Kerry loses about 400 votes to 3rd parties in Cleveland? 1 in 10^50 odds

[jfern]

Hmmm, is there something wrong here?

Precinct Badnarik Bush Kerry Peroutka
CLEVELAND 4A 0 10 320 0
CLEVELAND 4B 0 21 289 1
CLEVELAND 4C 0 12 313 0
CLEVELAND 4D 1 24 518 0
CLEVELAND 4E 0 14 356 4
CLEVELAND 4F 0 20 299 215
CLEVELAND 4G 3 58 261 3
CLEVELAND 4H 2 12 478 3
CLEVELAND 4I 0 7 494 0
CLEVELAND 4J 0 20 436 1
CLEVELAND 4K 0 10 413 1
CLEVELAND 4L 0 15 338 0
CLEVELAND 4M 1 17 478 2
CLEVELAND 4N 164 12 334 10
CLEVELAND 4O 0 13 439 6

So every precinct has almost no 3rd party votes except for the one with 0 Badnarik votes and 215 Peroutka votes, and the one with 164 Badnarik votes and 10 Peroutka votes? Right.

The odds are around 1 in 10^49 of this happening:

Precinct Badnarik Peroutka
4F 0 215
4N 164 10

Precints with the most 3rd party votes. There are other suspicious precincts.
http://www.dailykos.com/story/2004/12/1/52932/8333

Source of the precinct level data:
http://boe.cuyahogacounty.us/boe/results/history/2004/EL52.TXT

Graph showing how most of the precincts with high levels 3rd party votes had Kerry doing much better than Bush:

(how the bloody hell do I link to an image without displaying it? - if you want to see it click through at this link)
http://www.dailykos.com/comments/2004/12/1/155546/958/195#195

[jfern]

1. Kick

Is that image working for everyone?

[Timebound]

2. WHOA

Take down that image. Or link to it, or make it smaller. It's too large to read on my computer, and it took forever to load.

Other than that, you're very right. VERY suspicious... *clicks to send to various peoples*

[jfern]

3. Do you know how to link to images?

It's 55K, so even on a modem, it shouldn't take that long to load.

[Maraya1969]

4. It should be emphasized

that those two districts, 4N and 4F had statistically significant total number of votes. (There was one district with 1 vote for Badnarik and one for Kerry - 50% for the third party)

But in 4F and 4N the total vote counts were 534 and 520, (not sure which one was which and I don't feel like adding them up again but it really doesn't matter). So in those precincts to have a 40% and 33% voter percentage for third party is significant.

I know you highlighted those districts but I wanted to make it clear to everyone who checks the chart. In the last district in the chart the third party got 50% of the total vote of TWO. Not statistically significant.

[jfern]

5. Yeah, I just ranked everything by 3rd party vote percentage

As for statistically significant, just ignoring variations in 3rd party votes, and looking just at third party votes in 4F and 4N already gives you 1 in 10^49 odds.