Announcement

Collapse
No announcement yet.

2012 NCAA Tournament: Bracketology

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #61
    Re: 2012 NCAA Tournament: Bracketology

    Originally posted by Priceless View Post
    Hmmm there is a discrepancy between sites.

    Not sure where the difference is...
    Comparing the PWR Details for Minnesota, USCHO seems to have a different RPI for Minnesota, such that Miami wins the comparison. Digging into the RPI details, it looks like USCHO isn't dropping the wins over Sacred Heart (see also USCHO RPI and CHN RPI).

    I don't know if they know something we don't or are messing up.
    Last edited by JimDahl; 01-23-2012, 08:31 AM.

    Comment


    • #62
      Re: 2012 NCAA Tournament: Bracketology

      Originally posted by streaker View Post
      Funny thing about that is the W hosting both western regions and probably would be favored to come out of them.
      I wouldn't bet on MN to advance out of their region unless they figure out how to do something other than split every effing weekend.

      Comment


      • #63
        Re: 2012 NCAA Tournament: Bracketology

        Originally posted by Bakunin View Post
        I wouldn't bet on MN to advance out of their region unless they figure out how to do something other than split every effing weekend.
        Agreed. Minnesota has been far too inconsistent over the past two months to be a legitimate threat for the National Title. Obviously, there are two months left for the Gophers to correct that. They have the tools to do so, and I think it is just a mental hurdle that they need to pass before they are a legit title contender.
        North Dakota
        National Champions: 1959, 1963, 1980, 1982, 1987, 1997, 2000, 2016

        Comment


        • #64
          Re: 2012 NCAA Tournament: Bracketology

          Interesting fact. The Gophers have never won their conference the same year they won an NCAA national championship. They have always gotten hot at the right time at the end of the season. This Gopher team has the talent to win it all, they just need to be more consistent (as previously mentioned).

          WCHA Records of the National Championship Teams
          1974 14-9-2 2nd place
          1976 18-13-1 3rd place
          1979 20-11-1 2nd place
          2002 18-7-3 3rd place
          2003 15-6-7 2nd place

          Comment


          • #65
            Re: 2012 NCAA Tournament: Bracketology

            Originally posted by Bakunin View Post
            I wouldn't bet on MN to advance out of their region unless they figure out how to do something other than split every effing weekend.
            With MN hosting, I see very little chance of them being upset before the Frozen Four.
            That community is already in the process of dissolution where each man begins to eye his neighbor as a possible enemy, where non-conformity with the accepted creed, political as well as religious, is a mark of disaffection; where denunciation, without specification or backing, takes the place of evidence; where orthodoxy chokes freedom of dissent; where faith in the eventual supremacy of reason has become so timid that we dare not enter our convictions in the open lists, to win or lose.

            Comment


            • #66
              Re: 2012 NCAA Tournament: Bracketology

              The mid-season rationales for predicting season's-end NCAA tournament brackets are based upon mid-season teams' records - ignoring the possibility that prejudice could play a part in such predictions. In the past five or ten years how close a correlation has there been between best performing mid-season teams and NCAA playoff seeds?

              Comment


              • #67
                Re: 2012 NCAA Tournament: Bracketology

                Originally posted by Osorojo View Post
                The mid-season rationales for predicting season's-end NCAA tournament brackets are based upon mid-season teams' records - ignoring the possibility that prejudice could play a part in such predictions. In the past five or ten years how close a correlation has there been between best performing mid-season teams and NCAA playoff seeds?
                Obviously anything could happen, but given the system used to pick the teams is objective and chances of things happening can be predicted based upon the KRACH predictor, probably about 75% of the time you don't see a correlation is when there is an auto-bid involved.

                Comment


                • #68
                  Re: 2012 NCAA Tournament: Bracketology

                  Originally posted by FlagDUDE08 View Post
                  Obviously anything could happen, but given the system used to pick the teams is objective and chances of things happening can be predicted based upon the KRACH predictor, probably about 75% of the time you don't see a correlation is when there is an auto-bid involved.
                  More to the point... what do we have that could be better? Not much. Its not a perfect situation and if there are changes in chemistry etc that can't be handled in a model (especially with sparse data)... but, its often better than going on a straight hunch.
                  BS UML '04, PhD UConn '09

                  Jerseys I would like to have:
                  Skating Friar Jersey
                  AIC Yellowjacket Jersey w/ Yellowjacket logo on front
                  UAF Jersey w/ Polar Bear on Front
                  Army Black Knight logo jersey


                  NCAA Men's Division 1 Simulation Primer

                  Comment


                  • #69
                    Re: 2012 NCAA Tournament: Bracketology

                    Originally posted by Patman View Post
                    More to the point... what do we have that could be better? Not much. Its not a perfect situation and if there are changes in chemistry etc that can't be handled in a model (especially with sparse data)... but, its often better than going on a straight hunch.
                    And it certainly beats not being able to be mostly accurate because of committee votes (like some other sports out there ).

                    Comment


                    • #70
                      Re: 2012 NCAA Tournament: Bracketology

                      Originally posted by FlagDUDE08 View Post
                      Obviously anything could happen, but given the system used to pick the teams is objective and chances of things happening can be predicted based upon the KRACH predictor, probably about 75% of the time you don't see a correlation is when there is an auto-bid involved.
                      Thanks, I neglected to account for the auto-bid team(s)! Still, the chance of "things happening" - like the second half of the season - bemuse me when I see predictions made at or before the season's halfway over. I'm not sure I understand the accuracy of the KRACH predictor. Does it ( at midseason) correctly predict the NCAA tournament seeding of non-autobid teams only 25% of the time, or is it correct 75% of the time?

                      Comment


                      • #71
                        Re: 2012 NCAA Tournament: Bracketology

                        Originally posted by Osorojo View Post
                        The mid-season rationales for predicting season's-end NCAA tournament brackets are based upon mid-season teams' records - ignoring the possibility that prejudice could play a part in such predictions. In the past five or ten years how close a correlation has there been between best performing mid-season teams and NCAA playoff seeds?
                        I have the pairwise data since 2003, taken at key points throughout the season, and mid-January teams in the Pairwise end up in the field 79.4% of the time. The top 8 seeds have made the tournament 64 of 72 times (88.9%) and no #1 or #2 in the mid-January rankings has ever failed to make the field. Of the top 4 overall seeds, only 2010 UMD (#3) and 2004 St Cloud (#4) have failed to make the field.

                        Comment


                        • #72
                          Re: 2012 NCAA Tournament: Bracketology

                          Originally posted by Osorojo View Post
                          Thanks, I neglected to account for the auto-bid team(s)! Still, the chance of "things happening" - like the second half of the season - bemuse me when I see predictions made at or before the season's halfway over. I'm not sure I understand the accuracy of the KRACH predictor. Does it ( at midseason) correctly predict the NCAA tournament seeding of non-autobid teams only 25% of the time, or is it correct 75% of the time?
                          KRACH has nothing to do with tournament teams. The only thing KRACH does is to help predict, based upon past performances, who is the most likely to win a specific game. If you'd like information on tournament teams, PWR is the way to go. Obviously the stats change from game to game, but unless some team really goes to heck or a team around the bubble really starts to get hot, not much changes. I don't know what the percentages are, though. The "bubble" should be very large at this point since we still have two months worth of games to be played (playoffs DEFINITELY count toward the PWR), but it will get smaller the closer we get to the end of March.

                          Comment


                          • #73
                            Re: 2012 NCAA Tournament: Bracketology

                            College Hockey Weekly has their Bracketology up...

                            http://www.collegehockeyweekly.com/C..._Pairwise.html

                            Also, here is their forecast of what the tournament will look like...

                            http://www.collegehockeyweekly.com/T...-Forecast.html
                            North Dakota
                            National Champions: 1959, 1963, 1980, 1982, 1987, 1997, 2000, 2016

                            Comment


                            • #74
                              Re: 2012 NCAA Tournament: Bracketology

                              Originally posted by Priceless View Post
                              I have the pairwise data since 2003, taken at key points throughout the season, and mid-January teams in the Pairwise end up in the field 79.4% of the time. The top 8 seeds have made the tournament 64 of 72 times (88.9%) and no #1 or #2 in the mid-January rankings has ever failed to make the field. Of the top 4 overall seeds, only 2010 UMD (#3) and 2004 St Cloud (#4) have failed to make the field.
                              I wrote an article (When to start looking at PWR) a couple years ago that tackled this from a different angle -- I looked at the average week-to-week movement in PWR and the average deviation of each week's ranking from the final rankings. The latter had this to say about the January PWR, "each team has been ranked an average of 3-8 slots differently on April 1 than in January over the past six years." That fits pretty well with your statistic that the top 8 in January almost always make it, and reminds us that there is still hope for those teams in the mid-20s.

                              Comment


                              • #75
                                Re: 2012 NCAA Tournament: Bracketology

                                Originally posted by Fighting Sioux 23 View Post
                                Also, here is their forecast of what the tournament will look like...

                                http://www.collegehockeyweekly.com/T...-Forecast.html
                                5 WCHA teams making it?!?!?
                                Lammy is my daddy: Co-Founder of the Philippe Lamoureux Fan Club

                                IALTO

                                Comment

                                Working...
                                X