the season even begins! This means that there is a clear and strong subjective element to his rankings.
    I believe he claims that the initial seedings get washed out as the season progresses, but that claim is very difficult to evaluate, considering that Sagarin is unwilling to publish his formulas. If Sagarin's system is an "incremental" system (computing new rankings by making incremental adjustments to last week's rankings, rather than recomputing all rankings from
    scratch), then it is certainly subject to the complex systems' result that initial seeding can have an enormous impact on the system's behavior. This gives Sagarin's rankings the potential to be *extremely* subjective.
    If computer rating systems are going to be used in the long run to evaluate teams for national rankings (which I generally support), I feel that it is inexcusable to use any rating system that is not published and open to full scrutiny. Frankly, I'm somewhat skeptical of Sagarin's claims that we can trust
    his system because he went to MIT and he used to play basketball.
    Randy Jones

    Just wanted something other than Wordle, Stormy, or Tucker to be at the top.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From RoddyMcCorley@21:1/5 to JGibson on Fri Apr 14 16:08:15 2023
    On 4/12/2023 4:12 PM, JGibson wrote:
    On Monday, December 7, 1998 at 3:00:00 AM UTC-5, Randolph M. Jones wrote:
    Daryl D. Spillmann wrote:
    No, I think my description is pretty close. The only difference is what you >>> pointed out. Jeff Sagarin develops a set of criteria for what he thinks
    determines the strngth of a footabll team. However, those criteria are his, and
    his alone. To his credit, he does not deviate from these rules as the season
    progresses. However, the rankings he develops are still just his opnion. >>>
    There's another factor that's being left out of this discussion. It's true that
    Sagarin's system sets up a set of fixed rules by which to evaluate teams during
    the course of a season, and that makes the system somewhat objective
    (although it's being pointed out here that creating the rules in the first place is
    a subjective process).
    HOWEVER, Sagarin releases a set of rankings before the season even begins! >> This means that there is a clear and strong subjective element to his rankings.
    I believe he claims that the initial seedings get washed out as the season >> progresses, but that claim is very difficult to evaluate, considering that >> Sagarin is unwilling to publish his formulas. If Sagarin's system is an
    "incremental" system (computing new rankings by making incremental
    adjustments to last week's rankings, rather than recomputing all rankings from
    scratch), then it is certainly subject to the complex systems' result that >> initial seeding can have an enormous impact on the system's behavior. This >> gives Sagarin's rankings the potential to be *extremely* subjective.
    If computer rating systems are going to be used in the long run to evaluate >> teams for national rankings (which I generally support), I feel that it is >> inexcusable to use any rating system that is not published and open to full >> scrutiny. Frankly, I'm somewhat skeptical of Sagarin's claims that we can trust
    his system because he went to MIT and he used to play basketball.
    Randy Jones

    Just wanted something other than Wordle, Stormy, or Tucker to be at the top.

    Thanks. I needed that.
    --
    "In theory, there is no difference between theory and practice. In
    practice, there is." Ruben Goldberg

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)