Page 2 of 2 FirstFirst 12
Results 16 to 20 of 20
  1. #16

    Default

    Quote Originally Posted by Some guy View Post
    That is not always the case.
    On average, efficiency in all of the games will lead to winning records.
    In individual cases it may be completely flawed.


    It is funny you say that the reason my example is bad is too small of a sample a size. Because a 30 game schedule to judge 350 teams is exactly why efficiency is a bad metric to use as the primary ranker of teams. Too small of a sample size.
    In a sample size that small, things like wins, winning percentage, and SOS are more reliable to judge quality of accomplishment than point margins.


    If all the ncaa teams played everyone, efficiency would be a better metric.
    Efficiency measures how well you played in any time frame. You can not lose a game if your more efficient than your opponent. No way to just look at a teams win totals. You have many 20-10 teams far better than a team that won 25 games against a weak schedule. NET is used because not everyone plays each other. What measure do you want to use to try and find out 68 deserving teams and seeds ? Keep in mind that i do not think there are 68 good teams or deserving teams. I like the tourney when 24 teams got in, with the top 8 getting bye's as a reward for a truly great regular season. This is before they diminished the regular season for many fans and before they put all the attention into a crap shoot of a tourney. I liked getting teams in that proved great skill over 4 months instead of 3 weeks. But, i also embrace change and coaches wanted a way to measure deserving teams receiving due for playing tougher games and a examination of how they were played. Straight wins just does not do that.


  2. #17

    Default

    Quote Originally Posted by builtbadgers View Post
    Efficiency measures how well you played in any time frame. You can not lose a game if your more efficient than your opponent. No way to just look at a teams win totals. You have many 20-10 teams far better than a team that won 25 games against a weak schedule. NET is used because not everyone plays each other. What measure do you want to use to try and find out 68 deserving teams and seeds ? Keep in mind that i do not think there are 68 good teams or deserving teams. I like the tourney when 24 teams got in, with the top 8 getting bye's as a reward for a truly great regular season. This is before they diminished the regular season for many fans and before they put all the attention into a crap shoot of a tourney. I liked getting teams in that proved great skill over 4 months instead of 3 weeks. But, i also embrace change and coaches wanted a way to measure deserving teams receiving due for playing tougher games and a examination of how they were played. Straight wins just does not do that.
    Correct. I agree. Good metric to judge similar record teams.


    But when you say 17-13 team is better than 20-10 team with similar SOS it is ridiculous

  3. #18

    Default

    Quote Originally Posted by Some guy View Post
    Correct. I agree. Good metric to judge similar record teams.


    But when you say 17-13 team is better than 20-10 team with similar SOS it is ridiculous
    I think "better" is not quite the right concept. If this metric is used to select teams for the NCAA tournament, the concept should be "deserving." I would say that the 20-10 team with a SOS not significantly different from the 17-13 team is more deserving. If the 17-13 team had some expectations going into the season, then that team probably underachieved on the W/L dimension relative to those expectations. In retrospect, I think RPI really was a better measure for tournament selection because it considered only W/L and SOS. More quantitative complexity isn't always better, particularly if the greater "precision" isn't measuring the fundamental concept.

  4. #19

    Default

    Any idea when the schedule will drop?

  5. #20
    Join Date
    Nov 2008
    Location
    Bracketville, MN
    Posts
    21,532
    Blog Entries
    1

    Default

    Quote Originally Posted by Gophers7633 View Post
    Any idea when the schedule will drop?
    Last year Big Ten schedule (with dates) was released Aug. 21, complete schedule (with non-conference) not until Sept. 6.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •