« BACK  |  PRINT

RS

MEMBER DIARY

Can we now stop worrying about PA Senate?

By: Dave_in_Fla (Diary)  |  October 26th, 2010 at 03:51 PM  | 

RESIZE: AAA

I posted yesterday a quick analysis of the PA Senate race and the supposed tightening of the polls there. The key takeaway was the results of the Morning Call/Muhlenberg daily tracking poll, which I still question because I’m not confident they have their likely voter model correct. But the trend was enlightening:

Sestak 44 – Toomey 41
Sestak 43 – Toomey 43
Sestak 43 – Toomey 43
Sestak 42 – Toomey 45
Sestak 42 – Toomey 47

Well, since it is a daily tracking poll, they released another data point today:

Sestak 40 – Toomey 48

The trend is clear and not good for the Democrat. His low 40 support is eroding, most likely because his base is vacillating on whether to vote or not. I doubt anyone is changing their mind and thinking of voting for Toomey, they are just evaluating whether or not to actually vote.

All the usual caveats, Toomey isn’t over 50 yet, it is only one poll, we still need to GOTV.

But no need to panic, this race is well in hand.

Tags: ,

COMMENTS

POSTING RULES  |  LOGIN TO COMMENT  |  REGISTER TODAY
  • izoneguy

    POLLS mean NOTHING…..

    Voting is what will make a difference.
    AND then staying engaged after the election.

    • The_Rebel

      Dave put in the usual caveats, and the GOTV needs. He’s only telling us not to panic because some college polls or Dem leaning PPP polls say otherwise. Let’s not fall into their push poll and bad sample poll traps.

      • Dave_in_Fla

        One of the quibbles I have with Neil’s modeling methodology (and Nate Silver and RCP average) is that a number of polls are generated with a specific spin based on the results desired.

        Now Neil is nice in assuming it is a bad poll, but over the years I have noticed that there never seems to be a bad poll that messes up the averages in the GOP direction, it is always for the Democrats. It is almost as if someone knows that getting two polls into an RCP average will make a Dem trailing by 6, suddenly be trailing by 3 and inside the MOE.

        So this year we appear to have the normal “all bad polls that mess up averages tilt toward the Democrat” and a new factor that indicates that the turn out models for the LV screens are over sampling Democrats. A double dip error.

        Even so, the GOP is still leading in most of the polls, but the Dem is made to look competitive. I don’t think they are, I think the only two competitive races are California and Washington, and I’m not even convinced about Washington.

        Right now, O’Donnell could win and I would not be unsurprised.

        Anyway, we still need to execute, but we don’t need to panic. We are positioned to win, just 7 laps to go, don’t blow a tire.

        • GregInFla

          • GregInFla

            I thought I made it smaller but something went amiss. Sorry about that. I made object width/height correct.

          • http://908StraightSt.wordpress.com/ mbecker908

            if you’ll get me your email address I’ve got some emails in my spam folder that can help out. Well, that assumes you want to make it bigger.

          • http://moelane.com/ Moe Lane

            You have to put it in twice, though. Or just go to the YT page and punch it in there.

          • GregInFla

            The last set of height/width seemed to reference a full-size object, so I was not sure if that one needed changing. Obviously, it did. I’ll do better next time.

  • http://www.hakubi.us/ Neil Stevens

    There’s no need to panic, and I’m rating the state well for Toomey, but the work must go on until we see through the win.