Miami Herald article shows that Kerry may have won Florida!

[pointsoflight]

In an article published yesterday, reporters from the Miami Herald provided data from hand counts that they conducted in several northern "dixiecrat" counties in Florida. They concluded that nothing was awry, and in fact, went so far as to title the article "No flaw in Bush's state win."

Here's a link to a reprint of the story that doesn't require registration:

http://seattletimes.nwsource.com/html/nationworld/20021...

Here are the tallies for Union County:

Bush original: 3396
Bush hand count: 3393 (-3)
Kerry original: 1251
Kerry hand count: 1272 (+21)
Net change: Kerry +24

Here are the tallies for Lafayette County:

Bush original: 2460
Bush hand count: 2452 (-8)
Kerry original: 845
Kerry hand count: 848 (+3)
Net change: Kerry +11

It's a bit more complicated for the third county they looked at, Suwanee County, because they only report the totals for a hand count of "almost 60%" of the ballots.

Bush original: 11153
Kerry original: 4522
Bush hand count: 6140
Kerry hand count: 2984

In the original count, 71.2% of the votes cast for Bush or Kerry (n=15675) went to Bush. In the hand count, this drops to 67.3%. That is a significant drop. Let's translate that into numbers. If you take the percentages from the hand count and extrapolate, here's what you get:

Bush = 15675 x .673 = 10549 (loss of 604)
Kerry = 15675 x .327 = 5126 (gain of 604)
Net change: Kerry +1208

A switch of 1208 votes in a county with less than 16K votes cast is obviously huge. Now maybe there's a very large percentage of Bush votes in that remaining 40% that they didn't count, but we can't know that because they didn't count them. Which begs the question...why did they stop counting in Suwannee County when their tabulation of 60% of the ballots deviated so much from the original total? And without actually counting those remaining ballots, how can they possibly report that nothing is amiss when the data they have so far suggests a possible problem?

They conclude that there's "no flaw in Bush's state win." Sorry, but what I see is a possible gain of 1243 votes for Kerry from three small counties in which only 23627 ballots were cast. That represents about 0.3% of the ballots cast for Bush and Kerry statewide. If Kerry gained votes at the same rate statewide, he picks up nearly 400,000 votes and wins Florida.

Thanks Miami Herald, you just revealed to us in your hand counts that there's a possibility that Kerry won Florida.

[ClintCooper2003]

1. Brilliant analysis!

Send it to Keith! Keith! Keith!

[liberal al zib]

113. Yes Keith!!!

Keith Olbermann is the smartest journalist ever and with his 50 million viewers a day he will expose this fraud.

[DemocratSinceBirth]

125. keith olberman has about 200,000 l viewers...

where do you get 50,000,000 viewers?

[berniew1]

195. story on 3 counties in Florida is red herring; touchscreens were problem

The "audit" in 3 small rural counties is a red herring being circulated to quell concern over possible fraud in Florida. The 3 small counties in question showed no irregularities in pattern compared to 2000 voting and recent voter registration data. They were volunteered to be counted because it was known that they had no problems. Otherwise they wouldn't have been volunteered to have the counts.
But all of the major studies on unusual voting patterns in Florida
Univ. of Calif.(Berkeley) study, Princeton Univ. study and www.flcv.com/fla04EAS.html
found that the big unusual patterns in Florida were in the big touchscreen counties. This appears to be an attempt to take attention away from the touchscreen counties where anomalous patterns of votes was found by focusing instead on small optical scan counties. But it says nothing about the bigger issue, touchscreens in large counties. There were some larger optical scan counties that showed some unexpected voting patterns, but not these and not as much as the touchscreen counties.

[krkaufman]

120. Not to rain on the parade...

... but nearly all the votes in the Kerry swing are attributable to the extrapolation of the unfinished hand-count's in-process %'s, and so have little meaning.

[Wordie]

139. Could you explain that a little better?

[TruthIsAll]

167. My spread sheet says Bush dropped 2% from vote to count

Cty	BushOr	BushCt	KerryOr	KerryCt	Bush%Or	BushCt
Union	3396	3393	1251	1272	73.08%	72.73%
Lafay.	2460	2452	845	848	74.43%	74.30%
Suwanee	11153	6140	4522	2984	71.15%	67.30%

Total	17009	11985	6618	5104	71.99%	70.13%
						
Bush drops almost 2%						
Almost enough to reverse the state						
						

[rdmccur]

187. Madsen

Didn't Madsen say that most of the vote stealing went on in large counties in Florida?

[LIBERALNAVYVET]

176. FRAUD

Yes I do believe that was fraud in FL and OH. I can't wait for W to rewsign in shame

[ailsagirl]

189. And California, don't forget...

[katmondoo]

220. He has no shame

[NVMojo]

2. Holy Crap, Batman!!! This is amazing!

I just checked our local paper out and it had the same article with the same finals in it so this isn't a typo you caught!! Good job!!

[texpatriot2004]

3. Well, well isn't this interesting. Gee, why did they stop counting?

It wouldn't take long to overturn that state at a rate like that.

[RevCheesehead]

4. why did they stop counting?

You mean in 2000, or Nov. 3, 2004, or the Miami Herald?

answers: The Supreme Court, MSM, and the editor...

[texpatriot2004]

9. Hi Rev, thanks for the smile

I like your questions and your answers.

[Gman]

178. Same reason they stopped counting in 2000

[jsamuel]

5. COULD BE GREAT! - check which precincts they counted

[JohnnyCougar]

6. Send this to Keith!!!

[Karenca]

7. kick!

[politicasista]

8. Isn't it late for a recount?

If the truth comes out, Bush will leave the WH in disgrace.

[jsamuel]

11. There is always a contest of election

[Eloriel]

108. ?????

Not if state law doesn't support it. So what do you base your comment on?

[jsamuel]

10. Almost 600,000 swing to Kerry!

I calculated the numbers assuming all votes in florida showed the same change.

votes_here/total_votes
15,675/7,500,000 = 0.209% of the total vote

change_here/percent_changed
1,200/0.209% = over 570,000 votes to KERRY!

[Name removed]

105. Deleted message

Message removed by moderator. Click here to review the message board rules.

[KittyWampus]

109. True, The Numbers Switched To Bush Could've Been Greater In Counties

that had significantly more voters.

And it's not just switching votes from Kerry to Bush... it's switching from Kerry to Bush and third party candidates.

[LisaL]

12. One can't help but wonder-why did they stop counting

in Suwanee County?
Hmm...

[Stand and Fight]

13. You Guys!

Once more you delight me with your brilliance. Can't say that I'm very disappointed since I stopped lurking. You guys are great!!

[GetTheRightVote]

14. Spread the news, this is great, it looks like a Kerry Win !!!! Hurray.

[kerry2win]

15. ah now i see ...the exit polls were right. 50-49 Kerry

[ccarter84]

16. kickity kick kick

question, if you assume that the voting percentages stay the same as in the 60% sample, do you not even get a further repositioning of results? I might be wrong cuz its way past bedtime but I did
6140/.6 = 10233 bush votes - 11153 = 920 excess bush votes
2984/.6 = 4973 kerry votes - 4522 original votes = 451 votes kerry gain

add it up and you get 1371 for this little teenie district...ouch thats bad

...a new shiny for someone who is more awake than I who wants to explain it better

[politicasista]

18. How are we going to convince Kerry that he needs to contest the results?

He has been silent and hasn't been answering e-mails?

[pointsoflight]

22. You could do the math that way, too.

It gives you similar numbers. I went a different route since the article says they counted "almost 60%" of the ballots and we therefore don't have the exact number. In addition some of that counted sample would've gone to the third-party candidates, and they don't give those numbers either.

That's why I chose to extrapolate from the ratio of Bush to Kerry votes in the counted sample.

Either way you work the math, it points to a potentially large change in the vote count in Suwannee County. Can't help but think that the counters knew this, and that's why they stopped counting.

[LisaL]

24. Exactly. Why would they stop counting?

They got 60 % counted, and stopped? Why? Certainly not because they were getting identical results to the official count, since obviously Kerry/Bush ratio is not the same as in official count.

[alaintex]

100. One reason why they would stop counting is if a clear winner was determine

For example if they knew that 10,000 votes were cast, which ever candidate gets 5001 confirmed votes wins. If the other guy only has 1000 votes that means we’ve looked at 60% and there’s no reason to count anymore.

I’m not sure if that’s the case in this situation, it’s just one explanation why they may have stopped counting.

[Live Free Or Diebold]

17. Interesting.

That data is certainly interesting. But, maybe some of the 40% of the Suwanee precincts that they didn't count were in strong Bush areas. That could easily account for the 4% difference in the partial count.

The best thing to do would be to count all of the ballots - not just 60% - and see if that result holds up. That's what we really want, after all: all of the ballots to be counted. Why does the Miami Herald get to actually count ballots in some cherry-picked counties, while good patriots like Bev Harris can't even get election officials to do their legal duty and disclose public records?

[LisaL]

20. In the two counties that they finished counting in,

Kerry still picked votes and Bush lost votes. The third county, they stopped counting. Why? Was Kerry picking even more votes there so they stopped?

[BlueDog2u]

56. Your hypothesis makes sense of the information given/eom

[pointsoflight]

25. You're right, it doesn't prove there is a disparity in Suwanee.

Like you said, maybe a very high percentage of the remaining ballots would've gone to Bush, bringing the tally in line with the official results. Because they didn't count all of the votes, there's no way we can know for sure what those votes look like.

But my main point is this: Based on the data they have given us, you certainly can't conclude, as they did, that there are "no flaws in Bush's win in the state." If anything, the data they've uncovered so far does suggest a problem or flaw, and compels further counting.

What they've done here is find a several percentage point swing to Kerry, then turn around and say there's no change. That conclusion doesn't follow from the data.

(And of course their "no flaw" statement is also ridiculous given that they ignored the urban counties that statisticians pointed to as having the biggest anomalies.)

[Eloriel]

112. Have you written them?

Or better yet (or and), CALLED the reporters or even better the editors? And an LTE is certainly in order as well.

Great work.

[KnowerOfLogic]

215. Yeah, gotta wonder how they concluded "No problem." n/t

[BlueDog2u]

67. Good Point, But...

If we drill down to the precinct level data we should be able to resolve this question, or at least clarify it, ourselves. The data will either show that this explanation is a theoretical possibility, or it will show that the precincts are more or less uniform in their voting patterns, which will be almost as good as counting the entire county. I don't have time to do this at the moment as I am preparing a class....but I toss the idea out for some enterprising DUer. Surprise me with the results; I'll check back in a few hours.

[Liberty Belle]

19. Kick! How do we nominate for homepage?

[jsamuel]

21. There is a small link directly under the post

[Liberty Belle]

218. Do you mean the "alert" button?

I thought that was to alert for freepers and trolls!

[OrangeCountyDemocrat]

23. I'm Sure There's A Logical Explanation For This

Or NOT!

[pointsoflight]

26. If nothing else, it certainly motivates more counting.

Their conclusion is simply wrong given the data they have at the moment.

[aquart]

39. Did Bush lose and Kerry gain in EVERY handcount?

[pointsoflight]

43. Yes.

In both counties where full counts were done, Bush lost votes AND Kerry gained votes.

I still have yet to hear how machines can possibly come up with more votes than a hand count, but that's what happened in both cases for Bush.

[jsamuel]

27. GOSH! If they just would have finished counting, that would be solid PROOF

[LisaL]

29. Well, they obviously didn't want solid proof.

I still can't understand how they were even allowed to count, considering Gore couldn't get recounts in 2000.
But solid proof is not something they were going for.

[DoYouEverWonder]

50. At this point, anyone can come to FL and recount votes

because the election has been certified, the ballots have already been removed from their locked bags and are stored by the SOE in each county.

These ballots are now available for anyone with the time and money to view them. You are not allowed to touch the ballots, an election worker has to hold them for you, but anyone can pick a county and do what the Miami Herald just did.


Here's a link to the thread regarding recounting FL votes.

http://www.democraticunderground.com/discuss/duboard.ph...

[Live Free Or Diebold]

28. A hypothetical question.

A little hypothetical situation to consider.

Suppose that tomorrow, Katherine Blackwell said, "Hey everybody! After talking with my advisors, I've had a change of heart. We're going to start a recount in Ohio tomorrow!"

Now, being a conniving Repuke, he says "We're going to start in Cuyahoga County, in this special list of precincts that I just made up." His list contains mostly rich white neighborhoods that are very Republican. Now, you know that most precincts, of course, in the county at large are strongly Democratic, but he selects, say, the 20% that are most supportive of you-know-who.

Now, after all these Republican precincts were counted, suppose the results looked like this:

Cuyahoga County (partial recount results)
20% of all precincts counted
Bush - 65%
Kerry - 35%

Now Katherine Blackwell turns around and says, "Well, it looks like 65% of the voters support W. Obviously, the remaining 80% of the precincts must look like that too. So, no need to finish the recount. It's all over. Bush obviously won 65% of the votes in all of Cuyahoga County!"

Would you buy that? Of course you wouldn't. You'd see right through it. Right?

So, without knowing which precincts were counted in Suwanee County, Florida, does it make sense to assume that all of the uncounted precincts in the county voted the same way as the counted ones? Does the counted 60% necessarily reflect how the other 40% will vote?

[LisaL]

30. Nope. But why did they stop counting?

What, their poor little hands got tired? Their poor little eyes couldn't see no more? If they were supposed to show that Bush won fair and squire, couldn't they at least finish counting, considering their Kerry/Bush ratio was not the same as in the official results? For all I know, Kerry would have done even better if they had finished.

[Live Free Or Diebold]

34. Why don't you ask them?

I couldn't say why they stopped counting. Some possible reasons:

1. Lack of time. They were writing this article on a deadline and had to get it into print before the Suwanee counting finished.

2. Lack of funds. I'm sure they had to pony up some money to get these counts going.

3. Lack of volunteers. Ballots don't count themselves, you know (well, given the unreliability of the current Diebold opscanners, maybe they do...)

[LisaL]

37. And don't forget possibility #4.

Too many votes for Kerry that they couldn't explain away.

[jsamuel]

31. No, exactly, but you can't say that it supports their conclusion:no consp.

[Live Free Or Diebold]

32. That's true.

You're right: it doesn't support their (erroneous) conclusion that there was no fraud. Neither does it support the notion that there was fraud. It doesn't support anything. It doesn't even rate as an "anomaly".

[LisaL]

35. So, you tell us, why did they stop counting?

And why did they write an article saying there was no FRAUD?

[Live Free Or Diebold]

36. See #34 above.

Why they stopped counting: ask them. I'm sure they'll tell you (unless they do have some guilty secret. ) I gave some reasonable possibilities in my last post.

Why they wrote an article saying there was no fraud: Simple. Because they're dummies.

[pointsoflight]

40. That's right.

It doesn't prove fraud and it doesn't even prove there's an anomaly. A full count would be needed for that.

What it does show is that the Miami Herald report is complete garbage, though. Their conclusion that there is "no flaw in the state" is absolutely ridiculous. They didn't test the whole state, and that part of the state they did test currently isn't in line with the certified results.

The media has been ripping on us for making unsubstantiated claims, and here we have a set of reporters not only making unsubstantiated claims, but claims that are inconsistent with their very own data!

[rdmccur]

190. Concisely!

And (I posted this already) didn't Madsen say that most of the vote stealing took place in large pop counties?

[pointsoflight]

38. You're missing the point.

They concluded that nothing was wrong in Suwannee County (or in the whole state) even though the data they had in hand was not in line with the official counts. Their conclusion simply does not follow from their data.

How can one possibly count 60% of the vote, see a swing of several percentage points, then conclude "nothing's wrong, let's stop counting?"

As I said in previous posts, what I've presented here does not in any way prove that a full hand count in Suwannee County would show big differences from the official tally. What it does is show that the reporters were quite premature in their conclusions, and that we're getting very biased reporting from that newspaper.

It also compels further counting. How can you stop counting when the counts you have before you are not in line with the official tallies?

[Live Free Or Diebold]

42. Well, what are the numbers?

You can't compare 60% of the Suwanee precincts to all of the Suwanne precincts, for exactly the same reason that Katherine Blackwell couldn't compare 20% of Cuyahoga to all of Cuyahoga in my example. You're ignoring potentially important information by doing that.

If I counted the Presidential votes in Washington state, but left out, say, King County, the most liberal area by far, the percentages I'd get would be way off the reported totals, but I wouldn't say that was suspicious.

Here's the question I'll ask. Do the tallies for the counted Suwanne precincts agree (more or less) with the officially reported totals for those same precincts only? What are the numbers? If there's a 4% difference between those figures, then yes, I'll agree, this is big. Otherwise, no, it doesn't mean anything.

[pointsoflight]

44. Your still missing the point, though.

Even if we did a full hand count right now and showed there's no problem, it's still shoddy reporting for the Miami Herald to definatively conclude that there's "no flaw." How can they possibly say that?

I've said several times in this thread that this does not point to fraud and does not even prove there's an anomaly. My point is simply that based on the data the reporters have, they have no basis for arriving at the conclusion that there is "no flaw in Bush's win in the state." Yet they choose that exact wording for the very *title* of their article.

[Live Free Or Diebold]

45. I'll definitely agree with that.

Just counting a couple of northern counties and saying "no big difference" certainly doesn't prove anything about the existence of statewide fraud; the article is poorly titled. If they did an extensive audit in southern Florida, though, I suspect they'd be singing a different song.

[mirrera]

93. Reporters Bought and payed for!

Did you think they were investigating? Why? Because they said they were?

They were from the MIAMI Friggin HERALD??
Owned by which corporation?

It is like the letters you see in People magazine "How can you put Brittany Spears on the cover when people are starving!"
Answer...Because it is PEOPLE magazine!!

These were real hard hitting investigative journalists (sarcasm...)
Real fair and balanced... Just look at these heart felt quotes:

"People here traditionally register as Democrats to vote in local primaries, but they're very conservative," Union County Judge Dave Reimer said.

Election Supervisor Babs Montpetit was more direct. "People here are mostly fundamentalist Christians who work in the prisons," she said. "Do you think they're going to vote for the liberal senator from Massachusetts?"

Supervisor Glenda Williams greeted reporters at the election office in Live Oak: "Most people in this county are against abortion and gay marriage. So, they voted for Bush."

(HMMM I guess they couldn't find any quotes from people who think things smell fishy? I mean that is what the story is about! Isn't it? Getting to the TRUTH?? )

Or this unbiased observation:

Along the way were neighborhoods called "The Christian Village" and a warehouse called "The Christian Outlet." One billboard said: "There is life before birth."

(HMMM again, they hadn't made up their minds which way to spin this story, had they?)

Here is an excerpt that shows their intent to get to the bottom of this mess:

Reporters told Lafayette County Election Supervisor Lana Morgan the newspaper had come back to end the speculation.

(It is always good to "end the speculation")

"Good," she said. "You don't know how frustrating it is to convince people who are set on believing something — even if it's not true."

(WOOPS, they forgot to interview any of those unconvincible OTHER people!!)

Nice ending:

Morgan invited the reporters for an early Thanksgiving dinner in her office, but they had to return home, driving to the Jacksonville airport past Snoball stands, chicken farms and anti-abortion billboards.

FACTSthey checked 17,000 ballots
RESULTS: Less votes for Bush more votes for Kerry
"Conclusion: no conspiracy"

Love that down home feel!


PS my rage is spilling out at no one here, just this terrible situation that we are all in...
My son is 17, and I will break his legs before I let them take him!

[badc0der]

47. But how do we get that data?

Comparing precinct by precinct is obviously better than what we have but how do we find out which specific precincts the MH recounted? And even if we get that are the per precinct results for Suwannee county available somewhere?

One other note the article claims that they counted 60% of the vote not 60% of the precincts. Although the article is probably incomplete (insofar as it really is 60% of precincts as well) if the MH is take to mean a 60% sample from all precincts than the extrapolation of results is probably good.

[nodictators]

88. LFoD, you are confusing the issue. The M-H was looking for changes

in the count. So, it doesn't matter if the ballots that were uncounted by the Miami Herald were for Bush or Kerry. We are looking to see if the Suwannee County counts were correct. The M-H handcount found a statistically significant difference that showed Kerry was short-counted in the official results. It's the difference that matters here, not the counts, since we all know the official totals.

Since Kerry was shorted in the 60% of the ballots the M-H looked at, it is completely valid to assume that Kerry would continue to be shorted at the same rate in the remaining 40% that the scumbag M-H reporters didn't count. In fact, that would be the maximum liklihood estimate for that 40%.

Pointsoflight has a great find and analysis. Let's not poison his or her correct conclusions.

[The Doctor.]

122. But Those reporters quite Obviously set out to quell

the vioce of concern - so naturally they would pull a 'Blackwell' as you described.

Going to the Dixiecrat states - they thought they would get what they needed but had to stop short, just as you suggested would happen in Ohio.

I'm not saying I know what the results will be, but you've made the point for a full recount.

[Paligal]

33. Did you send this back to the Miami Herald? n/t

[pointsoflight]

41. No, I'd rather see them get criticized for their shoddy reporting first.

[Senator]

46. Wake These Morons UP!

We need to write to every paper that carried this nonsense.

Here's what I wrote to the Herald. I hope you join me.

To: [e-mail] [e-mail] [e-mail] [e-mail] [e-mail]
Subject: Herald's Fuzzy Math

Do you people own a calculator?

You falsely claim your count of Suwanee County ballots "nearly matched the county's official tally."

This is simply false.

Your count discovered that only 67.3% of ballots were for Bush and that 32.7% were for Kerry. This differs from the "offishyl" result by several percentage points.

Extrapolate your results to the entire county and you would have "found" (had you looked) a swing from Bush to Kerry of nearly 8% or
approximately 1200 votes.

And that's after examining only 00.12% of the votes in the state.

Hello!?! Is anybody home!?!

[rodeodance]

48. thanks for sending your calculations

[badc0der]

49. negative deltas more damning

I’ve been thinking of possible scenarios where the scanners/tabulators could count more votes for one candidate than a hand count would, especially given the reduction in invalid ballots. I can’t come up with any legitimate scenario in which that would be the case… maybe someone else can think of a way this could happen without tampering with or biasing the machines.

[Woody Box]

51. Mathematical analysis

Give these data to Professor Freeman. He will be very interested. A rough analysis shows that the original count and the control count for Suwannee County are incompatible.

Let's take the original count. 15675 votes, with 11153 votes =71.2% for Bush.

If we take a random sample (and this is what you do when you make a control count and stop when you've counted 58% of the votes), we can determine the mean and standard deviation for Bush's votes.

The mean is: (9124= total votes of the control count)

m = 9124*0,712 = 6496 votes. (This is what was to be expected in the control count).

The standard deviation is

sigma = square root (9124*0,712*0,218) = 37.6.

There is a 99% probability that a random sample lies between m-3*sigma and m+3*sigma. This is the so-called confidence interval. In our case this means that there is a 99% probability that Bush's votes in a control count with 9124 counted votes lies between 6383 and 6599.

Bush's actual vote count was 6140, however. This is far off. This is a deviation of (6496-6140)/37.6 = 9.46*sigma.

A standard deviation of 9.46*sigma means that the result is next to impossible and that you have to look for other explanations than random influence.

The most obvious explanation is that the original count was wrong. Is it therefore highly recommend to do another control count, but this time with 100% votes countes, not 58%.

[Woody Box]

52. Little mistake

Sorry, I made a little mistake, but this doesnt' change anything.

The standard deviation is

sigma = square root (9124*0,712*0,288) = 43.3.

There is a 99% probability that a random sample lies between m-3*sigma and m+3*sigma. This is the so-called confidence interval. In our case this means that there is a 99% probability that Bush's votes in a control count with 9124 counted votes lies between 6366 and 6626.

Bush's actual vote count was 6140, however. This is far off. This is a deviation of (6496-6140)/43.3 = 8.22*sigma.

A standard deviation of 8.22*sigma still means that the result is next to impossible.

[Live Free Or Diebold]

53. Your starting hypothesis is flawed.

There is no basis for assuming that the ballots counted were a 60% random sample of the whole.

A 60% random sample would require that they sample from all of the ballots in the county - for example, by taking a representative 60% sample from each and every precinct.

They certainly didn't do this. They just counted 60% of the precincts.

That is not the same as taking a 60% random sample, since a candidate's support by precinct is generally not a randomly distributed variable: for example, there are especially lopsided precincts which break 80-90%+ for one candidate or another. Leave a couple of these out, and whoops! Your sample totals are off by several percentage points from the mean of the whole.

This partial count in Suwannee is not indicative of anything besides being, well, a partial count. Which is not to say that there aren't lots and lots of valid indicators of election fraud out there: just that this isn't one of them.

[The Backlash Cometh]

64. You know what I find amusing about your response?

Gallop can call less than 1000 people in this entire nation and determine that Bush will win by a two digit percentile, and you, on the other hand, are asking for more than a 60% sample of a county before you accept what is no more than a projection which is valid enough to justify a full recount.

[bemis12]

68. Not really

They said it "almost matched". Polls are random, this was specific precints.

Depending on which precints you counted in Cuyahoga county Ohio, you could arrive at the equally mistaken conclusions that Kerry took 80% of the vote, or that Bush did.

[JBear]

76. Major Municipalities?

Does Suwanee County have any major liberal strongholds?? I think not. This is a rather homogenous county.....afterall, I am sure this is why they picked it.

(to the tune of Randi Rhodes's theme music)

Count the ballots, we're counting on you!
Count the ballots, a thing we just 'got to do!
Count the ballots, freedom is counting on you!

(oh I am getting sick!)

[The Backlash Cometh]

143. All Suwanne is saying, is give recounts a chance.

[ReneB]

66. He is right.. its flawed i think n/t

[Woody Box]

77. Not really

Even if the reporters reported the totals of some precincts - which is doubtable, see my post #72 - if the county has a homogenous structure, the randomness is still good. I think Suwannee County is such a county.

A standard deviation of 8.22*sigma can't be explained by a slightly skewed sample.

[nodictators]

95. LForD, this is about error rates in counting the votes. It is not about

"lop-sided" precincts or how many votes Bush or Kerry would get in the 40% of precincts that weren't reviewed by the M-H.

If the election officials had a given error rate in counting the 60% which were reviewed by the M-H, then it is perfectly logical and statistically valid to assume that their error rate would persist in remaining 40%.

In fact, there is probably is good reason to believe that they may have siphoned off Kerry votes and awarded them to Bush. And the slimeball M-H reporters realized that and scooted out of town.

[g9udit]

54. They should recount Jefferson, Liberty, and Calhoun county !!

Below is the Florida vote count broken down by county :


http://ustogether.org/election04/FloridaDataStats.htm

Check out the numbers at the above website for
Jefferson, Liberty, and Calhoun county .

All 3 have have a very high percentage of Registered Democrats,
but all 3 reported the great majority of votes for Bush.

Why didn't the Herald pick one of these counties, since
their reported results are very far off from what
you would expect of a Democratic county ?

[DoYouEverWonder]

71. We should recount Jefferson Liberty & Calhoun

We don't need them to do it for us.

Any citizen has the right to come to Florida and ask to look at the ballots. The SOE decide what the fee is and they will hold up the ballots for you to look at.

Call the local DEM party in these counties and ask them to organize recounts so we can find out what really happened.

The following link has links to every SOE in the State of FL.

http://election.dos.state.fl.us/county/index.shtml

[retread]

121. "...recount Jefferson, Liberty and Calhoun count!!" Before you do that

maybe you should come down here and get to know some of these "Democrats".

[BlueDog2u]

55. Very Good...

Should we media blast this? What do folks think? I think you just caught the media with its pants down.

[galadriel]

91. What BlueDog2u suggested

I agree with you, BlueDog2u. And by the way, this is my very first post. But I've been reading yours for a while.

[BlueDog2u]

135. Thanks, but I now suggest that before we blast it

We carry out the test indicated in my more recent post. Find out what the precint data are. That will help us to answer the question of whether these numbers could possibly result from variation in the precincts within the county. If we can safely rule out this explanation, then we have something explosively significant. If not, it still could be significant but we really don't know for sure.

[galadriel]

154. Thanks a bundle

for responding to a newbie. Just finding my feet around here.

[BlueDog2u]

222. You're quite welcome

And by the way,

Welcome to DU!

[tngledwebb]

57. Net gain and loss

reversed exactly?! Very interesting...

[Carolab]

58. This is JUST what I was saying yesterday

I calculated the net gain for Kerry equalled about 1/4% overall in those three counties. I was PISSED because the article made it seem that there was NO change--until someone posted the results. I bet they did a 0.3% calculation pretty much across the board. Nationally, how many votes do you suppose would that be? Enough to equal Bush's "mandate"?

[tngledwebb]

61. Gut instinct ( and other's research points the way)

says that the BIG mandate for Bush was exactly the high turnout numbers voting for Kerry, or ABB. BushCo evil-doers just flipped 'em, electronically or otherwise, and they are still laughing at all of us.

But not forever, and not for long...

[Eric J in MN]

59. Can we send this to the Florida Democratic Party? NT

NT

[mostly_lurking]

60. I'm sorry, but your analysis is somewhat flawed

The original article says:

"The Herald counted almost 60 percent of the votes in Suwannee County, where nearly 64 percent of the voters are registered Democrats."

"The newspaper's total from those precincts: 6,140 votes for Bush and 2,984 for Kerry, which nearly matched the county's official tally."

The key is the phrase "from those precincts." What the reporters apparently found is a match, precinct by precinct to the official totals (since the individual totals by precinct were available to them).

This wasn't a "random sampling" of the ballots but a complete total from a number of the precincts. Your analysis is flawed -- expecting a linear extrapolation from precinct to precinct is no more valid than expecting such a relationship to hold from county to county (or state to state, for that matter).

My point isn't to burst your balloon but rather to try and keep people focused on the real issues. These academic exercises using numbers posted online aren't convincing anyone in the MSM or in any position to make an impact.

[Karmadillo]

69. "Somewhat flawed" is being polite. Your restraint is admirable.

n/t

[Calvinist Basset]

78. Wait, guys.

As much as I appreciate a healthy dose of skepticism, I think you're both missing the major points.

1. Random or not, the Herald did not count all the precincts. Either way, we are left with a question about final results, and it is simply premature and wrong of them to conclude that no funny business took place.

2. The other counties, which were counted completely, showed vote gains for Kerry. Instead of verifying the official election results, this fact should compel the reporters to ask if this trend was consistent across the state.

[mostly_lurking]

81. Actually...

There is nothing wrong with sampling precincts. In fact, this is exactly what Nader is doing in NH. What is important is proving (or disproving) fraud involving the electronic voting systems. All you need to do that is to verify, by hand, the numbers tabulated by the machines in a closed system.

That means that taking samples of precincts throughout the state is a very good way to prove the point, one way or the other.

[Calvinist Basset]

224. Sampling may be okay.

However, it still doesn't answer how they can make the claim that there was no funny business when all of their particular handcounting reduced Bush's votes, and only Bush's votes.

[LisaL]

83. More votes for Kerry and less votes for Bush.

20 votes here, 20 votes there-it all adds up.

[mostly_lurking]

84. You are correct that it adds up

However, errors of this type have always (and will always) exists. They are caused by human error (usually) although can also be mechanical malfunctions.

If you assume no "fraud" then the errors will statistically cancel out (there would be errors in favor of both candidates). If fraud exists then the assumptions are invalid.

Proving fraud will take more than a few votes differing in each precinct unless EVERY case shows an advantage to a specific candidate.

[Calvinist Basset]

225. But that's just the point.

These numbers don't prove anything--they just raise suspicion. And suspicion should lead us to investigate more fully.

The problem is that the reporters' handcounting *all* showed a shift in Kerry's favor. Thus, we are compelled to ask the question whether these are indeed random errors.

Fraud may not be proven at this point, but it casts uncertainty about legitimacy. And that's good enough for me to have a full audit and recount.

[FloridaCrat]

86. No one has explained how Bush "loses" votes in a handcount

[Her Blondness]

151. With punch cards, chads fall out after the first machine count

[pointsoflight]

159. But Bush got *more* votes in the machine count.

[FloridaCrat]

205. Not in Union and Lafayette counties, his votes went down

Here are the tallies for Union County:

Bush original: 3396
Bush hand count: 3393 (-3)
Kerry original: 1251
Kerry hand count: 1272 (+21)
Net change: Kerry +24

Here are the tallies for Lafayette County:

Bush original: 2460
Bush hand count: 2452 (-8)
Kerry original: 845
Kerry hand count: 848 (+3)
Net change: Kerry +11

[Calvinist Basset]

227. Ooops.

I think you misread the previous post. The numbers show that Bush got more counts by the machines.

[eomer]

226. There are no punch cards in Florida anymore...

they've been outlawed.

These counties are optical scan.

[Woody Box]

74. I'm not admiring you - you're wrong

The key is the phrase "from those precincts." What the reporters apparently found is a match, precinct by precinct to the official totals (since the individual totals by precinct were available to them).

See post #72 and sho mew which precincts were chosen by the Miami Herald to count. I'm looking forward to your answer.

[mostly_lurking]

80. You make my very point...

Did you consider the (very likely) possibility that the absentee ballots were divided by precinct for the Miami Herald count? The web page you link to does not do such a division, so knowing for sure is again impossible.

That was my point from the start. Like many posters since 11/2, you are trying to draw inferences from online numbers w/o knowing the actual details. Fun, perhaps, but a dead-end.

If you remain curious about the precincts counted, why don't you contact the actual reporters and ask them?

[Woody Box]

99. Not so fast

Do you really think that, after doing their own count, the BoE members sit together and distribute the batch of absentee ballots over the precincts, just for the Miami heroes? What for? This makes no sense.

[mostly_lurking]

107. Why don't you ask them?

You can contact the BOE and the Miami Herald.

[tom II]

82. You just broke some hearts

Excellent response.

I am continually amazed by how determined people can "prove" anything with numbers, as long as you are selective over which ones they use and what they say they represent!

Please Please do not be sending this type of information to Olbermann. I watched him last night and he appears to be getting burned out with the chase. We need something more SOLID than just playing with selective numbers, extrapolating them, and presenting this as a case!

[Woody Box]

101. Oh, I think Keith would have fun to bash the Miami heroes

I think I'll send him the stuff

[ginnyinWI]

129. no I don't think he's burned out

If you read his blogs from yesterday you'll know why he was like that on tv last night: grief over the plane crash that killed his boss's son.

[pointsoflight]

160. You don't get the point.

I never tried to prove anything. I simply showed that it was the Miami Herald that didn't prove anything, as they were claiming.

[pointsoflight]

158. The only way that allows Bush to "catch up"

Is if the deliberately chose not to count the 10 or so precincts that went most heavily for Bush, and its these heavy Bush precincts that remain uncounted. Anything other than that, and the hand count can not be reconciled with the official count.

That alone would be cause for concern. Why would you cherry pick in such a way that you deliberately leave out those precincts where Bush won by the biggest margin?

[The Backlash Cometh]

62. Suwanneeeee, how I love ya, how I love ya, Mah Suwannee County!!

[BlueDog2u]

63. Here are the Suwannee data from 2000

Bush/Gore -- supposedly, anyway.

Suwannee 8,006 4,075

Also, a scenario: How did Florida swing to Bush in 2000? Doesn't it make incredible sense to pad those Panhandle votes? The Dixiecrat theory provides an immediate case of plausible deniability.

[retread]

123. Could be these Dixiecrats were unsure of what they were getting in 2000,

but liked what they saw! (no "compassionate conservative")

[kitp]

65. my email to them

prior to you guys doing this number crunching, I sent this email to the Herald (the Orlando Sentinel also ran the story and I wrote them as well).

I am writing regarding your recent articles dealing with the three county partial hand recount of ballots your reporters witnessed.

I have been following the problems with the our recent election in hopes of finding a way to eliminate all concerns and suspicions in this and future elections. I appreciate your newspaper being present at these recounts and publishing the results.

Anything that increases the transparency of our elections will improve the trust of the voters in that system.

I do have a few questions regarding your reports, though. In several studies, the UC Berkeley study for example, the three counties you reported on showed no signs of irregularities or problems. Therefore, a hand recount that is virtually identical with the reported results in these counties does not really address the issues raised by these studies.

A second issue is that in two of the counties the President lost votes in the hand recount. I do not understand how this is possible. A machine can (incorrectly) not count a vote for a variety of reason and a hand recount would add votes to the total, most likely adding votes for all candidates. However, if a hand recount decreases the votes cast for a candidate, that would mean that the machine counted a vote that didn't exist. I'm not sure I understand how this is possible and your article did not talk about this issue. Did the machine total include votes for which there was no corresponding ballot? Did the machine tally a vote when the ballot showed no selection? Either of these raises serious concerns about the machines' reliability. I have no concern that machines might miss votes, this can happen and hand recounts can find these votes and add them to the totals. I am concerned when the machine counts votes that don't exist.

The third issue is that, if our goal is to relieve the anxiety of those who feel that there is bias in the system, to show that regardless of what errors, glitches or malfunctions we may experience, it is systemic and affects all candidates, your report fails. Note that in each case of a hand recount, the reported results versus the hand-counted results favor John Kerry.

As I have been reading the reports of those who are concerned that there is a systemic bias in the errors, glitches and malfunctions toward the President in this election, I am afraid these results, though small, only feed that concern. It appears that there may actually be enough data to support the notion that there is a systemic bias and the results you reported do not challenge that, they support it.

My final point is that these partial hand-recounts in three counties not only do not address the serious issues raised regarding irregularities, as these counties were never in question, and not only do not address the possible serious issue of systemic bias, as the results seem to confirm systemic bias for every variance with hand-counting was in Kerry's favor, and not only do not address the possible serious issue of programmatic bias, as the machines counted votes for Bush that did not exist, but that these results are woefully inadequate to use as proof that "No flaw found in Bush's state win".

Again, I appreciate all efforts to make the 2004 election, and all future elections, transparent. I wish that you would continue this task by looking a little deeper into the issues raised and the problems reported.

[Turn CO Blue]

87. Great letter! Bush losing votes should be IMPOSSIBLE

yet the impossible has occurred: Bush lost three votes in one county and eight votes in another.

To look at the tiny percentage of change in total vote count in these two counties is beside the point. The machines cannot manufacture votes, or be able to count more votes for a particular candidate than exist upon substantiation by audit.

To your point, the machines have just been proven unreliable.

So, how does the machine count more votes for a candidate than exist? Any of the possible answers are unacceptable (and suspicious).

[klebean]

170. great letter - cc to KO as "talking points" n/t

[Iceburg]

70. Brilliant PoL ... this may be one of the most significant

findings to date. You are a genuine American hero.
I will write a letter to the Miami Herald later tonight (that's if I don't get arrested at the protest in Ottawa, Canada today)

[Woody Box]

72. Suwannee County precinct-by-precinct data - interesting

Look at this here:

http://www.suwanneevotes.com/new_page_3.htm

You see that the total votes vor Bush and Kerry (11145 vs. 4513) match the official votes quite good (11153 vs. 4522).

The columns show the results for each of the 16 precincts. The column left of the totals lists the absentee+early votes.

So how did the Miami heroes get their data? Did they simply add the totals of some of the precincts, as some are suggesting?

This doesn't add up.

Look at the Kerry votes: He got 1512 absentee votes out of his total 4513 votes. This leaves 3001 "real-time" votes. In the incomplete Miami Herald count Kerry got already 2984 votes. This almost equals the real-time votes.

So either

- the Miami Herald count included ALL of the 16 precincts and neglected the absentee votes. But this doesn't add up with the Bush votes - he gets 7524 in the precincts, but got only 6140 in the Miami Herald count. Not possible.

or

- the Miami Herald included the absentee votes and some of the precincts. I picked the most pro-Kerry precincts until they added up to the Miami herald count roughly (precincts 1, 2, 11, 12, 13 + absentee votes, yielding 2975 Kerry votes). If you add the respective Bush votes, you get 6283 votes - this is a difference of 143 to the Miami Herald count, so this is surely not the sample they've chosen. All other combinations of precincts work even worse.

Conclusion: No combination of precincts, with or without the absentee votes, matches the numbers of the Miami Herald.
(doesn't have to match it exactly, but at least up to a few votes).

The theory proposed by some that the sample of the Miami Herald is not representative because it excludes certain precincts is no longer valid.

The statistical absurdities continue to exist.

Send this data to Professor Freeman. He'll have a nice time

[BlueDog2u]

138. Nice work...that's what I suspected

I think we have a really strong case here, for empirical proof which backs up the statistical modeling of TIA and so many others here.

[3 DanO]

211. Precincts 1, 2, 5, 6, 7, 12 + absentee -> MH count (almost)

When compared to the unofficial precinct tallies, the hand count of these counties add 6 votes for bush and 3 votes for kerry.

My original reply was:
Precincts 1, 2, 5, 7, 12, 15 + absentee -> MH count
which added 2 votes for bush and 9 votes for kerry. The revised set above is a better match.

[3 DanO]

216. never mind

If only I had read the rest of the thread I would have seen that this was already resolved.