New, Improved Analysis of NH Democratic Primary Results By Voting Method.

[mhatrw]

Total Votes: 282,065 Total Machine Counted Votes: 224,881 Total Hand Counted Votes: 57,184

Percentage of Total Counted Votes:

39.17% Clinton 110,478
36.43% Obama 102,765
16.93% Edwards 47,755
04.60% Richardson 12,977
01.37% Other 3,867
01.36% Kucinich 3,826
00.14% Gravel 397

Percentage of Machine Counted Votes:

40.28% Clinton 90,571
35.85% Obama 80,614
16.78% Edwards 37,743
04.45% Richardson 9,781
01.37% Other 3,082
01.23% Kucinich 2,773
00.14% Gravel 317

Percentage of Hand Counted Votes:

38.73% Obama 22,151
34.81% Clinton 19,907
17.51% Edwards 10,012
05.59% Richardson 3,196
01.84% Kucinich 1,053
01.37% Other 785
00.14% Gravel 80

Percentage of Machine Counted Votes - Percentage of Hand Counted Votes

05.46% Clinton
00.00% Gravel
00.00% Other
-0.60% Kucinich
-0.72% Edwards
-1.24% Richardson
-2.89% Obama

That's an 8.35% difference between hand counted & machine counted votes when comparing Clinton to Obama.

[Uben]

1. So, what does this imply? nt

nt

[MADem]

2. They're also from different counties. The machine counted votes are

from larger population centers, the hand counted, from smaller, rural areas.

[mhatrw]

6. Yes, and sparsely populated rural areas are well known for being less racist than urban areas.

Right?

[MADem]

14. WTF??? Are you vacationing from reality, or what?

Obama didn't get one percent of the vote, here. He came in SECOND.

Jesus!

The converse to your sick 'RACIST!!!!!" implication is that sparsely populated rural areas are 'well known for being MORE SEXIST...'

Of course, that's bullshit too.

[mhatrw]

20. I was making fun of the so-called Bradley effect explanation that

MSNBC was floating last night. Do I need to explain further?

[MADem]

30. MSNBC dug Clinton's grave the day before. They don't get much right.

They're owned by General Electric--and they tailor their message accordingly.

It just wouldn't make sense that one population demographic would "Bradley" and the other wouldn't, though.

Obama didn't do the work on the ground in NH. Clinton did. She had a good GOTV program, she encouraged people to vote absentee (the snowbird vote is significant) and she campaigned in the state by answering questions, not lecturing people.

[mhatrw]

54. I'm not saying that there are no reasonable explanations for the discrepancy.

If Clinton had a superior urban/suburban GOTV operation, that would be a legitimate explanation.

[Skinner]

3. This does not prove anything.

To be relevant, you need to compare similar precincts with different voting technology.

Perhaps what we are looking at here is an urban/rural split. Big cities with large populations of voters are more likely to use machine counts. Rural precincts with small populations are more likely to use hand counts. Your data could well be showing urban vs rural, not machine vs hand.

[Uben]

7. Thanks

I was going to point that out, as well!

[mod mom]

8. This needs to be examined, regardless. When you see GOP candidates and Edwards'

results reflecting the polls, you have to question why Obama and Clinton deviated so much. Even the Clinton folks were expecting a loss. Maybe you are correct and this means nothing, but a true (not the pre conceived Blackwell type) audit would place faith in the system.

[MADem]

15. The results match the exit polls, though.

There's no deviation.

[mhatrw]

21. The exit polls are being tuned to the results. They learned from 2004.

They are not releasing the exit polls to the public until they tune them to the results. So the exit polls prove nothing.

[MADem]

23. ... ahhhh .... a CONSPIRACY theory~!!! Yeah, THAT's the ticket!

:rofl:

If it's not RACISM, it's a conspiracy!!

Are you serious? Or are you going to claim that was a wry comment, too!

Amazing how all these quiet henchmen manage to keep such a massive exit polling operation secret--why, they're more disciplined than the fellows you see in old James Bond films, running around in baby blue jumpsuits in underground caves or on remote islands at the behest of some evildoer or another!!

Hope you've stocked up on Reynold's wrap!!! You're clearly gonna need it....

[mhatrw]

24. It's not a conspiracy. It's SOP.

Nobody is EVER going to publicly publish exit polls that don't match the official results again.

[MADem]

25. Now you're really making me laugh.

Cue up the Henchmen!!!!

But surely, there must be someone, anyone, who will spill the beans on this nefarious plot!!!!


It couldn't possibly be that there isn't at least one MOLE in the bunch???



:rofl:

[mhatrw]

27. Show me one publicly published exit poll since 2004 that didn't match the election results.

Seriously. Let's see it.

[MADem]

29. I got some bad news for you--the reason they match is because they're ACCURATE.

You're childishly insisting that I "prove" something that doesn't HAPPEN, because the exit pollsters GET IT RIGHT.

I'd like to 'seriously' see some logic out of you, but that's plainly not going to happen. You're way too invested in conspiracies of pollsters, number crunchers, media people at all levels, who 'invent' a result just to fuck with your favorite candidate. Gee, were these nefarious henchmen taking the day off during the Iowa caucuses?

Here, you're gonna need more, I think:

[mhatrw]

40. It's proven. It's admitted. It's SOP.

I didn't make this up in my head, MADem. It's SOP. It's what US polling companies conducting exit polls for US elections now routinely do whenever the raw exit poll data doesn't match the election results.

http://uscountvotes.org/ucvAnalysis/US/Exit_Polls_2004_Edison-Mitofsky.pdf

“Adjusting”, “re-weighting,” “renormalization,” or “forcing” of exit poll data is also a weighting
process but one which involves the use of tabulated vote counts, to which the already demographically
weighted exit poll results are brought into congruence. Once the polls have closed and tabulated
results become available, first at the precinct and then at the county level, Edison/Mitofsky's exit poll
results are adjusted to mirror these tabulated results. Thus, it has been stated, “What you would get
after the election are preference (i.e., exit poll) numbers forced to the election result.” 19
Such adjustment of exit polls results to congruence with incoming tabulated election results did indeed take
place late on election night 2004, and was posted to media websites during the hour or so after approximately
12:24 a.m. on November 3, 2004. These adjusted results remain posted, as of this writing. 20

The demographically weighted results available prior to this adjustment will be referred to as “unadjusted.”
Many have questioned whether this process was in itself sinister, designed to conceal troubling
questions about vote counting in Election 2004 as revealed by the unadjusted exit poll results. The
answer is almost certainly no. And yet the effect of the process was at the very least confusing and
served to blunt public awareness of the dramatic exit poll-vote count discrepancies during the critical
period immediately following the election.

While we acknowledge that slight adjustment may legitimately be made to exit polls using the
reported election results, so that the results can be used to assess the demographics and opinions
underlying the reported voter shares in a consistent fashion, the justification for doing so rests entirely
on the assumption that the reported election results are in fact accurate, as reflected by a small and
undramatic discrepancy between exit poll results and vote counts. In order for the exit poll results in
2004 to be used in this manner they had to be substantially, in fact dramatically, adjusted. Such
substantial discrepancies and the need for such dramatic adjustment raised a bright red flag.
Edison/Mitofsky ignored this red flag and simply substituted the adjusted data set, which has been
generally employed without acknowledgement.

[MADem]

57. Your cite is about the cockup of 04. You are suggesting, falsely, that

normal statistical weights that are made to account for known variables are now somehow nefarious. Gee, I wonder if they'd be nefarious if someone else had won?

I'm not buying your argument--you're just pissed because your team didn't win, I think, so you're doing all you can to smear Clinton with a "cheater" designation.

[mhatrw]

61. I have no dog in this fight, MADem.

And if someone rigged the opscan machines, I highly doubt that it was Clinton.

Exit polls are routinely normalized to match election results. In such cases, they obviously cannot be utilized to confirm the accuracy of machine counts. Because CNN waited until after the election was called to publicly release their exit poll percentages last night and because the raw municipality data was not released, it's foolish to pretend that the exit poll confirms an accurate machine count because it may just as well have been adjusted to match the reported results.

[OnTheOtherHand]

39. where were you on election night 2006?

You don't know what the exit poll results were that night? Geez, this is wearing me out.

Ever heard of "Landslide Denied"? The analysis is awful, but the premise is accurate.

[mhatrw]

42. But the only reason we even know about that is because some people

made screen captures of the UNADJUSTED exit polling date BEFORE Mitofsky/Edison adjusted their raw exit polling numbers to match the election results. The "official" exit poll results for 2004 on CNN right now EXACTLY match the election results BECAUSE THEY WERE ADMITTEDLY ALTERED ON PURPOSE IN ORDER TO MATCH THESE RESULTS.

See page 19 & 20: http://uscountvotes.org/ucvAnalysis/US/Exit_Polls_2004_Edison-Mitofsky.pdf

[OnTheOtherHand]

56. what is your point?

In 2004 and again in 2006, CNN (and probably other networks) posted exit poll tabulations as polls closed.

Do you agree with this, or do you dispute it? If you dispute it, on what grounds?

Yes, I am well aware that the tabulations are updated to be consistent with the projections, and ultimately with the final (or semi-final) counts.

[mhatrw]

62. In this case, I did not see the exit poll data online until the race was called.

Therefore, the fact that the (quite probably adjusted) exit polling numbers match the results cannot be used verify the accuracy of the results.

[OnTheOtherHand]

63. false premise, correct conclusion

Personally, I don't care what you saw. But there are at least three distinct reasons why the tabs couldn't be used to verify the accuracy of the results:

(1) 8:00 tabs could well have been adjusted based on even a small fraction of vote results available before that time.

(2) If they didn't incorporate vote counts, then they almost certainly incorporated pre-election expectations.

(3) Exit polls are not sufficiently accurate to verify results.

[mhatrw]

69. Agreed. I misunderstood your correction.

http://www.pollster.com/blogs/looking_for_new_hampshire_exit.php

By the end of the night, the tabulations will be weighted to the official count.

http://www.pollster.com/blogs/exit_polls_what_you_should_kno_1.php

Once the polls close, the interviewer will attempt to obtain actual turnout counts, and if possible, actual vote returns for their precinct from polling place officials. One of the unique aspects of the NEP exit poll design is the way it gradually incorporates real turnout and vote data as it becomes available once the polls close. The exit pollsters have developed weighting schemes and algorithms to allow all sorts of comparisons to historical data that supports the networks as they decide whether to "call" a race for a particular candidate. When all of the votes have been counted, the exit poll is weighted by the vote to match the actual result.

[OnTheOtherHand]

72. OK, are we on the same 'page' now?

I'm juggling lots of things and I'm pretty sleep-deprived, but if we can agree on some facts, then we can build on that.

[John Q. Citizen]

120. In fact, nobody published exit polls that didn't match the official result in 2004, remember?

The only way people became aware of the discrepancy in 2004 was somebody took a sceen shot of the raw data accidentally released before it could be updated with the "official" vote count totals.

And yes, the pollster in 2004 readily admited to ajusting his exit poll in 2004 to reflect the "official" totals.

I'm surprised people forget this so soon.

Even if the New Hampshire totals are squeaky clean, (and my guess is they probably are) how would we know? Without an audit, I mean.

[Two Americas]

80. serious issue

The integrity of elections, and public confidence in the integrity of elections, is such a vital issue that none of us should ever be crying "tinfoil" about it when others raise the issue.

The things you are portraying as impossible and ridiculous have happened, and it is incumbent upon you to get up to speed on this issue as a Democrat and as a citizen.

[MADem]

90. Why didn't anyone wonder about those results that aren't even KEPT in Iowa?

This has everything to do with who WON.

You go to the NH SOS and ask to count all those OPTISCANS by hand if it bugs you so much....

:eyes:

[Muddy Waters Guitar]

93. Maybe b/c Iowa is a ***caucus*** and thereby transparent?

There's nothing to audit there, it's all in the open.

[MADem]

98. No, it isn't transparent. There is no repository of "the numbers" on the D side.

It's actually pretty fucked up, if you ask me. I know many people don't care for this crew, but they note the obvious screwups with this "less than transparent" process. http://www.indybay.org/newsitems/2007/12/31/18469456.php

[Muddy Waters Guitar]

100. It was filmed on TV, in the open, with every vote witnessed and monitored

Seems like a pretty good approximation of full transparency.

[MADem]

101. EVERY single caucus location? I don't THINK so.

If they put every location on TV, they'd still be showing the fucking thing.

[Muddy Waters Guitar]

109. It's called videotape

Not every single damn precinct caucus count is going to be shown on Channel 4 News at 11-- the results are recorded and tabulated in many ways. You have a number of independent assessments of the results and they are used to verify each other. That's the ultimate standard of transparency.

[MADem]

114. That's not what you said, though.

You're moving the goalposts.

You're telling me every corner of every location was filmed? With full audio?

Bullshit.

Video is ALLOWED. It isn't MANDATED.

The process is anything BUT transparent, and it isn't representative, either. If you work the midshift (doctors, nurses, cashiers, cabbies, cops, firemen, sales clerks) or if you're sick (hospital, nursing home) or if you're on active duty (fuck you, troops who are deployed!) or you dare to go on vacation, you're DISENFRANCHISED.

Heck of a system...not.

[Muddy Waters Guitar]

115. "TV" means "TV medium" not "10 o'clock news" MADem, do we need to go consulting Webster's now?

Results being filmed "on TV" simply means they were recorded and maintained "via broadcasting using radio waves from one point to another"-- it does not mean that every single precinct video was shown in detail on the nightly news. It means, you can see the entire process on a TV screen. IOW, it's transparent and easily audited, the whole point here.

[MADem]

116. No, it does not. TV means the thing that is hooked up to an antennae, cable or satellite

A few odd tapes of an event or events made by PRIVATE citizens is not TV.

The pictures of the family children frolicking with the dogs at the lake isn't "TV."

Stop moving the goalposts. It doesn't help your argument.

You made an incorrect assertion. I called you on it.

And even if I did take your huge stretch of a 'revised assertion' at face value, not all caucuses were filmed. A fact I pointed out and you ignored.

There's no "central repository" of caucus tapes, held by a state election agency. There is no blanket "official video record."

You're trying to infer that there is.

The courts have ruled that individuals have the right to tape so long as they don't interfere, but as I said--and you also conveniently ignored--there is no MANDATE for so doing.

[Two Americas]

112. you must be kidding

Caucuses are about as transparent as it gets. You can't compare that to the problems with electronic voting.

A Clinton win in NH was the outcome I preferred.

Still, there no doubt are Obama supporters here who would not be interested in this at all had their candidate won, and no doubt there are many Clinton supporters who would be interested had it happened the other way. I understand there were some Edwards supporters jumping on the bandwagon, as well. That sucks.

The whole thing is one ugly mess, and that is the problem. Once there is uncertainty about the integrity of elections it always will be a mess.

[MADem]

117. Yeah? Look up VOXEO and get back to me.

It's not just people, it's machines, too.

[Two Americas]

119. not sure what you are talking about

I don't know what you mean about people and machines. The problem with the machines is not in dispute, I don't think.

For the sake of the board and the party, and out of respect for the admin, I have decided that this is not the time or place to discuss election fraud.

Not sure why you are being so hostile toward me.

[mod mom]

58. IF you were going to steal an election, you would learn that this must be done.

[mhatrw]

9. Sure, or it could well be showing that the voting machines are rigged.

Why do I have to prove anything?

I am simply presenting data. I'd love to be proven wrong. Unfortunately, as long as our votes are routinely counted by black box software with no random auditing and no hand count verification whatsoever, we'll just have to cling to the same blind faith in the infallible accuracy of proprietary counting software that our corporate media demonstrates regardless of how many pre-election polls (or in the case of 2004, even the initial released exit polls) the results conflict with.

[Skinner]

11. You have to prove something because the burden of proof is on you.

You are the person making the claim that the accepted version of reality is wrong.

[mhatrw]

12. Where did I make any claims like that?

I posted data. Interpret it however you please. I don't claim to have proven anything. The data concerns me. I would like a hand count of the machine counted ballots to be certain. But I wouldn't make any bets either way as to the results of this hand count. Fraudulent voting machines are just one of several possible reasonable explanations.

[Skinner]

13. Cute. (nt)

[mhatrw]

18. Look, I don't like not being sure of results.

Do you? I trust the hand counted results 100 times more than I trust the machine counted results. Don't you?

I agree 100% that I haven't proven anything. But just because everybody in corporate media and our political establishment turns a blind eye to the problems of black box vote counting doesn't mean that they doesn't exist.

Really, I don't care if Clinton or Obama won. I was rooting for Edwards, and I'm lukewarm about him. My problem is with another mysterious election result that we are supposed to buy wholesale without question because Hillary cried or urban people are more bigoted than rural people or young people are fickle or everyone was lying to the pre-election pollsters or something. And I don't see what the exit polls prove unless you can point to a full exit poll that was publicly released BEFORE any results were. Edison/Mitofsky is tuning their exit polling data to the actual results.

[rinsd]

41. Hey stickdog! Did you give up on your HPV crusade?

[truckin]

33. Cute? He makes a very reasonable response and that's how

you respond? You should be better than that.

[mhatrw]

53. I actually editted my response while Skinner was posting.

The post started out being a little evasive, but then I thought better of it and decided to make myself perfectly clear. For the initial post, Skinner's response was not inappropriate.

[truckin]

55. Thanks, sorry Skinner

[MADem]

89. The guy OWNS this site...he can respond however he wants. We're guests in his house. NT

[MADem]

16. You threw a racial stink bomb in post six. That's more than just posting data.

You're suggesting nefariousness with no proof.

Explain why the exit polls matched the results, then--those evil polling and media people are part of the conspiracy, too?

[mhatrw]

19. All I was doing was making fun of the so-called Bradley effect explanation.

Get it?

[MADem]

22. It sure didn't smell that way from here. nt

[IsItJustMe]

17. It would be so easy to hand count sample, extrapolate, and then compare it to a machine count to see

if they are consistent. There is no reason in the world not to do this type of comparison, especially after the last two general elections. Something is terribly wrong with our voting systems in this country.

It's seems like that should be self evident.

[JackRiddler]

85. Well then, you should support the only means of proving or disproving anything.

What is the "accepted version of reality" and who did the accepting?

mhatrw should provide the proof, and there is a way she could - if enough people, including yourself, were to support a hand recount.

The difference between paper and optiscan counties is 8 percent. How large would the discrepancy have to be, for you to allow that there should be a hand count to verify the accuracy of the announced results?

[Muddy Waters Guitar]

118. It would seem that the burden of proof should be on the vote-counting systems

to ensure that the vote tally is done in a fully open and transparent manner to provide trust in the outcome. Using Diebold machines which are easily hackable as shown 2 years ago, and without an auditable paper trail, wrecks that trust. Tremendous distrust and cynicism has already been generated by the vote thefts in 2000 and 2004, and in the New Hampshire primary, this would be the last straw, period.

[harun]

59. I too am curious to see the demographics data but one cannot

dismiss that we see almost no variance in the other candidates percentages between the hand and machine counts. If we were looking at significantly different demographics, to explain the variance, I would have expected a little more float from the other candidates, just not all the float occurring with Obama.

[mirrera]

86. My sister-in-law is in a real rural area and she said it was Obama fever everywhere. n/t

[IamyourTVandIownyou]

103. Obama won the college grad vote

Do they live in cities or rural?

[mhatrw]

111. They live in towns that use opscan technology. n/t

[ayeshahaqqiqa]

4. It implies that I vote early in my primary

because that way I get to use a paper ballot and not a voting machine.

[goldcanyonaz]

5. The ABC crowd is out in full force this morning. Saboteurs.

[mhatrw]

10. I'm not ABC. I'm ABBB. Anything But Black Box. n/t

[Roland99]

26. More Obama supporters voted early?

*shrug*

[Kip Humphrey]

28. mhatrw, thanks for posting the numbers. Why are DUers attacking the messenger? We need a

hand recount of the machine vote, by the numbers. If Clinton's win sticks, great! If it doesn't, we've struck a blow for election integrity. Apparently Skinner and others here on DU have a serious problem with such transparency. Why?!!!

[HughMoran]

36. Because it's fucking complete nonsense

...and just makes those who can't accept a result look like complete idiots.

[Kip Humphrey]

43. Thanks, HughMORON, for clearing that up!

[HughMoran]

47. No problem

We won't have to do this again!

:hi:

[mhatrw]

44. I accept the result. Why can't you accept verifying the result?

The morans are the ones who asks us to accept any and every black box voting result with no auditing or verification of accuracy whatsoever.

[Zynx]

31. This is garbage.

Many of the hand counted areas are in the more liberal areas of the state that Obama was supposed to do well in anyway. Those would be those rural, moutain counties out in the western portion of the state. The machine counted areas are those that are closer in to Boston and were supposed to go more strongly for Hillary anyway. In either case, even the hand counted areas show a much closer result than the polling anyway.

[mhatrw]

48. The difference was over 8%. If the statewide election results matched the

hand counted elections results, the results would have been within the margins of error of all of the pre-election polls.

Show me some evidence that the hand counted municipalities are more liberal overall than the machine counted municipalities.

Overall, 54.8% of NH residents cast their primary ballots for the Democrats. 54.4% of NH residents of the municipalities that hand count their votes cast their primary ballots for the Democrats. So it appears that you were talking out of your butt.

[Zynx]

51. The rural and urban areas of NH vote very differently.

Look at the results in the 2004 primary for a comparison.

[mhatrw]

60. Again, you are presupposing that the 2004 machine count was accurate.

Furthermore, Dean being a longtime VT governor added a geographical component to the 2004 primary that wasn't part of this primary. Finally, more independents voted Democratic last primary because Bush's nomination was a foregone conclusion.

In this election, there was almost the exact same Democratic/Republican breakdown of voters in the hand counted municipalities as in the machine counted municipalities. That is not evidence for your contention that the hand counted (generally rural) municipalities were liberal bastions in which a black Presidential candidate would be expected to dominate.

[Muddy Waters Guitar]

94. Oh really? You mean, Obama has the "NH farm vote" locked up?

Wow, there's a new one. Seems to me that if anything, rural voters would be more inclined toward Hillary while urban voters (with lots of students and young adults) would be more inclined toward Obama. The only relevant difference here seems to be in the manner that the ballots were counted.

[JackRiddler]

84. So, you know.

You just know what the difference between the areas "should" be.

What do you fear in a hand count?

[truckin]

32. This is important information. It may not prove anything but it

raises valid questions. There should be audits of the machines and without them we can never be sure the votes were counted accurately. Someone referred to a mole and insinuated that someone would spill the beans if there was a big conspiracy. The problem with electronic voting machines is that one person from the voting machine company could insert a line of code to alter votes. That's all it takes. If no one hand counts the ballots to check the machines on a sample basis these questions will never go away.

This piece of data screams for an audit of the machines. Actually the audit should be done even when there is no evidence like this.

Percentage of Machine Counted Votes - Percentage of Hand Counted Votes

05.46% Clinton
00.00% Gravel
00.00% Other
-0.60% Kucinich
-0.72% Edwards
-1.24% Richardson
-2.89% Obama

That's an 8.35% difference between hand counted & machine counted votes when comparing Clinton to Obama.

[ProudDad]

34. The Deibold fix was in!

They're just practicing with this one...

Of course the ONLY "front runners" in both right-wings of the Big Business Party are tools of the corporate capitalist masters.

Heads they win, Tails you lose!!!

[HughMoran]

35. Totally meaningless since Hillay won the higher population areas that use machines

What a silly thing to post - I'm trying not to laugh out loud - this is embarrassing.

[mhatrw]

45. Only if you presuppose the validity of the machine count.

Your "argument" is ENTIRELY circular.

[HughMoran]

46. ok

:hi:

[suston96]

37. It's been already pointed out that your analysis is statistically flawed...

...in addition, how many machines were there in NH without paper trails? The answer is: NONE!

http://www.sos.nh.gov/voting%20machines2006.htm

That means there is a paper trail in ALL precincts. File a lawsuit to demand a recount in the machine tabulator precincts.

[truckin]

38. A paper trail is meaningless unless someone counts the paper. nt.

[suston96]

50. Paper trail meaningless.....Then count it. Tell the Obama campaign to demand a recount.

Uh, and why do you think that no recount has been demanded by the Obama campaign?

[mhatrw]

52. For what? To gain a single delegate at the expense of looking like a sore loser?

If the machines were rigged by a few percentage points yesterday, Obama wasn't the loser. Democracy was.

[Muddy Waters Guitar]

110. Right, that's the problem

The vote tallies are maintained essentially in secret on the data files held by Diebold and its associate companies, and easily altered there. There are videos that show how easy it is to alter the results, and especially when there's such a discrepancy with exit polls and paper ballot results, this breeds tremendous suspicion of a system that must be transparent.

[truckin]

64. I sent this email to Kucinich asking him to request a recount.

Dear Dennis,
You are the only candidate I have sent money to because you are not afraid to speak the truth. Because of this, I ask you to request a recount of the New Hampshire primary. Clinton and Obama’s results were wildly different than the final polls taken. In addition there were some strange numbers between towns where the ballots were hand counted and towns where the ballots were counted by scanners (see below). You are probably aware of these numbers but it is critical that you request a recount. If the hand counts agree with the machines we can congratulate Hillary and if they do not we can find out why.

This could be the most important thing you can do to reclaim our democracy and I thank you for your consideration.

[Bread and Circus]

49. I don't believe in a rigged election here... but that is just kind of scary.

[vee]

65. Compare the data

from the exit polls of each specific location (Machine and Hand counted areas)
to the data you have collected.

If the range of difference (8.35%)is similar, no fraud.
If the range of difference (8.35%)is off, fraud.


You should probably post the range of difference for the republican candidates.
something close to 8.35, no fraud. Something off, fraud.

[Maven]

66. This is nonsense and proves nothing.

For reasons that have already been pointed out on this thread, which you'll be sure to ignore.

[mhatrw]

71. I'm not trying to prove anything.

But you know what is really nonsense? Suggesting that a exit poll that is constantly adjusted to match the incoming returns somehow confirms the accuracy of the these returns.

http://www.pollster.com/blogs/looking_for_new_hampshire_exit.php

By the end of the night, the tabulations will be weighted to the official count.

http://www.pollster.com/blogs/exit_polls_what_you_should_kno_1.php

Once the polls close, the interviewer will attempt to obtain actual turnout counts, and if possible, actual vote returns for their precinct from polling place officials. One of the unique aspects of the NEP exit poll design is the way it gradually incorporates real turnout and vote data as it becomes available once the polls close. The exit pollsters have developed weighting schemes and algorithms to allow all sorts of comparisons to historical data that supports the networks as they decide whether to "call" a race for a particular candidate. When all of the votes have been counted, the exit poll is weighted by the vote to match the actual result.

[Maven]

75. Please see my post below

Your post was about the disparity between hand counts and machine counted votes. It has nothing to do with adjustment of exit polls.

[Muddy Waters Guitar]

97. It raises legitimate questions about validity, and that

is more than sufficient to demand further examination. With elections in a democracy, trust in the results has to be airtight. Transparency is at the heart of this. We're still reeling from the doubts and recriminations about 2000, 2004 and some Congressional elections in 2006. We can't afford another such debacle. This demands a closer look.

[Maven]

67. Here's a list of NH municipalities that use machines

http://www.sos.nh.gov/voting%20machines2006.htm

All others use hand counts.

You'll notice that the list includes every major city and town in southern NH where Clinton was expected to (and did) do well.

You'll also notice that the towns on the list in western NH where Obama was expected to win (such as Canaan and Hanover), he won with a considerable margin. In Hanover, Obama won by 58 to 26 (!). In New London, Obama won 51 to 27. In Sunapee, Obama won 41 to 30.

Now look at towns where Hillary won and there were hand counts (mostly southern NH). Strafford - Clinton, 37 to 31. Lyndeborough - Clinton, 40 to 32. Stratford - Clinton, 47 to 26. Hill - Clinton, 48 to 39.

You can cross-reference here:

http://www.politico.com/nhprimaries/nhmap-popup.html

There was no fraud.

[vee]

68. Then the level

of deviation should be the same, if the data is compared to the exit polls and rural/city folk
have the same level of honesty.

If the level of deviation is substantially different, something is wrong.

[Maven]

70. Those intent on pushing this theory have made clear

they have no respect for the exit polls.

So this is the last resort to counteract bunk like this OP.

[vee]

73. If the exit polls are

inaccurate, they will be uniformly inaccurate (there is no correlation between Machine/Hand counting
that would effect the accuracy/inaccuracy of an exit poll).

Thereby making a comparison of exit polls to the data from the Hand Machine counts legitimate.

[mhatrw]

78. Unfortunately, we are not privvy to any raw exit poll data, and

certainly not a municipality level. The only numbers that are publicly released are statewide percentages that have been adjusted to match the reported election returns.

[mhatrw]

76. Exit polls that are pegged to election results cannot be used to confirm said election results.

http://www.pollster.com/blogs/looking_for_new_hampshire_exit.php

By the end of the night, the tabulations will be weighted to the official count.

http://www.pollster.com/blogs/exit_polls_what_you_should_kno_1.php

Once the polls close, the interviewer will attempt to obtain actual turnout counts, and if possible, actual vote returns for their precinct from polling place officials. One of the unique aspects of the NEP exit poll design is the way it gradually incorporates real turnout and vote data as it becomes available once the polls close. The exit pollsters have developed weighting schemes and algorithms to allow all sorts of comparisons to historical data that supports the networks as they decide whether to "call" a race for a particular candidate. When all of the votes have been counted, the exit poll is weighted by the vote to match the actual result.

[mhatrw]

74. If you want to look at this anecdotally ...

109 of the 111 Diebold opscan municipalities have currently have 100% results posted on CNN. 66 of these municipalities went Clinton's way while 43 went to Obama. Interestingly, Clinton did not receive less than 26% of the vote in any of these municipalities, and Clinton received less than 30% of the vote in just 4 of these 109 municipalities. Clinton also received more votes than Obama in 20 of the 28 of these municipalities with the highest Democratic primary voting totals.

Currently, 117 of the 126 New Hampshire municipalities that count votes by hand have 100% results posted on CNN. Clinton received more votes than Obama in just 36 of these 117 municipalities, Clinton received more than 10 votes more than Obama in just 26 of these 117 municipalities, and Clinton received more than 20 votes more than Obama in just 18 of these 117 municipalities.

I'm glad you are so certain that the machine counts are accurate. I personally would like, a minimum, a truly random partial audit of all machine counted ballots required for every election. If this were implemented and the random partial audit matched the machine count, our certainty wouldn't have to be strictly based on anecdotes, MSM memes and blind faith.

[Maven]

77. .

"66 of these municipalities went Clinton's way while 43 went to Obama. Interestingly, Clinton did not receive less than 26% of the vote in any of these municipalities, and Clinton received less than 30% of the vote in just 4 of these 109 municipalities. Clinton also received more votes than Obama in 20 of the 28 of these municipalities with the highest Democratic primary voting totals."

Why does this fact alone make the data suspect? Maybe it's just that more people voted for her.

Again, you're not accounting for the assortment of voting machines to different areas around the state.

I am in total, 100% agreement with you about the problems with voting machines and the need for auditing. In 2000 and 2004 there was clear evidence that the numbers were manipulated, from exit polls to overvoting to vote "flipping." There just isn't that kind of evidence here, and suggesting there was fraud simply because you can't believe a particular candidate won the votes cheapens the effort to expose fraud.

[mhatrw]

79. You are misinterpreting my stance here.

There are only two voting methods in NH. Most of the smaller municipalities count ballots by hand. Most of the larger municipalities use a Diebold opscan machine to count the ballots. I trust the accuracy of the hand counts. However, the machine counts may or may not be accurate. The hand counted results match all the the pre-election polling within the margin of error. The machine counted results do not. The fact that the machine counted results match the exit polling results is meaningless because exit polling results are routinely tweaked to ensure such a match. Therefore, I believe some of the ballot counts generated by the opscan machines could possibly be erroneous. I would like a randomly selected subset of these opscan ballots to be counted by hand and compared against the machine counts.

I wouldn't bet on finding widespread/systemic machine count inaccuracy in such a manual audit, but neither would I bet against it.

[Maven]

81. And I gave you several examples (among many others) of hand-counted tallies

that break decisively for Clinton, contradicting the pre-election polling.

Perhaps Obama simply did better, overall, in smaller municipalities. You keep retreating to the macroscopic view as if it shows something is amiss. It doesn't.

[mhatrw]

82. Yes, there are several reasonable legitimate explanations.

However, 117 of the 126 New Hampshire municipalities that count votes by hand have 100% results posted on CNN. Clinton received more votes than Obama in just 36 of these 117 municipalities, Clinton received more than 10 votes more than Obama in just 26 of these 117 municipalities, and Clinton received more than 20 votes more than Obama in just 18 of these 117 municipalities. That's not exactly a groundswell.

Since no accurate pre-election polling was done on the municipal level, you can hardly make the claim that the positive results Hillary got in a few select hand counted municipalities were surprising.

[Maven]

83. Considering that quite a few of those municipalities have 200 total Dem voters or less

A 10 or 20 vote margin corresponds to a 5 - 10% difference. My point was simply to show that the results defied the pre-election polls considerably even in many hand-counted municipalities. It casts doubt on the proposition that Clinton won in machine-counted municipalities where she was expected to win simply because they were machine-counted. Likewise, evidence that Obama trounced Clinton by big margins in areas where he was favored despite machine counting casts similar doubt.

I guess we'll have to agree to disagree, because I simply don't see convincing evidence of fraud. What's unfortunate is that we have to pick apart every election like this because these machines undermine trust in our elections, and with good reason.

[ORDagnabbit]

88. easy to understand chart

http://www.dailykos.com/story/2008/1/9/181852/2086/306/433947

[sam kane]

87. Thanks, this is impressive work, and many appreciate it. n/t

[mhatrw]

91. kick. n/t

[Muddy Waters Guitar]

92. More evidence that there really was something fishy about

the vote tally in New Hampshire. So sad, we rightfully objected when 2004 was stolen from us, and now, seems possible that some within our own party are doing the same.

[EVDebs]

95. 517,226 total votes out of 850,836 registered voters = 61% turnout

"850,836 total registered voters -- 26% Democratic, 30% Republican, 44% Independent"

http://political.shoutingmat.ch/comments/2059829.html

This is the real story. Our so-called Democracy is being destroyed by couch-potatoes. Either that or Edwards is busting his ass off for people who can't make it to the polls to vote for him. Why bother ? The corporate candidates are sewing up the vote right now anyway, eh ?

333,610 registered New Hampshire voters didn't even bother to vote. Probably working on a second shift in that incredible Bush economic machine that runs 24/7.

Much the same happened in Iowa's caucuses where about 300,000 decided the state's 2 million plus vote for them.

[Occam Bandage]

96. This proves nothing, as they're from different districts. Nice try, though.

[Muddy Waters Guitar]

99. And why, pray tell, would those paper ballot-using districts mysteriously

have a greater Obama support level than the ones using the machines? The paper-ballot districts are rural districts, if anything one would expect them to be less inclined toward Obama than urban centers, with more young and college-age voters, Obama's natural mainstay of support. The only significant difference here seems to be in the manner in which the votes were counted, and that alone is enough to provoke enormous suspicion and skepticism.

[Occam Bandage]

102. Actually, Clinton won cities, and Obama did better in rural areas, according to the exit poll.

The "Obama does better in the cities and worse in the rural areas" CW does not at all apply in New Hampshire.

http://www.cnn.com/ELECTION/2008/primaries/results/epolls/index.html#NHDEM

[Muddy Waters Guitar]

104. But this is exactly the question, there was no such rural/urban split in Iowa

If anything, Iowa did much better in urban areas such as Sioux City or Des Moines than in the prairie precincts.

[Occam Bandage]

105. That's because the rural New Hampshire voter is completely different

than the rural Iowan. The rural NH voter is more likely to be a libertarian/independent/liberal, while the rural IA voter is more likely to be an evangelical or to favor their party establishment.

[mhatrw]

106. Why was the machine count Dem/Rep split the EXACT same as the

hand-count Dem/Rep split? Tell us again why rural NH folks are more likely to vote Obama.

[Occam Bandage]

107. Rural NH folks aren't just "more likely to vote Obama," they did vote Obama.

Are you challenging the validity of exit polling as well? If you are, then you're simply abandoning all reality in favor of conjecture. Exit polling is how we knew 2004 was fucked up.

I don't understand what you mean by Dem/Rep split; you haven't shown any data relating to that.

[mhatrw]

113. Exit polls are ADJUSTED to match the actual returns.

Read this and try to understand it, please.

http://www.pollster.com/blogs/exit_polls_what_you_should_kno_1.php

Once the polls close, the interviewer will attempt to obtain actual turnout counts, and if possible, actual vote returns for their precinct from polling place officials. One of the unique aspects of the NEP exit poll design is the way it gradually incorporates real turnout and vote data as it becomes available once the polls close. The exit pollsters have developed weighting schemes and algorithms to allow all sorts of comparisons to historical data that supports the networks as they decide whether to "call" a race for a particular candidate. When all of the votes have been counted, the exit poll is weighted by the vote to match the actual result.

Now read this and try to understand it as well. Overall, 54.8% of NH residents cast their primary ballots for the Democrats. 54.4% of NH residents living in the municipalities that hand count votes cast primary ballots for the Democrats. The overall Dem/Rep primary vote split was basically the same in the hand counted communities as in the machine counted communities indicating that the hand counted communities are not more liberal than the machine counted communities as some here have tried to argue anecdotally.

[Muddy Waters Guitar]

108. DKos has been blogging about this, more inconsistencies noted

Main problem seems to be the data being held in essential secrecy by a private company. Voting is supposed to be a public process at all levels.