October 24, 2006
Sprint to the Finish Rises From Its Grave
The previous three predictive posts on the upcoming election are here:
It's been a while (ten days) since my last prediction on October 14th; now that the election is just two weeks away, it's high time I got the leg out and updated myself.
You may recall (but I highly doubt it, so I'll repeat it, and to heck with the lot of you!) that last time, the Big Lizards bottom line was a Democratic pick-up of 12 in the House and 4 in the Senate, although I didn't clearly articulate that prediction until the next day, with Bride of Sprint to the Finish:
Big Lizards has been bucking the tide of Republican defeatism, predicting that the GOP will limit their losses to 12 in the House and 4 in the Senate, retaining both houses.
Today, I went through every last congressional race on Real Clear Politics' "election pages," making my determination how each would go... and I have favorable revisions since last time.
Let's start with the House. I consider the following races actually in play; the rest are pretty clearly holds for the incumbent (mostly Republican but a few Democrats). Note that I call it a hold if the Republican is ahead in the most recent round of polling or if the Democrat is ahead by only 2-3 points; I believe the GOP ground-game will make up for that small an edge, thus they'll hold such seats.
Districts in bold blue are pretty certain Democratic pick-ups; boldface districts are probable pick-ups; and italicized districts are potential pick-ups. There are no Republican pick-ups, not even potentially... unless things change rather drastically in the next fortnight.
Here is the list, sorted by alphabetically by district and ignoring the holds:
This works out to 2 certains, 7 probables, and 8 possibles. I gave the certains to the Democrats; and with the possibles (toss-ups), I gave the Democrats half. That makes a core of 6 pick-ups.
For the probables, I calculated them two different ways: with the Dems picking up 2/3rds of them (4-5 pick-ups), and with the Dems picking up 3/4ths of them (5-6 pick-ups). Thus, the total range is from a low of 10 to a high of 12 Democratic pick-ups. Thus, hitting right in the middle, Big Lizards is now prepared to predict a Democratic pick-up of 11 seats in the House, leaving the Republicans with a slim majority of 221 to 214.
Now to the Senate side of Congress. I'll mention each race here, because there are only thirteen that anybody cares about. The name is the name of the candidate from the incumbent party, whether that candidate is himself the incumbent or not.
This time, boldface means a certain pick-up (winner indicated by color); italics means probable pick-up; and color but no italics or bold means the party that currently holds that seat retains it. True toss-ups -- three of them -- are indicated by using italics with normal text color:
- AZ-John Kyl (R): Republican hold;
- MD-Ben Cardin (D): Toss-up; [I call MD a toss-up because it all hinges on the percent of the black vote that Michael Steele gets; if he gets 25%-30%, he wins.]
- MI-Debbie Stabenow (D): Toss-up; [I call this a toss-up because of the recent announcement by Ford that they're laying off many thousands of workers; if subsequent polling shows that voters aren't holding that against incumbent Sen. Debbie Stabenow, I'll move this to Democratic hold -- though it makes no difference to the bottom line.]
- MN-Amy Klobuchar (D): Democratic hold;
- MO-Jim Talent (R): Republican hold;
- MT-Conrad Burns (R): Republican hold;
- NJ-Robert Menendez (D): Toss-up;
- OH-Mike DeWine (R): certain Democratic pick-up;
- PA-Rick Santorum (R): certain Democratic pick-up;
- RI-Lincoln Chafee (R): probable Democratic pick-up;
- TN-Bob Corker (R): Republican hold;
- VA-George Allen (R): Republican hold;
- WA-Maria Cantwell (D): Democratic hold;
As you can see, assuming you can figure out my cockamamie type-decoration code, we have 2 certain Democratic pick-ups (Pennsylvania and Ohio); we also have one probable Democratic pick-up.
But the three toss-ups are all seats currently held by Democrats. The rest are holds, either Repubilcan or Democrat.
So let's use the formula to estimate Democratic pick-ups: 2 (certain), plus two-thirds or three-fourths of the lone probable, which makes 3; but for the toss-ups, we get either 1 or 2 Republican pick-ups. Which means we would predict a Democratic pickup of 1 to 2 seats; let's be conservative and say a pickup of 2.
But I don't have as much confidence in my back of the thumb guesstimate for the Senate races as I do for the House races (since there are fewer of the former); thus, I'm going to hedge and say a Democratic pick-up of 3. (Basically, I doubt whether the Republicans can really nab any of those three toss-ups.)
New Big Lizards bottom line: we predict a Democratic pick-up of 11 in the House and 3 in the Senate, in each case one fewer than we predicted last time.
I will revisit my predictions again in a week, and then one more time on Election Day itself. Let's see how close we come!
If you're curious why others, such as Election Projection, predict higher numbers of pick-ups for the Democrats, there is an easy answer: most of the other sites look at a race like, say, Montana Senate -- where Republican Conrad Burns is running 3 points behind on the last two polls -- and they simply project that forward in a straight line and predict he will lose. But 3% is within the margin of error of each poll; that means the polls are really saying nothing more than "they're neck and neck."
With a tie two weeks before the election, I conclude the challenger has not overcome the hurdle of incumbency; Jon Tester hasn't made the sale. So with the superior GOP GOTV program, the power of incumbency, the monetary advantage of the Republicans, and the natural tendency of all polling to slightly bias towards Democrats, and where the momentum is in the Republican's favor -- I instead believe that such a razor-close race will go to the Republican.
That is, Election Projection, et al, are telling you what the result might be if (a) the election were held today, instead of two weeks from today, and (b) assuming the Republicans have no advantage because of their superior ground game, which even the Democrats admit they do.
So here we are. (I like that better than "there we are," since I needn't explain where "there" is; "here" is obvious.)
Hatched by Dafydd on this day, October 24, 2006, at the time of 10:00 PM
TrackBack URL for this hissing: http://biglizards.net/mt3.36/earendiltrack.cgi/1389
Listed below are links to weblogs that reference Sprint to the Finish Rises From Its Grave:
» Revenge of Sprint to the Finish from Big Lizards
The previous four predictive posts on the upcoming election are here: Sprint to the Finish Addendum to Sprint to the Finish Bride of Sprint to the Finish Sprint to the Finish Rises From Its Grave In our last fit in... [Read More]
Tracked on November 1, 2006 7:48 AM
» Addendum to the Sprint from Big Lizards
The previous five predictive posts on the upcoming election are here: Sprint to the Finish Addendum to Sprint to the Finish Bride of Sprint to the Finish Sprint to the Finish Rises From Its Grave Revenge of Sprint to the... [Read More]
Tracked on November 2, 2006 4:50 AM
» Addendum 3 to Sprint: Will the Sleeper Awake? from Big Lizards
The previous six predictive posts on the upcoming election are here: Sprint to the Finish Addendum to Sprint to the Finish Bride of Sprint to the Finish Sprint to the Finish Rises From Its Grave Revenge of Sprint to the... [Read More]
Tracked on November 2, 2006 2:46 PM
The following hissed in response by: Rovin
If, When these predictions come to fruition Dafydd, the conservative blogosphere should anoint you to Carnak, the Reptillian Visionary of sanity.
And, of course, plan for a major meltdown of the left. "We wuz robbed again" will be the headlines far into the spring.
The above hissed in response by: Rovin at October 25, 2006 7:12 AM
The following hissed in response by: yetanotherjohn
I would like to quibble with the reptile (no inferences on how the NJ courts are going to rule). A poll showing a difference within the margin of error is not saying that it is a statistical heat. Each point away from a tie indicates an increasing percentage chance that the direction of the election is in accordance with the poll. So for a margin of error of 3, a candidate showing a lead of 3 is more likely really ahead than a candidate with a lead of 2 who in turn would be more likely ahead than a candidate with a lead of 1. Gerry Daly had a chart with the percentages of each step, but alas and alack his site is no more a bloggin. As another aside, for a poll with a margin of error of 3, a candidate with a 4 point lead only has a 95% chance of actually being in the lead. Pollsters recognize that theirs is an inexact science (as opposed to alchemy where they would always be able to turn lead into gold if they just get that elixer formula right) and that 1 out of 20 polls will just be totally wrong.
The above hissed in response by: yetanotherjohn at October 25, 2006 7:40 AM
The following hissed in response by: jp phish
The latest Strategic Vision Poll (came out yesterday) on Michigan has this result:
If the election for United States Senate were held today, whom would you vote for Debbie Stabenow, the Democrat or Mike Bouchard, the Republican?
Debbie Stabenow 48%
Mike Bouchard 42%
Bouchard has a get-it-done reputation, is tough on crime and border security (we are on the canadian border), is an ex leader in the state legislature, and is currently sheriff of Oakland county. He is wildly popular in Oakland and next-door Macomb Counties. Remember Macomb? Home of the Reagan Democrats?
Post a comment
Thanks for hissing in, . Now you can slither in with a comment, o wise. (sign out)(If you haven't hissed a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Hang loose; don't shed your skin!)
© 2005-2009 by Dafydd ab Hugh - All Rights Reserved