Interesting Results From My Home County (Pinellas, FL) and Home Precinct

[bj2110]

Please excuse the long post, and I know I'm new here, but I had to get this off my chest.

I've been pouring over a lot of numbers in the last week or so, trying to disprove the whole voting "irregularity" issue. This has been my approach, to disprove it, as should be done with any real hypothesis. With my latest escapade, I decided to examine my local votes, both county & precinct, including my own and my wife's.

I really don't make any conclusions here, they're just observations, but important observations, nonetheless.

The thread concerning the NC early/absentee vs. election day consistency in all races except the President & Senate inspired me to perform a similar analysis on Pinellas County, but with just the Senate & President races. Pinellas County uses electronic touchscreen voting machines by Sequoia.

What I found:

President
Early+Absentee (Approx. 1/3 total vote)
Bush 76582 49.1%
Kerry 79523 50.9% (+1.8%)

Election Day
Bush 149045 50.5%
Kerry 145844 49.5% (-1.0%)

This shows a 2.8% Bush swing from pre-election day votes to election day votes. Not a huge #, but it doesn't really have to be to make a difference, right? It certainly seems that in such a large county, assuming a random distribution of data, that 1/3 of the sample would be little closer to the other 2/3. Maybe <1%? I'd like to see a statistical analysis. Yes, the data is not randomly distributed, as it has been pointed out that the Reps generally have a higher representation in the absentee votes (military, etc...) and the Dems may have had a predictably larger early turnout. In Pinellas, these were both true, Kerry was +7898 votes (55.7%) in early voting, and Bush was +4957 votes in absentee voting (52.9%). Wouldn't these be averaged out over such a large sample (~150,000 votes), creating a much smaller MOE than exit polls (~2000 votes)? Even more support for a solid distribution of this 1/3 of the vote is in the Senate race:

Senate
Early+Absentee (Approx. 1/3 total vote)
Martinez (R) 68639 45.7%
Castor (D) 81659 54.3% (+8.6%)

Election Day
Martinez (R) 128960 45.8%
Castor (D) 152710 54.2% (+8.4%)

This shows a 0.2% Republican swing from pre-election day votes to election day votes.

Seems more logical, doesn't it? A sample of 150,000 countywide votes should match pretty evenly with the other 300,000, shouldn't it? It sure does in the Senate Race.

In neighboring Hillsborough county (Also Sequoia touchscreen), the resulting swings were 2.6% Bush and 0.9% Martinez. Pretty similar results, huh? Same voting machine, too.

Next, I looked just at my precinct, #650 in Pinellas. First of all, I've heard reports of 2 voting machines at a precinct in Ohio that was to serve thousands of people. In my precinct, we had 3 voting machines, serving 223 people. Who decides these things? We literally had zero wait time. There was only one other voter in the room when we got there. We were in and out in less than 10 minutes.

The results in our precinct were Bush - 129, Kerry - 92, Other - 2, with zero undervotes. Senate results were Martinez (R) - 113, Castor (D) - 90, Other - 2, with 13 undervotes. Let's assume for a second that everyone who voted Martinez also voted Bush and everyone who voted Castor also voted Kerry. Yes, this may be a flawed assumption, but I surely can't think why voters would tend to
switch any one direction more than the other between the two races.

With this assumption there were 18 people who picked either the Republican or Democrat who didn't in the Senate race (13 undervotes + 5 less Other votes). And these 18 people voted 16 - 2 for Bush? Come on.

This then got me thinking about another thread I recently read here. A programmer hypothesized that the ~4200 vote foul up in an Ohio county might be due a rogue programmer making a logic mistake. This mistake was instead of switching every xth Kerry vote to Bush, it added the running total of the Kerry votes to Bush's total. Let's assume that in my precinct, every tenth Kerry vote on election day (assuming the 1/3 - 2/3 early/poll split) was switched for Bush. This would put the actual final tally at about Bush - 122, Kerry - 99, Other - 2. The 18 people who picked Rep or Dem for President, but not for Senate would have voted 9-9 Bush/Kerry. Doesn't this seem a little more reasonable?

IF you take this hypothesis of switching every xth vote, even just on the touchscreen machines (Forgetting the opscans for now), it doesn't take very many votes, i.e. pretty long spaced intervals, say every 10 or 20 or 30, being switched from Kerry to Bush to explain the discrepancy between the exit poll and reported #'s. There are 4358 precincts that used touchscreens in FLorida. This means there only had to average 43 switched votes per touchscreen precinct to account for Bush's victory margin. With an average of around 1000 votes per precinct, doesn't really sound that impossible. Nor would it be easily seen as an outlier in any county-county or precinct-precinct analysis.

Food for thought...

[Eloriel]

1. Very interesting, thanks

That thread you mentioned (the rogue programmers' error) is this one, if anyone wants to read what I think is a superb thread:

The 3893 extra Bush votes in one Ohio Precinct--by WhiteKnight1
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x48519

[bj2110]

9. Thanks for the link, I knew I forgot something...

[NewYorkerfromMass]

2. The Senate/Pres. switch in your precinct is very suspect

but besides conjecture and speculation, how does one prove a pre-programmed switch?

[bj2110]

3. That's the real questin, isn't it?

I guess the best way is to somehow get a copy of the source code, which I'm sure is practically impossible, isn't this what BBV has been trying to do for some time now?

Aside from the code itself, I guess you'd have to have some sort of legal testimony from everyone in a particular precinct as to which way they voted and compare it to the recorded results. Of course, this doesn't prove tampering, it only proves a "glitch." But, then this would be enough to open the door on the source code, I imagine.

[jsascj]

5. Maybe some of the purged CIA folks

can help us with some of this stuff. They rig elections, right?

[bj2110]

7. Yeah, now their purpose has been served...

And they're no longer needed.

[jsascj]

14. But since they know the tricks of the trade

Maybe they could be coerced to HELP!

[Lil]

4. Thanks for your info!! Which precinct are you in? :-)

[bj2110]

6. Precinct #650.

I think basically, our precinct is a condo/apartment complex & a mobile home retirement community. The average age should be pretty high, but there's a lot of young people like me in the condos & apartments. I'd guess the demographics are fairly large female to male, and least 95% white.

[Lil]

8. Thanks, bj2110.

edit: oops. I see it in your initial post now. Sorry.

[bleever]

10. Very nice work. If all politics is local, then political theft is...

[milkyway]

11. Skimming a fixed percentage off the top of Kerry's votes and giving them

to bush is the most likely way they would fix the election. People are focused on the big "mistakes" where thousands of votes were messed up (these could, however, be indicating that someone has been messing with the vote tallies, but they screwed it up while doing it).

But the fix would be much more subtle, such in the way you suggest. The best thing about changing a relatively small percentage of the votes would be that it falls within the realm of believability. Thus, we've had two weeks of mainstream discussion about how "moral" values won the election, and what can the Democrats possibly do? (I would love to see the corporate media have to retract every single bit of crap they've come up with to explain the election when fraud is proven.)

Maybe the best--and only--chance of nailing them with hard evidence is that Bev Harris's audits turn up something.

[NewYorkerfromMass]

12. Also- it would not be believable to most people

MOst people cannot comprehend that such a blatant vote stealing scheme could be rigged so easily.
Look at all the conspiracy naysayers.

[bj2110]

13. A simple plan is key to such a large operation...

I'm sure this quote isn't right, but here goes anyway:

"You know what the best part of this plan is, Dude? It's simplicity." - John Goodman, "Big Lebowski"

And, I think I disagree, this hypothesis is so simple and easy to execute, heartland Mr. America has to be able to understand it. Now we just need to:

a) Prove it definitively (Or prove something else definitively)
2) Get MSM to actually report it.

[Hobbes199]

19. Agreed

I think the glitches that are noticed are only the slopy mistakes that were made, but the successes have to outweigh them nationwide. Maybe not noticeable at all in WA, but enough to throw it in NM.

[m berst]

15. if we were talking about banking...

If this were something relatively mundane like routine banking procedures, the combination of "it could be done like this" and numerous unexplained statistical anomalies would be more than adequate evidence to question the entire operation and would justify a complete audit and thorough investigation.

In fact, without that set of standards for evidence and cause it would be just about impossible to ever find or stop embezzlers.

Imagine if the suspected clerk at the bank said "where's your hard proof??" in his own defense, and the customers in the lobby all joined the chorus "get over it! Let's move on!" and the bank investigators were called "conspiracy theorists."

[bj2110]

16. Great analogy. Thanks exactly what we're dealing with here.

Except somehow are respected new media are playing the part of the bank customers waiting in line for their turn with the suspected party.

[milkyway]

20. Yes, great analogy. In another thread someone mentioned the exit polls

would be a huge red flag if this were a tax investigator looking through returns. The exit polls scream out that the vote counting needs to be thoroughly audited.

[bj2110]

21. It's important to call it an "audit". Not a "recount."

A recount implies that the votes written to memory (for lack of a better word, if in fact there even is a "memory") will be counted again to ensure they were added up properly. What if they were not recorded, written to memory, or encoded properly? A recount does no good, unless there is an auditable paper trail or some other tangible evidence of a person's vote. "Audit" is the word. Audit the process, the process creators, the machines, the code, everything.

When you make a deposit in the bank, you can get a written receipt with a stamp on it. Then, miraculously, your account balance total changes online within minutes, hours, whatever. If somehow, it added it wrong, or someone hacked into your account, you would know it, and have a receipt, or something of the likes, to prove it. Why can't our election system grasp this basic idea?

[milkyway]

22. And, coincidentally, most ATMs are made by Diebold. They know what

should be done to make elections secure and fair--and that's exactly why they don't do it.

[sandnsea]

17. Warren County OH

There, the increase of votes was 1.3% for both Bush and Kerry over 2000 numbers. That's the county they closed to observers. What are the odds a county would have the exact same increase of voters for both parties? It amounted to 5,000 more for Kerry, 18,000 more for Bush. I still can't figure out how they would match that to poll books, as I discussed with someone else earlier. But it still strikes me as odd.

[bj2110]

18. That doesn't seem all that odd to me...

It just means the population growth from 2000 generated voters that were equally inclined to vote one way or another as in 2000. The fact that it is the exact same % is a little weird, but that could just be significant digits / rounding.

[althecat]

23. Haven't read this yet but I will

I know that Pobeka reckons Pinellas is peculiar in some way too.

He has an amazing graphic to show how.

al

[bj2110]

24. Please post a link to this graphic if you can...

[althecat]

27. It's here

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=201x3372

cheers

al

[BeFree]

25. "...these 18 people voted 16 - 2 for Bush"

C'mon is right. There's no way 16 people out of 18 would have not voted for senator and made sure they voted just Bush. That's 1 out of every 14 voters not voting in the senate race.

If that sample were spread out over the whole election in your precinct, bush would have gotten many more votes, about 198. That's rather odd.

[bj2110]

26. Odd is right.

Makes more sense for the vote to be at most 11 - 7 based on the total precinct. I can't think of any reason why non-votes for senator turn into votes for Bush. Maybe the uninformed (people who don't know what a senator is), tend to vote for the incumbent? I don't know, seems a little far-fetched.

So, the real question is, in my local precinct where did these add'l 6-8 Bush votes come from? Answer that, and we have a potential answer to our fraud question.

[robicat]

29. The obvious answer is ....

...that people who voted for the Dem senator and Bush, account for it. The anomaly you highlight falls methodologically from excluding this possibility. (I think) That doesn't mean your wrong - just that your not going to prove it unless you prove your premise - ie that no one voted for the Democrat senator and Bush.

The gap between the early votes and the polling day votes is a stronger argument for shenanigans IMO, and reflects what someone observed in NC.

[robicat]

30. I think the result would hold if ...

..the 18 stray votes split evenly and 7 people voted Dem in the Senate race and Bush for president.

If they split 10 - 8 for the democrats then it would take 8 cross voters etc.

[bj2110]

31. No, it most certainly includes this possibility, in fact...

<"Yes, this may be a flawed assumption, but I surely can't think why voters would tend to switch any one direction more than the other between the two races.">

Yes Kerry voters could have voted for Martinez & Castor voters for Bush, but why would one group be more prevalent than the other? I can't think of a reasonable answer to that...

[liquiduniverse]

28. Another interesting note about Pinellas

Last week, a group of citizens went to one of the voting offices in Pinellas county to test out some of the voting machines. They were given a certain amount of time to test the machines. What is a little odd is that the voting office didn't allow the press to be present at the test and they didn't allow video documentation. When one of the group asked why he couldn't tape the test, the office refused to give him an answer. This was reported on WMNF radio.

Another "odd" thing that happened in Pinellas is that of all the precincts on the beaches, only two went for Kerry. One of those two was in Treasure Island, and I know they were only given 5 voting machines, which made for extra long lines, while neighboring precincts were given 10.

[bj2110]

32. Can you post a link to some reference of this organization?

I'd like to contact them and see if I can help...

[liquiduniverse]

34. WMNF link

I'm not sure if they consider themselves an organization.

You can listen to the Nov. 10 broadcast recording at http://www.wmnf.org/programming/daily.shtml?ShowId=242 (scroll down towards the bottom of the page)

This story can be heard a little bit before the halfway mark in the broadcast.

[bj2110]

35. Thanks. I'll see what I can find...

[bj2110]

33. And my pro-Bush precinct was given 3 for only 223 people. I had no wait.

[lostnfound]

36. A question about Pinellas precincts..

I really want to do a precinct-level analysis comparing results in 2000 to 2004 for Florida, and particularly for Pinellas,Hillsborough, Indian River and Palm Beach, but unfortunately the number of precincts has changed substantially during that time period.

The Pinellas County Dept of Elections website has a nice map of current precincts, but I need to find out what the lines of the old precincts were (in 2000), so that I can somehow correlate the precincts.

Anyone know where such a thing might be found? Or anything that defines how the new precincts were derived from the old ones? That would be incredibly helpful to supporting fraud allegations for these critical Florida counties.

[bj2110]

37. I know the current Pinellas Co. SoE, Deborah Clark, took over in 2001...

but I would imagine that the office has records on that sort of thing. Have you tried calling the SoE office? I'd love to help you on this one. I work from home, and live in Pinellas, it's easy for me to spend some time with this. I know Pinellas, Hillsborough & Palm Beach has posted the 2004 vote by precinct, has Indian River? Why these 4? Why not add Broward, Volusia, Miami-Dade, Orange, Brevard, Duval, and/or of few of the supposed "Dixiecrat" counties in the panhandle?

[lostnfound]

39. I have 2000 and 2004 votes on all of them

But I want to see the same data at the precinct level, and since the precincts have changed, I can't correlate the information.
These 4 counties all used the same voting machines and software. I noticed that the growth in votes in these 4 counties line up in a fairly consistent 9-11% offset from an 'even growth' line over the last election, as shown below.

If similar patterns exist at the precinct level, I think it would be very interesting. I just need a map (or at least descriptions) for all precincts in 2000.

[bj2110]

40. Interesting. If the same 9% - 11% increase exists at the precinct level...

we're looking at seriously consistent inconsistencies. Now that would be a tell tale sign.

I'm looking at a county by county analysis similar to the quick one I did with my precinct. Very intriguing so far. In some counties, the additional republican voters outnumber the additional total voters. This means that not only did Bush get every single one of the new votes in these counties, he took some away from Dems on top of that. This is just the tip of this iceberg, though. I'll post more later.

[lostnfound]

44. I found maps to be able to compare at the house district levels..

which might be even better since it won't be so overwhelming (since each house district combines multiple precincts--I think there's currently 7 house districts in Pinellas County).

Pre-2002 districts:
http://www.flsenate.gov/data/redistricting_data/current_maps/House_population_deviations.PDF

This page breaks down the districts into precinct numbers:
http://www.votepinellas.com/content.aspx?section=110700_RESULTS

Gotta go to sleep now but I'll work on it tomorrow night.

[bj2110]

47. Kudos. Stay with it. Let me know if I can help.

[bullimiami]

38. historic and other perspectives on pinellas county

These are the historic voting trends for the last 20 years in pinellas.
+ republican
- democratic
for 92 and 96 i use rep+perot vs the dem vote.


1984 1.87 pretty republican
1988 1.39
1992 1.62 an bit of abberation because of perot maybe
1996 1.02 almost dead even
2000 -1.09 democratic vote up by approx 4.5%
2004 1.00 following the trend I would have expected the dem to be up by maybe 6-8%
like maybe 54-46.

All across florida bush has reversed or halted the inexorable slide toward voting democratic. I have to wonder if it was Bush with his popularity and fantastic approval rating. Or was there tampering like the exit polls indicate?

Another red flag for me is Betty Castor getting 54% of the vote. Pretty much what I would have expected Kerry to get. It could be explained by this being her neck of the woods but I think republicans and democrats should not have crossed over by almost 5%.

Pinellas is an 85% white county. Pinellas has touch screens but I dont know if the whole county has them by now. I assume so.

For me the results are suspicious. With the touchscreens I suppose there may be no way to investigate this one.

[Zan_of_Texas]

41. If this chart is right

http://ideamouth.com/floridacounties.htm

Then I would focus on these optical scan counties:

Clay Okaloosa Alachua Escambia Polk Duval

plus
Hillsborough (shown on a different chart at the site, used Sequoia electronic voting)

If someone got control of these 7 county tabulating machines (avoiding completely EVERY county that got publicity in 2000), and subtracted 13% of Kerry's votes, and turned them in to 13% additional Bush votes, that's more than the margin of victory. Less than 28,000 votes flipped in each of 7 counties would flip the national election (without even addressing Ohio and the rest).

[RageKage]

43. Have also noticed strange numbers in Pinellas.

I've been wondering about this county. (which went 50.3/46.4 for Gore in '00) Consider the following:

Change GOP registrations '00-'04: - 8,578 (decrease of 3.57%)
Change DEM registrations '00-'04: +12,668 (increase of 6.01%)
Change IND registrations '00-'04: +20,310 (increase of 17.59%)

Change in BUSH vote '00-'04: + 40,861 (increase of 22.1%)
Change in DEM vote '00-'04: + 24,830 (increase of 12.4%)
Change in NADER vote 00-'04: - 7,620
Change in IND vote 00 -04: - 1,186

Now if we make the following hypothetical assumptions:

1) The votes that NADER lost went to Kerry; and other IND canditate losses to Bush;
2) Everyone who voted BUSH in 00, and voted in'04, voted for Bush;
3) Newly registered voters turned out at the same rate as as overall turnout in the county (77%); and voted according to party.

then either of the following would also have to be true:

A) BUSH takes 100% of the newly registerd Independents who show up at the polls (18,800) AND 2,924 voters switch from Gore to BUSH

OR (to put it another way)

B) BUSH takes 50% of newly registered Independents (9,400) AND 12,324 Gore voters switch sides. (slightly less than the total increase in Democratic registrations!).

----------------------------------------
I know it doesnt prove vote manipulation, but it really does not seem likely or realistic... unless of course there is something particular about that county than I am ignorant of (not at all improbable)...


All stats are from http://election.dos.state.fl.us/

[imenja]

42. Assumption regarding senate race isn't valid

I don't think you should assume everyone who voter for Kerry voted for Castor and those who voted for Bush voted for Martinez. Obviously the Castor-Martinez race was closer than Bush-Kerry. Secondly, the fact that Martinez is Hispanic is a particular impediment to straight party voting. Some Latino Democrats likely voted for Martinez and white Republicans may have voted for Castor because they wouldn't vote for a Latino.
I can't comment on the numbers. I'm retarded in that area.
Sure looks like you're working hard at it. I appreciate the effort.

[bj2110]

46. Well, the hypothesis is that the Bush-Kerry race really wasn't any closer

I agree that is more likely than not that plenty of people voted across party lines between the President and Senate race. But to believe the posted numbers, you have to believe that they did so in a grossly exagerrated manner in favor of the Castor/Bush ballot in lieu of the Martinez/Kerry switch. I don't buy that.

I'm working now on a similar analysis by county vs. the 2000 race, i.e. how many new/available votes were there in the 2004 election. This has similar assumptions, i.e. all bush voters remained bush voters, gore voters became kerry voters, other party voters in 2004 also voted other party in 2000, and there was no difference in voters switching from Gore to Bush as from Bush to Kerry. The kicker, however, is that you'd expect these new/available voters to follow the trend of the county 2004 results to some margin of error. But the trend is significantly in favor of Bush. And I do mean significantly.

[floridadem30]

48. Take a look at senate race in Jefferson county

It is similar to what you have posted, but we lost votes for president.

[floridadem30]

45. As you start checking other counties you'll find the same.

Welcome to DU