Announcement

Collapse
No announcement yet.

72% chance we lose 1st rd

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • 72% chance we lose 1st rd

    according to some Harvard model


  • #2
    Make a computer model with a small sample size.
    Test it one year against a really small sample size.
    If it fails, don't publicize it.
    If it succeeds, publicize it.

    Comment


    • #3
      "Never tell me the odds!!"

      -- Han Solo
      "When life hands you lemons, make lemonade." Better have some sugar and water too, or else your lemonade will suck!

      Comment


      • #4
        And Nate Silver says we have a 74% chance of WINNING the first game: http://www.nytimes.com/interactive/2...h-madness.html
        There are nights when I think that Sal Paradise was right.

        Comment


        • #5
          To create the model, Ezekowitz analyzed every opening game played since 2004 and found three common threads among teams that pull upsets: They consistently win the turnover battle; they rebound well and they play a strong schedule. Shooting mattered little.
          Sounds like either somebody didn't understand the model results Or the model needs to be relooked at. Wsu played a tougher schedule and is a better rebounding team. Now you could argue how close the turnover game is between. The two teams - but 74% is fails a general sanity check.

          Comment


          • #6
            There is a 100% chance I will have several adult beverages in me by tipoff.
            "When life hands you lemons, make lemonade." Better have some sugar and water too, or else your lemonade will suck!

            Comment


            • #7
              Actually reading from the modeler blog he has warned that vcu result us a big outlier and may not do as well as predicted - nice waffling. Basically he saying "even though I predict them to win they might very well lose". :)

              Comment


              • #8
                He seems to weight the turnover margin high, which will heavily favour VCU.

                Comment


                • #9
                  If you want more detail on the methodology, go here:

                  http://harvardsportsanalysis.wordpre...ing/#more-2959

                  Scroll down to the second topic.
                  "Prediction is very difficult, especially if it is about the future."

                  --Niels Bohr







                  Comment


                  • #10
                    So basically in the near future, a computer might predict the perfect bracket?

                    Comment


                    • #11
                      Originally posted by shox27 View Post
                      95% chance we fire Harvard.
                      Some posts are not visible to me. :peaceful:
                      Don't worry too much about it. Just do all you can do and let the rough end drag.

                      Comment


                      • #12
                        Or you could go with a source that predicts an enormous number of games and quantifies the results, like Raymond Cheong.



                        I've posted this image before, but this is a chart ranking prediction systems for NCAA games. Cheong's system (which has us as 9 point winners over VCU) is the second best and Bayesian LRMC, which has us ranked as the #6 team in the nation is the best. In fact, since most of these were drawn from the Massey Comparison Ratings, I don't think there is a single one of these that doesn't predict a WSU victory (although winning percentage - the single worst predictor - is among the few where the two teams are closely ranked).

                        "Cotton scared me - I left him alone." - B4MSU (Bear Nation poster) in reference to heckling players

                        Comment


                        • #13
                          This makes me wanna go into the basement of the science building and start beating up nerds!
                          "When life hands you lemons, make lemonade." Better have some sugar and water too, or else your lemonade will suck!

                          Comment


                          • #14
                            Originally posted by ShockerPrez View Post
                            This makes me wanna go into the basement of the science building and start beating up nerds!
                            They would probably beat you up if you tried. And that's just the female graduate students. :p :D
                            Some posts are not visible to me. :peaceful:
                            Don't worry too much about it. Just do all you can do and let the rough end drag.

                            Comment


                            • #15
                              Originally posted by Ricardo del Rio View Post
                              If you want more detail on the methodology, go here:

                              http://harvardsportsanalysis.wordpre...ing/#more-2959

                              Scroll down to the second topic.
                              the model success rate is 42% 15-6 identified upsets, missed additional 15 upsets or 15 out 36.

                              Comment

                              Working...
                              X