Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
Display:
This situation reeks of manipulation.

The optical counters 4 hours behind the complete hand count?  That staggers the imagination.

The flip-flop in voting percentages to 5 frickin' digits over the total population!?!  That smells like a calculatory artifact.  Getting a .99999 correlation is exceeding rare in the Social Sciences.  Anyone who can get a .97 is ecstatic and the conclusion is considered bullet proof.  

Somebody in those counties need to get off their ass and file a class action lawsuit to force the SecState to a hand powered recount.

She believed in nothing; only her skepticism kept her from being an atheist. -- Jean-Paul Sartre

by ATinNM on Sun Jan 13th, 2008 at 12:07:29 AM EST
[ Parent ]
European Tribune - What happened in New Hampshire?

Optical Scan
Clinton 91,717 52.9507%
Obama 81,495 47.0493%
Total 173,212

Hand Counted
Clinton 20,889 47.0494%
Obama 23,509 52.9506%
Total 44,398

Smoking gun. Or the coincidence of the millennium.

Has anyone posted any of this to the Obama and Clinton campaigns?

by ThatBritGuy (thatbritguy (at) googlemail.com) on Sun Jan 13th, 2008 at 06:33:16 AM EST
[ Parent ]
Obama cannot call the recount or he will be skewered as a sore loser by the MSM. It has to be the Clinton campaign that calls for a recount. And the reason has been pointed out downthread. Maybe this was planted by Republicans in order to "discover" it after she gets the nomination.

However, it is possible that the actual ballots were tampered with and that a recount will confirm the results.

We have met the enemy, and he is us — Pogo

by Migeru (migeru at eurotrib dot com) on Sun Jan 13th, 2008 at 07:04:50 AM EST
[ Parent ]
Maybe this was planted by Republicans in order to "discover" it after she gets the nomination.

That is very possible. But why make the numbers so bizarre, as if whoever did this thought something like that wouldn't be noticed until several months from now (unless they were incompetent)? Wouldn't they want something that they could disguise for a few months? I'm just thinking/typing out loud.

by lychee on Sun Jan 13th, 2008 at 07:35:41 AM EST
[ Parent ]
The fact is, this is too close to being statistically insignificant. If instead of fixing the numbers to the last digit you only fix it to +- 7 votes, it drops below 95% confidence in one of the scenarios. In the other you just need to go to +- 13 votes.

The scenarios are: estimate the probabilities from the machine counts and assume the hand counts are a random sample, or vice versa.

We have met the enemy, and he is us — Pogo

by Migeru (migeru at eurotrib dot com) on Sun Jan 13th, 2008 at 07:40:46 AM EST
[ Parent ]
Plus - if this happened, it would have been a Republican scheme. And Republicans are notorious for not being terribly bright when it comes to science.
by ThatBritGuy (thatbritguy (at) googlemail.com) on Sun Jan 13th, 2008 at 07:46:04 AM EST
[ Parent ]
I was thinking whoever did this was bright enough to doctor the data, but not bright enough to doctor it properly.

The easiest way to steal the election is to just reverse the Obama/Clinton counts on a precinct basis. If you do it in every precinct it even gives you plausible deniability - sorry, computer bug! A human error programming the mappings to the database.

But maybe they thought that's easy to reverse and they really want Clinton to win. So they say okay, let's just tamper with the largest county in the state. They could just have reversed the counts and, given statistical noise, the vote percentages would have been similar but to less than 95% confidence. Being a single county, you'd have to go down to precinct level to show all the precincts are switched.

But maybe they thought they needed to eliminate the discrepancy. In statistical parlance, they worried about a one-tailed chi-square test (misfit) but not about a two-tailed chi-square test (misfit or too-good-to-be-true fit).

I mean, it just takes a pocket calculator to get the "correct" vote counts.

Tampering with the count would mean that a recount would give the win to Obama. Tampering with the ballots would confirm the results, possibly with a 200-vote difference from the initial ones.


We have met the enemy, and he is us — Pogo

by Migeru (migeru at eurotrib dot com) on Sun Jan 13th, 2008 at 07:56:17 AM EST
[ Parent ]
However, it is possible that the actual ballots were tampered with and that a recount will confirm the results.

This is actually quite important. I want to pull it out and emphasis it.

How are the ballots secured after an election? Black Box Voting seems to have some serious concerns over this.

"We have no control over the ballot chain of custody and we have learned the pain from the 2004 Nader recount, in which only 11 districts were counted, chosen by a highly questionable person, and then nothing showed up. Now all we hear is how the Nader recount validated the machines."
http://www.bbvforums.org/cgi-bin/forums/board-auth.cgi?file=/1954/71260.html

aspiring to genteel poverty

by edwin (eeeeeeee222222rrrrreeeeeaaaaadddddd@@@@yyyyaaaaaaa) on Sun Jan 13th, 2008 at 09:51:29 AM EST
[ Parent ]
ATinNM:
The optical counters 4 hours behind the complete hand count?  That staggers the imagination.
Well, I have to ask Drew to substantiate that with a link :-)

We have met the enemy, and he is us — Pogo
by Migeru (migeru at eurotrib dot com) on Sun Jan 13th, 2008 at 07:17:52 AM EST
[ Parent ]
http://www.bbvforums.org/forums/messages/1954/71287.html?1200198688

Be nice to America. Or we'll bring democracy to your country.
by Drew J Jones (pedobear@pennstatefootball.com) on Sun Jan 13th, 2008 at 10:42:01 AM EST
[ Parent ]
This comment by Bev Harris is better:
Black Box Voting : 1-7-08: Silvestro the Cat & New Hampshire Elections
Kicking myself for not tediously taking screen shots of every friggin' municipality in the dumb Googlemap thing. Why can't they just add a table below it?

One noticeable thing on the 59 screen shots I grabbed between 10:45 pm NH time and midnight NH time, is that the ones that had late results (not submitted as of 4 hours after poll closing) -- well, you'd expect them to be hand count locations, right? Nope. Mostly Diebold locations. That's a major red flag to me. How the heck can you not push "print" for four hours??? It normally takes only 30 minutes to wrap things up and print the poll tape when the polls close.

My method was grabbing the municipalities left to right, right to left, starting at the south end of the state and working up. I only got about three rows up. Anyone who has additional time slice information documenting late reporters I'd like to see it.

Late reporters from the first 59 locations I grabbed:

BRENTWOOD - Diebold location - had the Dem results, but no Republican results as of 11:53 pm (polls closed at 7)

CHESTERFIELD - Hand count location - no results as of 11:00 pm

DERRY - Diebold location - no results in as of 11:42 pm

FREMONT - Diebold location - no results in as of 11:48 pm

GREENFIELD - Hand count location - no results in as of 11:52 pm

HAMPTON - Diebold location - results in on time, but I flagged this because every Dem candidate had a result divisible by 5 and for Republicans, Huckabee 217, McCain 1217, Romney 1217, it just looked weird. So much for my statistical capabilities.

HOLLIS - Diebold location - results not in as of 11:54 pm

NEW IPSWICH - Diebold location - results not in as of 10:52 pm

NEWTON - Diebold location - results not in as of 10:58 pm

PELHAM - Diebold location - results not in as of 10:56 pm

TEMPLE - Hand count location - results not in as of 11:26 pm

WINCHESTER - Diebold location - results not in as of 10:46 pm


We have met the enemy, and he is us — Pogo
by Migeru (migeru at eurotrib dot com) on Sun Jan 13th, 2008 at 12:58:09 PM EST
[ Parent ]
Migeru:
How the heck can you not push "print" for four hours???

Weren't these votes machine-counted, by optically scanning paper ballots, rather than machine-voted?

On the othe hand, I see no correlation with number of tallied votes and late reporting:

Brentwood - 838
Chesterfield - 952
Derry - 5230
Fremont - 742
Greenfield - 368
Hampton - 3974
Hollis - 1923
New Ipswich - 717
Newton - 888
Pelham - 2484
Temple - 395
Winchester - 826

*Lunatic*, n.
One whose delusions are out of fashion.

by DoDo on Sun Jan 13th, 2008 at 01:21:56 PM EST
[ Parent ]
By county, separated:

Rockingham:
Brentwood
Derry
Fremont
Hampton
Newton

Hillsborough:
Greenfield
Hollis
New Ipswich
Pelham
Temple

Cheshire:
Chesterfield
Winchester

Be nice to America. Or we'll bring democracy to your country.

by Drew J Jones (pedobear@pennstatefootball.com) on Sun Jan 13th, 2008 at 02:01:59 PM EST
[ Parent ]
By result, Clinton and Obama percentages:

Brentwood - 39.62%, 36.16%
Chesterfield - 39.46%, 38.52%
Derry - 45.64%, 31.20%
Fremont - 41.78%, 30.73%
Greenfield - 27.45%, 42.93%
Hampton - 42.78%, 32.59%
Hollis - 35.52%, 41.19%
New Ipswich - 38.63%, 29.43%
Newton - 48.42%, 29.17%
Pelham - 50.72%, 29.03%
Temple - 24.56%, 50.38%
Winchester - 48.79%, 28.57%

Not much of a trend.

*Lunatic*, n.
One whose delusions are out of fashion.

by DoDo on Sun Jan 13th, 2008 at 03:03:55 PM EST
[ Parent ]
The sum of these is 43.21% to 33.10%. Looking at absolute numbers, 8356 to 6401.

*Lunatic*, n.
One whose delusions are out of fashion.
by DoDo on Sun Jan 13th, 2008 at 03:13:07 PM EST
[ Parent ]
Significance?

Be nice to America. Or we'll bring democracy to your country.
by Drew J Jones (pedobear@pennstatefootball.com) on Sun Jan 13th, 2008 at 03:25:25 PM EST
[ Parent ]
Not much I can see. In other words, I don't think the results coming in after 4 hours changed much.

*Lunatic*, n.
One whose delusions are out of fashion.
by DoDo on Sun Jan 13th, 2008 at 04:56:06 PM EST
[ Parent ]

Display:

Occasional Series