Wednesday, October 6, 2010

Silver smacks down Strickland's hopes and dreams...

Below we mentioned Nate Silver's update of the probability of a Kasich win. The New York Times numbers guru showed our GOP candidate currently enjoying an 85.8% chance of victory in a little less than four weeks.

But earlier this afternoon Silver added a bit of commentary to the state of the race that's going to disappoint some Democrats.
Some polling in Ohio last week had suggested that the Democratic incumbent, Ted Strickland, was pulling into a rough tie with his Republican opponent, John Kasich, spontaneously overcoming what had been about a 10-point deficit.

The model, however, was cautious about taking those numbers quite at face value. And this week, that looks prudent, since three new polls have come out showing Mr. Kasich still holding an advantage of 6 to 9 points. Obviously, Mr. Strickland could still win ..... But the lesson is not to jump to conclusions. Especially at this stage of the campaign, things like changes in advertising spending patterns can touch off short-lived bounces in the polls.
It seems that Strickland's massive surge in spending on commercials only paid off for a short time. And as Silver states, the new polls this week confirm that.

Kasich will be smart to do everything he can to keep Strickland's level of support as flatlined as it has been all year. As long as Strickland stays decently below the magic 50% number, and Silver's best analysis of that support shows it at 41.5%, undecideds will do what they always do, break towards the challenger. That goes even moreso in a wave election like the one we're about to experience.

27 days, people. Keep working.

6 comments:

  1. Oh, and Silver ALSO updated his projected election night forecast (head to head numbers). The new update not only narrows Kasich lead, but it puts the race within the margins of error.

    But, you're right, no need to panic for Kasich at all. Seeing a candidate lose half his lead in multiple polls weeks before the election day is totally expected... LOL.

    ReplyDelete
  2. Um. Yeah. I included it all below.

    And I don't think you understand how his election night forecast number works. That isn't a number designed to be within a margin of error. It's based on a formula he uses that utilizes poll numbers and other factors to predict the final result. He thinks it will be 6.

    I don't care if it's 1. A win is a win.

    One question for you, are you comfortable with him only improving your chances of victory by 1% to 14.2% after all those polls came out?

    ReplyDelete
  3. Um.... he says on the chart +/- 6. That's what those SHADED regions on the graphs mean.

    Geez, you know this, don't you?

    You're also overstated (as Silver himself wrote this week) what those odds mean.

    Strickland's odds are improving every day. The aggregate is within four. The only reason it went up is Rasmussen did another poll for Fox News and you had another Quinny which showing more favorable results for Kasich than ANY other pollster is coming close to showing.

    The race is tightening at a pace that Strickland can win by Election Day.

    ReplyDelete
  4. Still nothing on the 85.8% chance of Kasich winning?

    I mean, even after all those favorable polls, it still only improved 1% point in a week.

    Too bad for you this election isn't next June.

    ReplyDelete
  5. Um, I did address the "85%" figure. I said that Nate Silver says you're overestimate the significance of that figure.

    BTW, I take that to mean you acknowledge that I was right that Silver's election night projections are even MORE within the margin of error now for than they were a week ago?

    Or are you still insisting that Silver's graph does not disclaim a 6-point margin of error?

    ReplyDelete
  6. Link to him saying not to take his own prediction at face value?

    He seems to be confirming the value of his model here:

    The model, however, was cautious about taking those numbers quite at face value. And this week, that looks prudent, since three new polls have come out showing Mr. Kasich still holding an advantage of 6 to 9 points. Obviously, Mr. Strickland could still win ..... But the lesson is not to jump to conclusions. Especially at this stage of the campaign, things like changes in advertising spending patterns can touch off short-lived bounces in the polls.

    ===

    But yes, I do submit that his forecast is within the margin of error. I hope that makes you feel better. Unfortunately for you, it only improved from a 5.9 gap to a 5.6 gap in one week.

    At that rate, by election day that may be down to a 4.7 prediction in Kasich's favor. You comfortable with that?

    ReplyDelete

No profanity, keep it clean.

Note: Only a member of this blog may post a comment.