Sunday, November 7, 2010

Why this race isn't over yet...

Following up on the last post, it occurred to me that this race REALLY isn't over yet.  Based on the new SMOVs and VCRs [see Terminology post] from the five counties that reported yesterday, here's how we can expect their EBOH to be allocated:

Kitsap
14,000 EBOH * Nov. 6 VCR (79.87%) = 11,182 votes
11,182 votes * Nov. 6 SMOV (47.22%) = 5,280 +SAND & 5,902 +WIGG
= net 622 new votes for Wiggins

Pierce
13,500 EBOH * Nov. 6 VCR (78.45%) = 10,591 votes
10,591 votes * Nov. 6 SMOV (53.06%) = 5,620 +SAND & 4,971 +WIGG
= net 649 new votes for Sanders

Snohomish
30,000 EBOH * Nov. 6 VCR (77.36%) = 23,208 votes
23,208 votes * Nov. 6 SMOV (50.84%) = 11,799 +SAND & 11,409 +WIGG
= net 390 new votes for Sanders

Spokane
32,000 EBOH * Nov. 6 VCR (71.43%) = 22,858 votes
22,858 votes * Nov. 6 SMOV (53.80%) = 12,297 +SAND & 10,560 +WIGG
= net 1737 new votes for Sanders

Whatcom
12,000 EBOH * Nov. 6 VCR (73.05%) = 8,766 votes
8,766 votes * Nov. 6 SMOV (51.50%) = 4,514 +SAND & 4,252 +WIGG
= net 262 new votes for Sanders

Given Sanders's current lead of 13,251 (as of 8pm Sunday night), the projected EBOH allocations from these five counties would extend his lead to 15,667. 

Just for kicks, let's add in the remaining three members of the Big Eight [see Terminology post]:
  • Clark: 1,750 EBOH * 77.10% VCR = 1,349 votes * 54.13% (SMOV Nov. 5) = net 111 votes for Sanders.
  • Thurston: 6,000 EBOH * 76.88% VCR = 4,613 votes * 48.35% (SMOV Nov. 4 total) = net 153 votes for Wiggins.
  • King:  See previous post.  We'll assume the low end of what Wiggins should expect, namely 19,773 votes.  (just to make this analysis interesting).
That gets us to a 4,148 net vote lead for Wiggins.  Now, let's factor in EBOH allocations from the five other counties that may be able to significantly impact the numbers: (all using VCR & SMOV totals from Nov. 4)
  • Benton:  7,000 EBOH * 78.68% (VCR) = 5,508 votes * 57.96% (SMOV) = 877 net votes for Sanders
  • Grant:  1,500 EBOH * 76.59% (VCR) = 1,149 votes * 65.34% (SMOV) = 353 net votes for Sanders
  • Okanogan:  4,000 EBOH * 74.10% (VCR) = 2,964 votes * 59.00% (SMOV) = 534 net votes for Sanders
  • Stevens:  6,000 EBOH * 74.62% (VCR) = 4,477 votes * 58.32% (SMOV) = 745 net votes for Sanders
  • Yakima: 10,000 EBOH * 82.75% (VCR) = 8,275 votes * 59.78% (SMOV) = 1,618 net votes for Sanders
Drum roll please.... that's a 21 vote victory margin for Wiggins.  And none of the remaining counties can realistically swing the vote by more than 200 votes each.

So what does this mean?

1.  Anyone who says that this race is over is C-R-A-Z-Y.
2.  Wiggins cannot win this race unless King performs better than the "low end" of the VCR and SMOV scale.  In effect, King now has to offset not only Sanders's current margin but the SMOV gains he will accrue throughout the state.
3.  Given the (somewhat) neutral performance of the other Big Eight counties besides King, it looks like Yakima and some of the rural non-Peninsula counties may hold the key for Sanders.
4.  Aside from King, the single best thing happening for Wiggins right now is the decline in SMOV in Snohomish and Spokane.  If those trends continue -- and if Pierce doesn't completely reverse its SMOV decline since Nov. 4 -- that will help Wiggins substantially.
5.  Once again, anyone who says this race is over is crazy.  We'll know more tomorrow, but don't get too comfortable.

3 comments:

  1. While I appreciate your calculations and model (it's pretty good), and I probably am crazy for calling this race, the answer will be after tomorrow's count.

    Plain and simple, if the trend continues, and Sanders is only up in the mid four figures (as my model projects), he is done.

    One factor to consider, aside from the under-votes (which my model takes into account, just not as detailed as yours) is the total votes outstanding in King County. Currently, the SOS number is below the KC number. Why that is, I confess myself unsure, but if we see another five figure "received" tomorrow, then that will only add to the momentum for Wiggins (most other counties have stopped receiving ballots).

    If memory serves, currently Sanders is only gaining (percentage wise) in three counties, and by minimal tenths of percentages. Wiggins has managed to increase his performance by 1-2% or more in many counties, including Spokane, Pierce, Snohomish, Whatcom, Skagit, etc.

    In the end, it is VERY unlikely that Sanders could win, and it would require an increase in his performance in the remaining ballots to be counted, especially in King, Snohomish, Pierce, Kitsap, Thurston and Skagit Counties. Considering the two October stories about his...issues...along with the trend from Tuesday through Saturday, it is politically improbable at best that the trend would reverse itself.

    ReplyDelete
  2. Hey Michaelp,

    Thanks for stopping by - I'm a big fan of your work, by the way.

    I agree that Wiggins is definitely the beneficiary of the late trend. In fact, without that trend, he wouldn't have a shot (using data only thru Nov. 4th, I had projected the remaining ballots to result in a final margin of 7-8k for Sanders.)

    Interesting comment about the KC EBOH count being higher than the SOS's count. That could indicate a different methodology used to calculate the #s (i.e. SOS only includes validated ballots ready to be counted and unopened ballots; KC includes ALL ballots, including those that have failed verification or are pending voter affidavit). I must admit, I actually haven't seen the separate KC projection of EBOH - can you post a link?

    I agree that Sanders's marginal SMOV gains between Nov. 4-5 in Clark, Ferry, Stevens, Wahkiakum, and Yakima are probably not enough to counteract the shifting percentage gains in the other 23 counties. Those five counties only account for 17,783 EBOH. But, Yakima is the real killer. It's 10,000 EBOH (obviously an estimate - I'm always suspicious of round numbers), and his Nov. 5 SMOV was over 60%. I'll definitely be watching the KPSS numbers come in today, but I don't think we can discount Yakima as a potential source for a Sanders resurgence.

    But, in the end, I agree with you - we'll have a much better idea after 7pm today.

    ReplyDelete
  3. OMG, I'm totally a big fan of your work, and it makes me wish I would have added a more complete undervote calculation to my tables.

    http://www.kingcounty.gov/elections/elections/201011.aspx is the King County one, and there's math involved to get to their actual EBOH. As of Friday, however, they were over 200k following their counts, per a phone call between Sanders and the office.

    I agree with your thought that the SOS is probably only counting ballots that are ready to be counted, not all ballots received, which makes my 1.5% number higher than it should be, but I'm okay with being conservative.

    In the end, while I have had the same concerns over Yakima County (in case you can't tell, I'm a Wiggins supporter), the ever-increasing EBOH of King County, even after they drop 75k ballots, counteracts most of that.

    I can't wait for 7...Snohomish County and their damn delayed reporting!!!

    ReplyDelete