Hamilton's Pamphlets: Washington Recount Math

This is me doing "back of the dirty envelope math". I am not an engineer or mathematician or statistician. You want REAL math with Phd level explanations check out Stefan Sharkansky on Sound Politics or Mac at Pull On Superman's Cape or Timothy Goddard at Flag of The World . If you're looking for nuts n bolts "this is what it looks like to me" math you've come to the right place.

I'm sure that Tim or Stefan or Mac will correct me if I am wrong (and I could very easily be) but here's how I see it:

24 of 39 counties have reported their recounts. Rossi has picked up an additional 46 net votes (that means votes gained by Rossi and votes lost by Gregoire for a net gain or loss). That means as of yesterday evening Rossi's lead was 88 votes. Since 24 of 39 counties have reported, that means we have 15 counties left to report. Since King County is whacked, we'll leave them off that total and go with 14 remaining counties. Rossi averaged just shy of 2 net votes gained per county (1.92). That includes the counties that had zero change to get that average. Assuming the same average net vote gain over the remaining 14 counties in consideration gives us a net vote gain of 27 (actually 26.83, but since a fraction of a person can't vote - except possibly in King County - I rounded it up). Add the 27 to the 88 already in and you get a 115 vote lead for Dino Rossi.

Now, let's look at King County. Since they already did "a partial hand recount", I would expect their change to have been very, very small - maybe even zero. Then they come up with 561 votes that were "erroneously disqualified". Many are already arguing that these votes cannot be counted based on the WA State Supreme Court ruling announced this morning, I'm not so sure. I hope that is the case, but assuming the Dems prevail in King County on this issue (and they probably will). Here's how the numbers break down: Gregoire's advantage in King County over Dino Rossi has been about 60/40. In applying that split to the 561 votes, it looks like Gregoire needs about 60.5% to a Rossi 39.5% to eke out a slim 2 vote victory. Yes, you read that right. Out of nearly 3 million votes cast, King county has managed to "discover" just enough ballots to push Gregoire over the top. Here we go:

88 vote lead already in. 27 vote gain in remaining counties (excluding King).

That's a 115 vote lead for Rossi.

561 "discovered ballots". 60.5% for Gregoire = 339 (rounded down) and 39.5% for Rossi = 222 (rounded up).

Add Rossi's 115 vote lead with the gained 222 votes and you get 337 votes.

Gregoire gets 339 of the 561 "discovered ballots".

That's 2 MORE than Rossi. Gregoire wins by 2 votes! Bet they won't call that a tie!

If the remaining counties (other than King) have more zero changes or result in net gains for Gregoire the picture only gets worse. If Rossi fails to get at least 39.5% of the 561 votes it only gets worse. Say neither candidate picks up or loses votes in those 14 counties and the King County split goes 70/30 in Gregoire's favor. She wins by about 136 or so votes...

After the mandatory recount, Gregoire and the Dem's were calling Rossi's 42 vote win "a virtual tie". I have a virtual reality for them. "Discovering" these 561 ballots is leading our state to "virtual anarchy". If Gregoire and the Dems succeed in stealing this election, their only legitimate claim can be "Governor of Seattle and King County".

Again, out of nearly 3 million votes cast, King county comes up with precisely the number (based on this election's average vote split) that Gregoire needs to get the win. This goes beyond incompetence and sloppiness. I'd like to see Stefan or one of the other math wiz kids tell me the probability of finding the EXACT number of votes needed to steal the election. What are those odds? A mere coincidence, I think not.

Comments

You forgot to mention that the King County Council Chair was one of the 561 people who had their vote initially disqualified.

I'm not quite ready to say that these 561 "missing signatures" were "fixed" after the election but as Hugh wrote "if it isn't close they can't cheat"....

Posted by: Duane | 14 December 2004 at 04:10 PM

Breaking news. I heard Chris Vance say on Carlson that it's no longer 561, but 577!

Assuming that all of the 577 ballots are valid and they break proportionally as all of the previously counted absentee ballots, you'll get 550 two-party votes, with the remainder split between no-vote, Libertarian, write-in, etc. Of thoe two-party ballots, Gregoire would have a lead of 19.4%, or 106 votes from this group.

Posted by: Stefan Sharkansky | 14 December 2004 at 05:51 PM