Welcome to the Hardcore Husky Forums. Folks who are well-known in Cyberland and not that dumb.

We are down to 32% now

2

Comments

  • YellowSnow
    YellowSnow Moderator, Swaye's Wigwam Posts: 37,689 Founders Club

    http://fivethirtyeight.com/features/how-a-weekend-of-upsets-changed-the-college-football-playoff/
    The season is definitely not over, but our statistical odds to make the CFP per Nate Silver's super computer have decreased from 47% pre USC to 32% post said game. In my view, Nate's stock is not diminished post Trump win. He gave him a 29% chance as of Monday, 11/7 and indicated an electoral college win was well within the polling margin for error.

    Yes, it's FS that I even looked this up - just win El Norte and worry about the other crap when we get there.

    silver had the election nailed so im sure he's right on this one.
    I hardly think he had it "nailed" as was the case in 2008 and 12, but he gave him a lot better chance at victory than most predictors.
  • PostGameOrangeSlices
    PostGameOrangeSlices Member Posts: 27,658
    Louisville sucks. Clemson sucks. Michigan sucks
  • doogie
    doogie Member Posts: 15,072
    This is the guy who gave a 67% chance to Hillary day before the election, right?
  • doogie
    doogie Member Posts: 15,072

    doogie said:

    This is the guy who gave a 67% chance to Hillary day before the election, right?

    Yeah. What's the problem?

    Even Trump's camp didn't think they were favorites to win. 538 gave Trump a better chance of winning than pretty much every polls-based website out there and also better than the betting market. People who think Trump's win is a reason to ignore 538 are clueless.
    Trump knew he was going to win.
  • RaceBannon
    RaceBannon Member, Moderator, Swaye's Wigwam Posts: 115,408 Founders Club

    doogie said:

    This is the guy who gave a 67% chance to Hillary day before the election, right?

    Yeah. What's the problem?

    Even Trump's camp didn't think they were favorites to win. 538 gave Trump a better chance of winning than pretty much every polls-based website out there and also better than the betting market. People who think Trump's win is a reason to ignore 538 are clueless.
    I'm not sure that giving the guy who won an electoral landslide 33% is reason to celebrate

    As someone else pointed out Silver never changed his models to reflect the Trump voters and the lack of carry over of Obama voters for Hillary. He missed on Trump on the primaries too

    It doesn't make him a bad person or bad at what he does but if we wanted stupid percentage metric bullshit we'd move to Friday Harbor with Chest

    He was as bad as everyone else
  • jecornel
    jecornel Member Posts: 9,737
    Silver should have found a new career last Wednesday morning. Lost all credibility. LOSER.
  • AIRWOLF
    AIRWOLF Member Posts: 1,840
    jecornel said:

    Silver should have found a new career last Wednesday morning. Lost all credibility. LOSER.

    You probably don't have a very good grasp of probability.
  • phineas
    phineas Member Posts: 4,732

    That's bullshit along with most of the crap that comes out of there

    I had Trump winning in June of 2015. Fuck Silver.

    What were his odds on my Cavs down 3-1?

    We win we are in and we have 2 cupcakes and a likely rematch with Utah to do so

    81% chance

    BLACK POWER
  • BuffBuffPass
    BuffBuffPass Member Posts: 326

    doogie said:

    This is the guy who gave a 67% chance to Hillary day before the election, right?

    Yeah. What's the problem?

    Even Trump's camp didn't think they were favorites to win. 538 gave Trump a better chance of winning than pretty much every polls-based website out there and also better than the betting market. People who think Trump's win is a reason to ignore 538 are clueless.
    I'm not sure that giving the guy who won an electoral landslide 33% is reason to celebrate

    As someone else pointed out Silver never changed his models to reflect the Trump voters and the lack of carry over of Obama voters for Hillary. He missed on Trump on the primaries too

    It doesn't make him a bad person or bad at what he does but if we wanted stupid percentage metric bullshit we'd move to Friday Harbor with Chest

    He was as bad as everyone else
    You have a really poor grasp on how probabilities work, particularly when it comes to events that have a high degree of correlation, i.e. non-independent events.

    And he did change his model to reflect the uncertainty in this election, which is why his model gave Trump a better chance of winning than virtually every poll-based model out there.

    Many of the arguments used about turnout were used back in 2008 and 2012 and were categorically wrong back then. There is such a thing as polling error and polls have become less reliable in recent years by the look of things (Brexit, Israeli election, British election, etc.) which, again, is why he gave Trump a better chance of winning than other models out there.