Blue Mass Group

Reality-based commentary on politics.

Baker’s own internal poll shows his campaign collapsing

By

Our friends at RMG report that Charlie Baker’s internal pollster shows the race as a dead heat at 40-40.

This is, of course, terrible news for Team Baker. Why? Two reasons. One, the very same pollster had Baker at +7 two weeks ago, as the Bakerbots announced with great fanfare in an effort to counter news of real other polls showing Baker down 6-8 points. Second, Nate Silver, who has forgotten more about polling than most people ever knew, says that an internal poll typically favors its candidate by about six points. If you figure that into today’s announcement by Team Baker, that would put Patrick ahead by about 6 points – exactly where he’s been for months.

What would be really interesting to know is whether Baker’s support is collapsing to the extent today’s announcement suggests, or whether that first one was a total outlier. Either way, bad news for Team Baker.  

Please share widely!
fb-share-icon
Tweet
0

Comments

  1. Serious question: When was the last time you saw a winning candidate release internal polling?

    • But down 7 points from 2 weeks ago according to the same internal pollster?  That’s collapsing.  Any way you look at this, it can’t be good news for Team Baker.

    • When your own internal polling shows a 7-point lead evaporating in little over a week, I’d call that collapsing.  And these are the numbers they’ve chosen to put in a press release.

  3. I’ve heard rumors that the new Suffolk poll tells the same story we’ve seen for months – Deval up by around 7.  

    <

    p>Has anyone actually seen the poll yet to confirm?

  4. MEMORANDUM
    TO:                  BAKER FOR GOVERNOR CAMPAIGN
    FROM:            NEIL NEWHOUSE B PUBLIC OPINION STRATEGIES
    SUBJECT:      RECENT BAKER POLL DATA
    DATE:            OCTOBER 28, 2010
    We just came out of the field with a 600 sample of likely voters over the last two nights (October 26-27) for the Baker campaign.  The poll has a margin of error of +/- 4% in 95 out of 100 cases.

    <

    p>Key Findings and suggestions

    <

    p>1.      9% say they would be less likely to vote for a candidate if that candidate also supported someone who tried to cover up a “sexual assault”  Only 30% were less  likely to support someone who covered up a “strip search”  (suggestion: keep holding our nose by supporting Jeff Perry by referring to the “strip search” incident)

    <

    p>2.    55% of respondents agree that because Charlie talks about his IVY league back ground a lot he is more electable. (Suggestion: keep mining the Princeton and Dartmouth alumni list)

    <

    p>3.  21% of the women surveyed find Charlie: arrogant and whinny. (Suggestion: Charlie should take Flonase to mask the whinnies)

    <

    p>4. 85% of the respondents said they would rather get fired from their job by Charlie then by any other candidate.

    <

    p>5. When presented with the Charlie’s 75 point plan for turning around state government, only 10% of the respondents could remember more then 4 of the points (Suggestion: we should try to get Charlie to focus only a few issues at a time, and not rattle off 10 things at once. Opps, we’ve been working on that for 18 months.)

    <

    p>

  5. Here’s an important observation.   Baker didn’t release his internal polls.    He released internal memos that describe selective parts of those polls.

    Then, of course, Baker has a problem with memos!

  6. We hear polls from many organizations before the elections but let’s match them up with next Tuesday’s results and brand the good ones and the bad ones. So many times we hear polls and experts give their opinions and they are flat out wrong. Check out the guys who pick football games in the papers, they’re terrible.

    <

    p>The Globe, Herald, Suffolk and many other organizations have done polls recently and by next Tuesday we should know who to trust (or not trust) in the future.

    • …using previous elections, to tell which polsters have a better track record in matching the actual outcome?

    • Note that the Globe and Herald and other media organizations (like 7 News) don’t do the actual polls; they just commission them.

      <

      p>So, for example, if you want to know whether “the Globe’s” polls have been accurate in the past, it’s really asking whether UNH’s Survey Center is accurate, since that’s who the Globe commissions. This is relevant because if the Globe changes who they commission as their pollster over time, comparing the accuracy of “Globe” polls when done by different pollsters is really comparing apples-and-oranges.

  7. that this internal poll overstates the position of the client (as internal polls are liable to do) is that it has Baker’s favorability at +11 (48% favorable to 37% unfavorable).

    <

    p>That’s far higher than all other independent polls have shown. If Baker has an inflated +11 favorability rating in a poll that has him merely tied, then he’s probably down even more than 6 points in reality.

  8. Nate Silver, who has forgotten more about polling than most people ever knew, says that an internal poll typically favors its candidate by about six points. If you figure that into today’s announcement by Team Baker, that would put Patrick ahead by about 6 points – exactly where he’s been for months.

    <

    p>That’s “typical”, but since he’s a statistician by trade, I expect that he’d be hesitant to apply what is something like an average (mean?/median?) to a specific poll in a specific race, particularly where a single third party candidate is polling over 5 percent.

    <

    p>

    <

    p>It’s certainly not good news for Baker, but I don’t think it’s appropriate to extrapolate this specific poll to be a 6 point Patrick margin.

Archives