Hillary has started to feel the Bern (Politics threat)

by **jerseyhoya** » Wed Sep 02, 2015 14:27:24

She's looking at a relationship between precinct size and 2 party vote share, with an assumption that as precincts get larger, the vote share should become more Democratic. She says this is because the largest precincts will be urban and not rural. She does nothing to demonstrate that the larger precincts are urban (rather than suburban, which one might expect wouldn't be more Democratic), just takes it as given. When she finds the very largest precincts are more Republican than the mid sized ones, she takes this as evidence of fraud. I'd guess what she's finding is many of the largest precincts in the states she looks at are probably in suburbia - where precincts can be encompass more area because people are assumed to drive and such - rather than in inner cities. She does nothing to look at how these precincts voted in past elections, just says the most likely explanation for this is multiple levels of manipulation of the voting process. She shows some voting systems have different patterns of 2PVS by precinct size, but doesn't control for which counties use the different voting systems. We're talking very basic stuff here.

A good way to look at potential fraud would be to examine counties that have changed their vote tabulation methods between elections and see if there were any anomalous results. I don't think the idea of manipulating machines to commit election fraud is beyond the realm of possibility, and the system should be transparent so people have faith that the election results are valid. But when you put forth a 'study' that ignores how things have gone in the past and just looks at a one off election and uses a weird default assumption to base things on you're not uncovering anything sinister - just demonstrating that you're an idiot.

Website · by **Wolfgang622** » Wed Sep 02, 2015 14:43:01

CalvinBall wrote:
Houshphandzadeh wrote:not really, imo

lol

Well, reading through her argument, I think there might be something there, although I'd like to see how the numbers line up in elections won by Democrats. The argument, which wasn't clear to me from the initially linked article, so I'll summarize it here for anyone who would like it boiled down to a nutshell, is, essentially, this: as you plot out the percentage of Republican vote in each precinct in increasing order of size in three selected recent state-wide elections (2012 Ohio Prez, 2014 Wisconsin Gov, 2014 Kansas Senate), you find that as the size of the precinct increases in terms of population, the percentage of R vote increases. This could be indicative of shenanigans because, theoretically, what you should find is that as the size of a precinct increases, the percentage of the R vote should flat-line approaching the "limit" of the overall percentage of "R" vote across the state - that is, the larger the precinct, the more closely the percentage of R vote (and, although not studied, D vote) within the precinct should match the state-wide percentages, because the larger the sample the more representative it should be of the whole. The anomaly is more pronounced in precincts when you isolate by voting machine type - in precincts where various electronic machines are used, the effect is more stark (although how much more stark varies by quite a bit on eyeball inspection).

The trends aren't jaw-dropping by any means in terms of how big they are. In 1 case it failed to affect the outcome (Ohio), in 1 case (Kansas senate) the Rs were on their way to an easy victory regardless, and in Wisconsin it probably would not have affected the outcome. These three case studies do seem to be cherry-picked: over the period 2012-2014, there must have been at least 140 or so state-wide elections (50 presidential excluding DC, 66 senate, and let's say 24 gubernatorial, though that is just a guess), but we are looking at 3, 2 R wins and 1 Dem win. Is the trend present in the closest statewide elections in that period? Does it exist for Ds in states they won? I'd love to know the answers to those questions. The most interesting of her examples is WI, because the percentage of R vote as the sizes of precincts increase goes from roughly 50% to 52.5%. This would seem counter-intuitive - generally, the larger the precinct, the more likely it is to be urban, and the more likely it is to be pretty Democratic - but without knowing the exact make-up of Wisconsin precincts (larger precincts may not necessarily be in high-population density urban areas, for example: maybe Milwaukee has lots of little precincts and some other less densely populated areas have bigger, by population precincts), it's hard to judge. I will say that it would appear that Walker achieved 50% of the vote or better in every precinct in Wisconsin in 2014, if I am reading the chart correctly, and having been to Madison that does not pass the sniff test.

Website · by **Wolfgang622** » Wed Sep 02, 2015 14:57:34

jerseyhoya wrote:She's looking at a relationship between precinct size and 2 party vote share, with an assumption that as precincts get larger, the vote share should become more Democratic. She says this is because the largest precincts will be urban and not rural. She does nothing to demonstrate that the larger precincts are urban (rather than suburban, which one might expect wouldn't be more Democratic), just takes it as given.

I agree with a lot of what you said, but to be fair I think this misrepresents her point. It's not that urban precincts should be more Democratic, it's that precincts with larger populations should have voting percentages, with perhaps room for an outlier or two or three, that trend very closely to the state-wide percentages for each party within the state. So, if Scott Walker won 52.3% of the vote in 2014, then in the biggest handful of districts he should have won approximately that percentage of the vote, and as you look at districts in increasing order of size, the trend should be that the result in each individual precinct should get closer and closer to the state-wide result. Instead, Walker's percentages start trending up after a certain point as the precincts get bigger (What she does NOT say, and what I realized in reviewing my posts, is that Walker's percentage of the vote dips a bit to about 50% in some of the mid-sized precincts, but then recovers, gradually, to get up to ~52.5% in the largest precinct - which is almost exactly the percentage by which he won - kind of damning to her point I should think, at least in the case of WI, because although there is an upward trend in Walker's percentage in the larger precincts, it is an upward trend that approaches his state-wide percentage, rather than increasing further away from it). She also claims in precincts where certain types of machines have been used, this effect is more pronounced (I think this point is strong in Wisconsin, weak in the others, based on a layman's look at her graphs).

There could be all sorts of non-fraud reasons for this, some of which you elucidated, but that is her point, not that Democrats "should" win bigger precincts. Indeed, I would imagine that the trend she imagines should be the case (results in big precincts getting "truer" to state-wide results) would not be the case at all in states where Ds have typically won, because they often win huge in the biggest, urban districts, while losing many more smaller districts.

Your point about looking at election-to-election trends, and particularly pinpointing where machines were changed and seeing if wild swings followed, is a good one.

by **Bucky** » Wed Sep 02, 2015 15:23:09

maybe richard pryor programmed the machines and they only take switch one vote from every machine in order to avoid suspicion

Website · by **Wolfgang622** » Wed Sep 02, 2015 15:35:04

Actually, looking at her post again, with the numbers in front of me, this is her problem: her point is, as I said before, that "the law of large numbers ensures that the sample proportion will quickly approach the true proportion" - i.e., that the percentage of vote for any given party should get closer and closer to its state-wide result as the size of the precincts increases. But in each of the examples she uses - 2012 Ohio, 2014 Wisconsin, and 2014 Kansas, she "forgets" to give us the information about what the state-wide percentage actually was for the R candidate - instead trying to lead us to water by showing that an "upward trend" in percentage of R vote must be anomalous. It certainly would be, if, for example, Scott Walker had won 58% of the vote in the biggest handful of precincts while achieving only 52.5% state-wide. But he only won 52.5% in the biggest precincts - the trend line is up, yes, but up from a percentage below his overall state-wide performance.

In 2012 in Ohio, Romney won 47.69% of the vote, and there is an upward trend line in her data - up to what appears to be 46, maybe 46.5 percent of the vote in the biggest precincts. In Kansas in 2014, Pat Roberts won with 53.15% of the vote (is that all?), and his trend line in the biggest districts moves up, yes (but only very, very slightly), and up from 53.5 (close to his "true" percentage) to maybe 54 (slightly bigger than his "true" percentage). Not really very compelling.

Now I am forced to agree with jersey: I think this woman silently conflated the basic idea of "the law of large numbers" - which, she is right, should predict vote percentages get closer to the "true" percentage as the size of a sample increases - with the idea that in larger precincts, Dems always win. She thinks share of R vote should flat-line or decrease as the size of the precinct increases, because she assumes that the percentage of R vote will be simply decreasing BUT ALWAYS MORE THAN the state-wide average as you go from the smallest to the largest precincts. The percentage of R vote in the precincts in these cases DOES approach the "true" - i.e., state-wide - number as the size of the precinct increases, but from the "south." It has not occurred to her that R vote can be significantly less in mid-size precincts than the statewide average.

by **TomatoPie** » Wed Sep 02, 2015 15:51:53

A bit awkward that a president is changing the name of a mountain on a climate-change trip to what most know as a gas-guzzling SUV

by **Houshphandzadeh** » Wed Sep 02, 2015 15:56:31

so awkward

by **jerseyhoya** » Wed Sep 02, 2015 16:18:16

Because everything is more interesting than doing the work I'm supposed to be doing, I spent 45 minutes going through the Wisconsin 2014 Gubernatorial results. One can find the ward by ward results here if the spirit moves anyone else.

One thing that becomes readily apparent is that a lot of the reported results are a handful of wards (presumably all contained within the same precinct/voting facility) grouped together. The largest precincts are almost all in towns I've never heard of. Of the top 25, a precinct from Sheboygan (Wards 1-10), is the only municipality that rings a bell.

I went through, sorted all the precincts by total votes, broke them into 500 vote groupings (with 2500 votes+ as the largest group), and calculated votes cast and Walker's vote share.

Code: Select all: Votes GOP Dem Total GOP% N 2500+ 120,303 91,948 214,358 56.1% 69 2000-2500 92,786 79,781 174,383 53.2% 79 1500-1999 178,870 135,306 317,371 56.4% 184 1000-1499 265,054 232,651 503,303 52.7% 419 500-999 379,335 385,698 774,501 49.0% 1080 1-499 223,358 197,529 426,398 52.4% 1652

So a few notes
-Walker's best group is the largest precincts. Noteworthy is there is not a single precinct from Madison or Milwaukee contained in the over 2500 votes group.
-Most city wards aren't grouped into larger bunches. This makes some sense as each individual precinct is probably at the neighborhood school/firehouse/church that people can walk to and they report separately. Scrolling through it looks like Milwaukee is heavily overrepresented in the 500-999 range, which also happens to be Burke's best grouping.
-There aren't terribly major swings from group to group.
-Ms. Clarkson's #analysis is p. dumb.

by **The Crimson Cyclone** » Wed Sep 02, 2015 16:19:40

Ok cool, glad it isn't probable. Would be a bad thing if it were

by **Monkeyboy** » Wed Sep 02, 2015 16:19:59

Houshphandzadeh wrote:do you feel good about having an electable Democratic POTUS candidate who is just going to maintain the status quo

Among the people who have a legit chance, I think that's true of all of them. Both parties. Need to get the money out of the game by publicly financing elections and whatnot.

we're screwed

by **Monkeyboy** » Wed Sep 02, 2015 16:25:09

mozartpc27 wrote:
Soren wrote:If Hilary is going to do a perp walk for this email thing then when are we going to nail Cheney to the cross

I laughed a lot at this. Still giggling. Truth, sir - truth.

she should have placed all the classified info in a safe. You know, so nobody could see what she was doing.

by **Monkeyboy** » Wed Sep 02, 2015 16:39:23

jerseyhoya wrote:Because everything is more interesting than doing the work I'm supposed to be doing, I spent 45 minutes going through the Wisconsin 2014 Gubernatorial results. One can find the ward by ward results here if the spirit moves anyone else.

One thing that becomes readily apparent is that a lot of the reported results are a handful of wards (presumably all contained within the same precinct/voting facility) grouped together. The largest precincts are almost all in towns I've never heard of. Of the top 25, a precinct from Sheboygan (Wards 1-10), is the only municipality that rings a bell.

I went through, sorted all the precincts by total votes, broke them into 500 vote groupings (with 2500 votes+ as the largest group), and calculated votes cast and Walker's vote share.

Code: Select all
Votes GOP Dem Total GOP% N 2500+ 120,303 91,948 214,358 56.1% 69 2000-2500 92,786 79,781 174,383 53.2% 79 1500-1999 178,870 135,306 317,371 56.4% 184 1000-1499 265,054 232,651 503,303 52.7% 419 500-999 379,335 385,698 774,501 49.0% 1080 1-499 223,358 197,529 426,398 52.4% 1652

So a few notes
-Walker's best group is the largest precincts. Noteworthy is there is not a single precinct from Madison or Milwaukee contained in the over 2500 votes group.
-Most city wards aren't grouped into larger bunches. This makes some sense as each individual precinct is probably at the neighborhood school/firehouse/church that people can walk to and they report separately. Scrolling through it looks like Milwaukee is heavily overrepresented in the 500-999 range, which also happens to be Burke's best grouping.
-There aren't terribly major swings from group to group.
-Ms. Clarkson's #analysis is p. dumb.

She's a GOP plant so that the real fraud that occurs in the next election will be more easily dismissed when people complain. The GOP sees the chance to grab all 3 chambers. They'll do anything to make sure it happens.

Wait, is this the conspiracy thread?

by **Phan In Phlorida** » Wed Sep 02, 2015 16:53:50

The Crimson Cyclone wrote:will be Interesting to see where this story will end up

http://www.mintpressnews.com/university ... ud/209187/

With Ms. Clarkson's body being pulled out of the Arkansas River.

by **pacino** » Wed Sep 02, 2015 17:21:24

Why can't the 12th place person be allowed in the Republican CNN debate?

by **SK790** » Wed Sep 02, 2015 17:28:37

Who really cares? The cutoff has to be somewhere.

by **pacino** » Wed Sep 02, 2015 17:31:03

I care. Why is Christie still in it if they are going by more recent polls?

It's arbitrary. Why is Christie more viable than Bobby Jindal? Iowa voters don't seem to see much of a difference.

by **SK790** » Wed Sep 02, 2015 17:43:52

It's arbitrary no matter where you make the cutoff.

by **drsmooth** » Wed Sep 02, 2015 17:47:19

jerseyhoya wrote:Why is it an overblown conspiracy theory?

That HRC is a Russian spy or was endeavoring to steal trillions in bitcoin using the Watson-level computing gear she had secreted in her Fortress of Solitude or furtively maneuvered to have thousands of people killed in an oil war she dreamt up reasons to foment - no wait, that last was Cheney, strike it

by **SK790** » Wed Sep 02, 2015 17:53:40

The best part is conservatives act like Cheney is still a reliable source on foreign policy. Fox News trotted him out several times to try to convince their idiot viewers that the Iran deal will lead to the end of America.

Not like he's ever overblown a Middle East country's capacity to produce WMDs before. This guy is very reliable and good.

by **Bucky** » Wed Sep 02, 2015 18:10:28

maybe they could only find 11 matching podiums (podii?)

Archived

Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)

Re: Hillary has started to feel the Bern (Politics threat)