1. Welcome! Please take a few seconds to create your free account to post threads, make some friends, remove a few ads while surfing and much more. ClutchFans has been bringing fans together to talk Houston Sports since 1996. Join us!

NYT Nate Silver Doubles Down: 75% Chance Obama Wins

Discussion in 'BBS Hangout: Debate & Discussion' started by MojoMan, Oct 30, 2012.

  1. GladiatoRowdy

    GladiatoRowdy Contributing Member

    Joined:
    Oct 15, 2002
    Messages:
    16,596
    Likes Received:
    494
    Fixed it for you.
     
  2. jopatmc

    jopatmc Contributing Member

    Joined:
    Sep 4, 2002
    Messages:
    15,368
    Likes Received:
    387

    Landslide is a very subjective term. He's trying to par away from landslide and 92%.



    My point all along is Nate Silver's predictor model for presidential elections is useless. Presidential elections happen every 4 years. They are unique singular events. They're not playing 162 games a year for the next 100 years. The data sets are much smaller and much more subject to error. Doesn't work. And that's why his model is way off.
     
  3. SamCassell

    SamCassell Contributing Member

    Joined:
    Feb 14, 1999
    Messages:
    8,841
    Likes Received:
    1,253
    I think that I've finally seen the one reason to want you to stick around the board. Clutch could use a good laugh. I'm thinking the "idea" is a board dedicated to making asinine predictions (Mitt Romney, Dwight Howard, other supposed "sure things") and then claiming that whoever made said prediction was ballsy, rather than idiotic.
     
  4. jopatmc

    jopatmc Contributing Member

    Joined:
    Sep 4, 2002
    Messages:
    15,368
    Likes Received:
    387

    No, wrong. I'll disclose it at a given time.
     
  5. Cannonball

    Cannonball Contributing Member

    Joined:
    Jul 27, 2006
    Messages:
    21,649
    Likes Received:
    1,907
    To get to 321, Romney would have to take Florida, N. Carolina, Virginia, Colorado, Ohio, Pennsylvania, Michigan, Minnesota, and Iowa. All but 2 of those states are polling in favor of Obama. Ain't no way in hell.
     
  6. Major

    Major Member

    Joined:
    Jun 28, 1999
    Messages:
    41,383
    Likes Received:
    15,808
    These two statements, less than 3 hours apart, are amusing together.

     
    1 person likes this.
  7. Cannonball

    Cannonball Contributing Member

    Joined:
    Jul 27, 2006
    Messages:
    21,649
    Likes Received:
    1,907
    According to the "scenario analysis" part of his site, it looks like he defines landslide as a double digit popular vote win. He has that as a 0.3% change of happening for Obama and <0.1% for Romney.

    If you look at his numbers, Nate predicts a close race in the popular vote and in most of of the battleground states as well. It's just that 2-3% point leads in these states tend to hold on election day.
     
  8. jopatmc

    jopatmc Contributing Member

    Joined:
    Sep 4, 2002
    Messages:
    15,368
    Likes Received:
    387


    It's a useless model.



    Nate is predicting (100%) an Obama victory. He made a bet with Joe Scarborough (friendly bet but still a wager), that Obama wins. That's how much he believes in his model.

    There's no way you can have Obama at 92% and then say the election is going to be a close election. He's in effect saying all the states will be close but Obama is on top of any combination of states to get to 270. If it's going to be a close election, then you have to ratchet down those odds to closer to 50/50. Remember, close doesn't count. A winner is a winner and a loser is a loser. Doesn't matter if it is 1 vote or a million votes.

    His predictor model is useless for presidential election purposes.
     
  9. hairyme

    hairyme Member

    Joined:
    Jan 22, 2011
    Messages:
    691
    Likes Received:
    93
    Notice how your argument keeps boiling down to "Well these projections can't be right... because I think Obama sucks." How is this remotely relevant to the topic at hand?! Obama could be hot poop on a stick, and if these previously validated and well-respected projections (Silver's that is) predict his victory, then that's just how the math shakes out regardless of the candidate's hypothetical crappiness!

    Also, you keep balking at the high percentages. I don't fault people for not knowing math and stats, but if you are in that category, you shouldn't try to have a discussion on it... much less question its validity! I don't remember exactly what Silver's method is, but it's basically a Monte Carlo simulation if I recall correctly. The current 92% figure does NOT measure strength of victory, but rather indicates certainty or likelihood.

    For example, if the simulation is run 100 times and you get:
    - Obama narrowly defeating Romney 92 times.
    - Romney strongly defeating Obama 8 times.

    This is an extreme example, but the end result would be Obama with a 92% chance of victory. That's all this means. And NO, those 8 times aren't a built-in excuse for Silver when Romney mops the floor with Obama today (in your world)!

    But alas, none of this matters. This has been explained to you and your ilk repeatedly in this thread, but you are far too delusional to understand.
     
    1 person likes this.
  10. BigBenito

    BigBenito Member

    Joined:
    Sep 5, 2002
    Messages:
    7,355
    Likes Received:
    175
    OK, adding 'does not know how gambling works.' to your tally.

    At the time he had Obama at 80%. Anyone. And I mean anyone who knows how gambling/statistics works - would take a straight up bet with an 80% chance of winning.
     
  11. Rashmon

    Rashmon Contributing Member

    Joined:
    Jun 2, 2000
    Messages:
    19,153
    Likes Received:
    14,303
    [​IMG]
     
  12. weslinder

    weslinder Contributing Member

    Joined:
    Jun 27, 2006
    Messages:
    12,983
    Likes Received:
    291
    Would Nate Silver give me 11.5-to-1 (92-to-8) odds for a Romney win. Hell, I'll take 10-to-1.
     
  13. jopatmc

    jopatmc Contributing Member

    Joined:
    Sep 4, 2002
    Messages:
    15,368
    Likes Received:
    387

    A Monte Carlo simulation with a data set that is 80% in error is a useless Monte Carlo simulation.
     
  14. BigBenito

    BigBenito Member

    Joined:
    Sep 5, 2002
    Messages:
    7,355
    Likes Received:
    175
    He was not offering odds, sorry.
     
  15. jopatmc

    jopatmc Contributing Member

    Joined:
    Sep 4, 2002
    Messages:
    15,368
    Likes Received:
    387

    He's not offering odds. He is saying the odds are 9.2 to 0.8 that Obama wins.
     
  16. BigBenito

    BigBenito Member

    Joined:
    Sep 5, 2002
    Messages:
    7,355
    Likes Received:
    175
    FOR THE BET. try to follow along.
     
  17. Deckard

    Deckard Blade Runner
    Supporting Member

    Joined:
    Mar 28, 2002
    Messages:
    56,793
    Likes Received:
    39,077
    I tried to get basso to make a bet with me. If Romney wins, I won't start a thread in D&D for a year. If Obama wins, he doesn't start a thread in D&D for a year. He wouldn't do it. He wouldn't take the bet. The fellow is a troll, without a shred of conviction in what he's babbling here.
     
  18. ChievousFTFace

    ChievousFTFace Contributing Member

    Joined:
    Jan 25, 2010
    Messages:
    2,794
    Likes Received:
    560
    You really don't understand statistics do you? If this was going to be a landslide, it would be >99.99% Obama.

    Of course it could be a close election and still have a heavy favorite. This is what happens when you run multiple simulations using data. You aren't even worth explaining to due to the fact that you said you'd doubt his model even if he was accurate.

    Betting sites are giving you 3 to 1 on your money if you are so confident in Romney.
     
  19. ItsMyFault

    ItsMyFault Contributing Member

    Joined:
    Jan 28, 2009
    Messages:
    15,645
    Likes Received:
    978
    Such cheating.

    <iframe width="420" height="315" src="http://www.youtube.com/embed/QdpGd74DrBM" frameborder="0" allowfullscreen></iframe>
     
  20. Rashmon

    Rashmon Contributing Member

    Joined:
    Jun 2, 2000
    Messages:
    19,153
    Likes Received:
    14,303
    [​IMG]
     
    2 people like this.

Share This Page

  • About ClutchFans

    Since 1996, ClutchFans has been loud and proud covering the Houston Rockets, helping set an industry standard for team fan sites. The forums have been a home for Houston sports fans as well as basketball fanatics around the globe.

  • Support ClutchFans!

    If you find that ClutchFans is a valuable resource for you, please consider becoming a Supporting Member. Supporting Members can upload photos and attachments directly to their posts, customize their user title and more. Gold Supporters see zero ads!


    Upgrade Now