Hand-counted ballots mean loss of under/over votes.

[Kelvin Mace]

One of those things no one thinks about.

With OpScan, the machine can be programmed (and in some states are required) to reject under/over votes at the time the ballot is submitted (while the voter is still present). This means that the voter has the opportunity to CORRECT his ballot and have his vote count.

If you eliminate OpScan and rely on handcounts, you will lose these votes, since the precinct official will not be allowed to "read" your ballot and point out these problems. Your vote will be LOST.

The percentage of vote "spoilage" of this type in elections is significant, and certainly made a difference in 2000. Court battles about over/under votes are frequent and quite partisan.

Will the chamnpions of hand counting tell me how they overcome this critical problem?

[MissWaverly]

1. well, I can think of only 1 thing

When I went to vote for Kerry in 2004 in Baltimore, MD, the machine defaulted to George W.
Bush 5 times, was that a over vote or an under vote. Did I really get a chance to vote at
all; I don't think so. So I can do Jury Duty every 11 months, pay the flush tax, pay a tax
of $3.50 to use my cell phone which is the just because we need it tax, but I can't vote.

[Kelvin Mace]

5. To clarify

I am talking about OpScan, not TouchScreen. I have stated repeatedly I would ban touch screen if I could and I intent to continue to work toward banning them.

How will we recover lost under/over votes with hand counting when election officials (rightly so) can't examine the ballot?

[MissWaverly]

23. Okay, I think it's doable but ballot boxes should be transparent

that way people could see if they were empty before op-scan ballots were run through like at
the start of the voting day. It doesn't make much sense to have these machines if they are
pre-loaded with votes at the start of the polling session. Like every vote counter could be
set to start at a certain number that would be in a sealed envelope at the polling place like
2061, 2062, 2063, that way if there were votes out of sequence they would show up as bogus.

[Kelvin Mace]

25. At the start of election day

a zero tape must be run on the OpScan, i.e., a print of of the current count total of votes stored. If it is anything other than ZERO, election officials must get involved with reps from both parties to zero the machine before the first ballot is inserted.

Procedures differ from state to state, but the door of the ballot box is locked, and a seal placed on it to prevent tampering.

People have objected to transparent ballot boxes since it allows people to see their votes. (not saying the are right, just passing along the complaint), so that will get the transparent box shot down.

At the end of the day, the ballots are checked, the seal is broken (with witnesses present) and the number of ballots are checked to ensure they agree with the eletronic vote tally of total ballots cast, if not, we have a problem and a fracas ensues. Cue media and lawyers.

[MissWaverly]

29. but I saw on tv where the seals were broken by poll workers

due to a "Jam" they didn't know what to do they are untrained, etc. Also who is to run this
qc check before the vote starts. Is it the company tech or the old people working the poll.
As for my wild ideas of transparent boxes, I saw that they were transparent in the land of the
purple fingers. I think it is good you are working on this but I am a little nervous after losing my vote once & losing my efforts to have my state assembly vote down Diebold. Another
idea I had is why don't you offer the voters a voluntary verification ballot. They could place
a box in the polling station and let people but a ballot there to help verify that the machines
were actually working properly. These would be counted up to see total votes for the candidate
at the top of the ballot.

[Kelvin Mace]

54. Seals can be broken in the case of jams

There is a specific procdeure for doing so and it must be witnessed and documented.

Locally, I am trying to get my BoE to post a sign showing a properly filled out ballot using fake candidates.

[MH1]

30. Numbers are also checked against the number of voters.

It is very important that if someone is checked in to vote, they must at least cast a "no vote" before leaving. Otherwise at the end of the day the machine "total votes cast" will not agree with the number of voters; and like you said, fracas ensues. (well, it could. most likely you just have a set of tired cranky and nervous election board people at the end of election night, but when the overall tallies are posted, nobody else really notices.)

[MissWaverly]

32. oh, don't get me started on this one

the Board of Elections in Baltimore will not show the number of people that signed the poll book
in my precinct that's in the report I paid for. They have also never posted the results of the
2004 election on their website. They have 2002 but they have never posted 2004. Tell me about
how the machine votes cast are checked against the tallies of the poll book. NOT

[papau]

2. loss of under/over votes by hand counting is impossible - indeed one

determines under/over votes by hand counting.

The definition of over/under votes is that intent could not be determined until there was a hand count (or in terms of punch cards - a hand inspection).

The most common pick up is the write-in for the same name as is checked elsewhere on the ballot, the easy to read by hand vote that the machine could not read, and the slightly misspelled name that is easy to identify as a write in candidate. Also stray marks that are used to make machine read fail, put on the ballot by the GOP vote counter, can be noted and ignored in the hand recount.

[Kelvin Mace]

7. Ah, but court cases are built on trying to discern voter intent

I can program an OpScan unit (and some jurisdaictions require this) to reject an under/over vote at the time the vote is cast (so can TS systems but they need to go PERIOD).

Thus, if I accidentally make a stray mark on my ballot, I will know as soon as I cast my ballot that I have an over vote and I can correct it.

(I hope that no one is going to insult our intelligence by claiming that wingnuts will NOT argue like mad that their guy got the vote no matter if the intent is clear to my cat).

If ballots are handcounted they go straight from the voters hand to the ballot box. If the voter has failed to properly mark his ballot, then there is a good chance the vote will be lost.

I have watched people bicker over "voter intent" and while it worked pretty smooth in the distand past, partisan ranker has led to people claiming black is white, day is night, and that Gore vote looks like a Bush vote to me.

[papau]

85. True as to both bicker and as to OpScan - indeed the Florida switch off

of the automatic correction feature in minority and heavy Democratic areas - resulting in 10% to 12% vote rejection rates on final count compared to GOP white areas that had the OpScan (and like designed machines) auto check for error switched on and got around 2% vote rejection rates - was a point used in the 2000 Court battle.

The GOP courts of course ruled no one could prove the bad instructions to poll workers in Democratic areas was due to bias by the State and county and city poll workers.

But even the auto check does not recover for Opscan readers that are programed to look for the Dem vote mark in a place where the ballot design is not putting it - albeit it does warn if no vote is found - it does not correct for it. And a hand count would catch that.

[Febble]

3. At the risk of flames....

I can confirm that the exit poll discrepancy was significantly less in urban precincts that used optical scanners or DREs than in urban precincts that used levers or punchcards. One causal mechanism for this finding might be that differential residual vote rates on levers and punchcards was higher than on DREs or optical scanners.

However in precincts serving rural, suburban, or small urban communities (the only size of community that included HCPBs in the NEP sample) there was NO significant difference in exit poll discrepancy between paper and other technologies.

So FWIW, the exit poll evidence suggests that the counts from digital voting systems (overall) may have reflected the intentions of voters more acccurately than the count from punchcards or levers. It doesn't tell you anything about how they compared to HCPBs though.

[kster]

6. Is this a big problem in your country,

when you are counting the ballots by hand? Do you lose over and under votes?

[Kelvin Mace]

8. You are proposing that 100,000,000+ ballots will be counted by hand

so it WILL be a big problem. MIT/CalTech said the the aggregate residual vote (under/over) was 1.1%, that translates into 1.1 million votes. In some places the residual was OVER 2%. You get a quarter million people counting 100 million ballots and you are going to see fist fights, riots and law suits over voter "intent" on residual votes.

[papau]

88. fist fights, riots and law suits over voter "intent" on residual votes is

a good thing - it is sunshine on what is now hidden theft.

The Canadian overnight count of paper ballots in a national election can be done in the US - or are we just not as educated or trainable as our Canadian friends?

[Kelvin Mace]

129. Canadian ballots are standardized

and pretty simple. My county has sixty different ballots during elections just by itself, multiply that out by 3500+ counties nationwide and you have a problem. We also have MANY more races on a typical ballot than the Canadians.

BTW, how many elections have you worked in? Do they hand count ballots in your county?

[Febble]

10. Well, there's no such thing

We only have one race on the ballot. On the rare occasions when there is more than one race at the same election you get two different ballot papers (on different coloured paper so they are easy for the tellers to separate).

People do spoil ballots, either accidentally or deliberately, but there is a limit to how far you can go wrong when all you have to do is to mark a cross against the name of the candidate you want to vote for with a 2B pencil.

I'd love to see hand-counted paper ballots in America, but my mind reels at the thought of how you'd count all those races. And perhaps the reason we don't vote for so many people in the UK is BECAUSE it would be so hard to count them. That's a thought I've never had before. Perhaps if we adopt optical scanning here, we will become more democratic....

[kster]

12. UK.gov ditches 'Big Brother'-style e-voting

http://www.theregister.co.uk/2005/09/06/govt_voting/

LOL! You guys are way to smart to do that.

[Febble]

15. Yeah, but I'm sure it will come.

At the moment we can do it because we have the skills. I don't suppose it will stay that way.

However, I do think we start from a better baseline, because of our tradition of transparency, and also the way our democracy is structured. We count by constituency (around 30,000-50,000 voters), which means that the units are large enough to be worth a TV camera (there are TV cameras at all counts) but small enough to be done in a night by hand. Also the candidates are actually there, at the count. If we move to optical scanners, I expect the ceremonial aspect (and the TV cameras, and the public speeches and all) will be retained.

I am worried, though, I admit, though rather more worried by the possibility for fraud offered by postal-voting-on-demand. There is no voting method that is intrinsically uncorruptible, and methods that erode secrecy are potentially as bad as methods that erode transparency. We all just have to be eagle-eyed.

One huge incentive for fraud that we share with you is the first-past-the-post system. Fraud only needs to be targetted at swing seats to change the result. I have to say, I see absolutely no reason for your electoral college system. If you got rid of that, and simply elected your president by popular vote, at a stroke you'd make presidential elections much harder to steal.

[kster]

16. You mean like this

Still would be to hard for us American folk to count. We need that there electronic "thig-a-ma-jig" to do our counting.

[Kelvin Mace]

20. Please address the question

How do you vette 240,000+ people to count 100,000,000+ ballots.

[kster]

24. I'm not sure of all the details, all I know is

[Kelvin Mace]

27. Oh, that's bloody helpful

No plan, just a conviction.

[kster]

28. Lighten up dude, its the weekend........nt

[kster]

33. OK, even though its the weekend, heres a plan

look at this link and add in Febble's colored ballot idea and "whala".


http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=425015&mesg_id=425015

[Kelvin Mace]

55. OK, you have a transparent box with

Slots for straight Dem, staright GOP, independent and split ticket.


My ballot currently has 35 races, referendums, and bond issues. How do I split these ballots up according to your system? In my neck of the woods, voting is going to be all over the place between the party. What party slot does my school bond vote go into?

Also, you still have ignored the question of HOW do we vette all of the 240,000+ people that will be needed to count by hand the 100,000,000+ ballots nationwide?

[MissWaverly]

34. use bank tellers like in Mexico

[kster]

35. College credit $ for students and or there parents

who show up to count the ballots, talk about getting big bang for your buck.

[MissWaverly]

38. please, this sounds like a good idea but won't work

I have been involved in community projects, it takes a lot of effort to train, supervise and
organize volunteers who have no experience even if they are the best and the brightest that we
have. Bank Tellers do this for a living, they are skilled and our country relies on their
accuracy every day.

[kster]

40. OK, I'm with you, How about them tellers, prior to an

election give the students a course in counting the ballots so that they would be up to "snuff" if needed to help, in the counting.

Just trying to make sure "Future America" gets some $ for college and a chance to get involved in the process.

[MissWaverly]

45. well, I think there could be a place for students

but not in the vote count; we are up against a wealthy corrupt lobby that will use every trick
to maintain its power. Bank tellers have to answer ethics questions to get their certificate,
it is already on file and verifies their honesty. Though I do think that students as poll
watchers and runners to provide ballots when needed would be good. Also I think student escorts
of ballots to the tabulation area would be a very good idea, I am thinking half a dozen and
the ballot boxes should have GPS tracking labels just like the commercial where the boxes called
the help desk to let them know they were lost.

[Kelvin Mace]

57. Absolutely

but how do we deal with the person(s) who comes along and points out that banks are conservative, and have close ties to the GOP, so the GOP would then have an undue influence on elections?

(You know someone is going to, right?.

Also, isn't that "privatizing" the audit?

[MissWaverly]

67. well, how is having private vendors for DREs

not privatizing the election, and I think the honesty issue is something people are going to ask
and for a bank teller; they would have an ethics statement on file. You could say that they
use this procedure in the UK and in Mexico to help verify the elections and that a bank teller
would have his reputaion of honesty in the community to uphold. If people complained about
the vote count, it would reflect negatively on the banks.

[Kelvin Mace]

68. Oh, I agree with you, I am just preparing your for the arguments

I have no problem with CPA or bank tellers doing the counting.

[Kelvin Mace]

56. Have you ever actually WORKED in a precinct during an election?

Have you spoken to your election officials and explained this plan to them?

[Febble]

37. Yes, we use bank tellers too

but how do you count banknotes in American? Don't you use scanners?

[MissWaverly]

39. you mean the automated tellers (they're ATMs, here)

what you mean ATMs where you get the money out of the machines, yes, we use those, but the
banks still have branches open with tellers who count the money right there.

[Febble]

41. Well, that's good

no, I didn't mean the ATMs, I mean real people who actually count banknotes. They're the ones with the right skills! Counting valuable pieces of paper.

We have ATMs too, but we still have people who count banknotes. I fear it may be a dying art though, with so many transactions done by plastic. I think optical scanners are inevitable.

Incidentally, our Electoral Reform Society has been running elections for years for bodies like Trades Unions, and has an interesting website:

http://www.electoral-reform.org.uk/

But they use optical scanners!

http://www.erbs.co.uk/election_services/ballot_and_election_services.asp

Here's an ERS page on election integrity:

http://www.electoral-reform.org.uk/topstories/elecvoting.htm

[MissWaverly]

42. Thanks, our biggest weakness is our hodge podge elections

We have the elections process run by every state with no uniform standards, it's a nightmare,
a total nightmare, then you have our US Congress passing HAVA which forced all the states
to buy these electronic machines to be in compliance.

[Wilms]

44. And then you have all the mis-info further clouding things.

Lots of BoE's and, unfortunately, election reform activists got it wrong when they say things like, "our US Congress passing HAVA which forced all the states to buy these electronic machines to be in compliance".

Forced?

Show me the part in HAVA that says that you must use electronic machines.


I'll save you some time. It doesn't.

The vendors smile knowing that even their detractors promote sales.

[MissWaverly]

47. This article says HAVA requires electronic voting machines

The 2006 deadline has passed, and pressure is being placed on states to comply with the Help America Vote Act (HAVA). HAVA requires states to transition all voting machines to electronic and optical scan technologies in time for primary and mid-term elections of this year.
As states are rushing to spend millions of dollars for this transition, controversy over the accuracy and security of the new machines is creating obstacles to reaching this deadline.

http://www.truthout.org/docs_2006/033106L.shtml

[Wilms]

48. Don't mean to sound snarky but I don't care what the article says.

I prefer to reference the law.

http://www.fec.gov/hava/hava.htm


Show me.

[MissWaverly]

50. Wilms, I read the law

it says no manual or lever machines and that the states were to recommend machines to be
used through a certification process, well, Maryland did all that and came up with Diebold.
And now we find out that Diebold did not meet the certification guidelines in Maryland
up to and including the 2004 primaries. On paper it looks like we are following HAVA
but we are not.

[Wilms]

51. Read it carefully.

Levers seem to be banned (though some have/do/will argue that).

Punch card types are banned.

Electronics may be used, it's even kindof suggested that they be used.

But mandated? Nope.

[MissWaverly]

52. yes but it says the states are to recommend the machines

and what are they going to recommend with these machines and their powerful lobbyists

[Wilms]

66. Cite it. The wording is very important.

HAVA does not mandate machines.

[MissWaverly]

73. no it leaves it up to the states

SEC. 254. <<NOTE: 42 USC 15404.>> STATE PLAN. (a) In General.--The State plan shall contain a description of each of the following: (1) How the State will use the requirements payment to meet the requirements of title III, and, if applicable under section 251(a)(2), to carry out other activities to improve the administration of elections.

How the State will adopt performance goals and measures that will be used by the State to determine its success and the success of units of local government in the State in carrying out the plan, including timetables for meeting each of the elements of the plan, descriptions of the criteria the State will use to measure performance and the process used to develop such criteria, and a description of which official is to be held responsible for ensuring that each performance goal is met.

http://www.fec.gov/hava/law_ext.txt

[Wilms]

74. Where do you see electronic voting machines being mandated here?

Your not a salesperson for the vendors, are you?

Just kidding, but this is a serious, serious, misunderstanding.

[MissWaverly]

75. You don't know what we have gone through in Maryland

The Governor who is a (R) Ehrlich indicated that the Diebold machines should not be used
The State House which is Democratic voted to stop Diebold in 2006 in Maryland
The Maryland State Senate voted to continue with Diebold
and our Secretary of State Linda LaMone LOOOOVES Diebold


we are getting nowhere fast, it's does no good to point to all that high sounding rhetoric that
is HAVA, I do not see the Feds Testing and/or certifying these electronic machines, the state of maryland picked Diebold, and then we found out it was not even certified until the general election in 2004. We need help, what do we do, we have the machines, we have spent at least 23 million on them and we need to figure out what to do. I am nervous that I will once again not have the ability to make my vote count in 2006.

[Wilms]

76. I agree that MD's is an extremely frustrating situation.

That, however, does not re-interpret the law.

HAVA does not mandate electronic voting machines.

[MissWaverly]

77. No, it just says get rid of your machines if you want our money

well, then what do you buy, since you cannot have punch or lever then what is available,
the States are asked to pick and get back to them. I see several flaws in this, by not
going through the state houses, there was no real framework for oversight in place for the
original choice for the new system and now there is no real oversight of the new machines and the voters never had any say in this at all.

[Wilms]

78. Vote-PAD: The Simple Voting Device that May Save American Democracy

http://www.bradblog.com/?p=2330

[MissWaverly]

79. it's not going to happen, not enough moola for the vendors

and the whole thing driving Diebold is supposed to be their accessibility for the handicapped.

[Wilms]

80. I'll accept your view of that, as that's not what I came by to discuss.

Still, HAVA does not mandate electronic voting machines.

You'd do the reform effort well by remembering that, and repeating it, often.

[MissWaverly]

81. I understand your frustration but do you understand mine

I cried when Gore lost and I was so excited by this HAVA thing, Diebold was picked by the
SOS for Maryland, even Avi Rubin from Johns Hopkins here in Baltimore has made a steady
stream of protest that these machines are not reliable but the machines march on, it's like
the arrival of the gas driven auto. What do we do, are you saying there's a chance to
get rid of these machines now? The only chance I see is from RFK's lawsuit.

[Wilms]

84. I don't have an answer.

Of course I have high hopes for RFK's suit, but I really think it would help if BoE's knew that HAVA does not mandate electronic voting.

And if ER activists were aware of it, that would help, too.

[MissWaverly]

86. Keep us posted

If there's a chance we could donate to RFK's lawsuit money, I would donate, I donated to
the effort in Ohio both for the lawsuit and the recount.

[WillYourVoteBCounted]

111. Maryland, Early Voting and the Final Nail in Diebold's Coffin

Maryland is screwed for now.

But please don't lose hope.

This year, Maryland's Democrats have given you a great big gift,
your chance for a North Carolina -style election disaster!

Yes sirree, get your election protection folks in place,
and get ready!

Maryland will now have early voting for the first time, and
voters will be allowed to cast their ballots early from any location
in the state.

If you thought your Diebold machines had failures in other elections,
wait until they have to run all day long for several days in a row!

Here is the type of Diebold Debacle you can expect, with early voting

November 2004. North Carolina.
There was a problem with 11,945 votes from Dallas, NC.
Seven days after the election the County Board of Elections noticed
that 11,945 voter were missing
(the precinct rolls reported 12,867 where the GEMS system only showed 922).

Apparently, 15 early voting memory packs didn't work in the standard
(AccuVote-TS) process of accumulating the memory packs onto one unit
and then "downloading" them into the GEMS server.
That didn't work in a reproducible fashion.

These memory packs had to be directly loaded onto the GEMS server by
a DESI technician. This seems to have been done in a controlled manner. http://josephhall.org/nqb2/index.php/2005/10/03/p691#more691

Wait until you have capacity issues, and on election day - your counties
cannot get all of the votes to load.

You may have to wait a week to get the final "un-official" results, much
less the official results.

You can sit back and prepare for the meltdown, and this will be the
final screw in Diebold and paperless voting's coffin for Maryland.

So, try to get as many folks working in the polls, or poll watching,
and when we get the EIRS going, get fliers on that out all over the place.

You will be in the position of saying "I told you so".

Just get all of the documentation of this screwup that you can.

And, please don't give up... Lamone's day is coming.

Oh, and you may want to go on record warning Lamone's office in advance that
Maryland is not ready for the 2006 election, and suggest some election protection
measures (which she will surely ignore).

[diva77]

70. correction (I believe) -- if you accept HAVA moola, then you have to deep

six the punchcard machines. You can keep 'em if you deny the HAVA $$.

[Wilms]

71. I think you're correct.

That could explain the fact that punch cards are still around.

[Kelvin Mace]

58. Bill counters, which

a very accurate.

[MissWaverly]

60. that's true and it is also made by Diebold

But it gives me a receipt and when I punch the button for $20.00 does not give $100.00. I am
not accusing Diebold of skullduggery, it could be anyone who accesses these machines or the
central tabulators.

[Kelvin Mace]

62. Yes, but the rub is

that there is no need to hide your transaction from the bank. In this transaction, how you vote is a secret, which complicates matters.

And don't get me started on the idiot officials who prattle on about selling your vote if you had a receipt, something that could be done right now with absentee and postal ballots.

[Febble]

63. And it's a problem

over here, right now.

Secrecy matters. There's more than one way to rig an election

[MissWaverly]

64. well, then give me a randomly numbered ballot once you

verify my voter registration, but keep a tight track of this number sequence so there could
be no ghost votes.

[Wilms]

69. Thanks for that point about prattling.

Seems a good one. (And that's coming from someone who has, er, prattled on about coercing and selling votes.)

Still, how would a receipt help. The receipt could say one thing, the tally another. If an election is contested we mail them in to be counted? What?

[Kelvin Mace]

123. My comment about "prattlig"

was about election officials who claimed that having a "receipt" would mean you could show it someone later and sell your vote.

The problem with receipts as I have stated many times is that they are NOT ballots, so they don't make things safer. Second, the "vote selling" could be done with absentee or mail in ballots, so when are they going to outlaw them (I was maing a poin sarcastically to these folks).

[eomer]

97. Not only could be done -- has been done.

http://www.cnn.com/ALLPOLITICS/1998/03/04/miami.mayor/

Judge Orders New Miami Mayoral Election
Absentee-ballot fraud voids last November's balloting

MIAMI (AllPolitics, March 4 {, 1997}) -- A judge has overturned last fall's election of Xavier Suarez as Miami's mayor because of evidence of absentee ballot fraud.

-snip-

I'm interested in knowing why you think it is prattling to talk about ballot secrecy.

Thinking out loud, it seems like a voter-retained receipt could present various scenarios that would be troubling. For example, how about a battered wife whose husband could force her to vote the way he says? How about an organization that could coerce its members to vote a certain way? A cult, for example? Pressure from a spouse, an employer or some other entity could be more subtle than those examples and still affect a person's vote.

Do you discount that such scenarios would actually occur or that they would occur with enough frequency to make a difference? It seems to me that just one wife who is pressured into changing her vote is enough to nix the idea.

I'm also skeptical about what a voter-retained receipt would accomplish.

But I'm open to being persuaded if you have some ideas.

[Kelvin Mace]

125. Where did I say it was prattling to talk about ballot secrecy?

As I explained above, I commented that they have this problem with absentee ballots, so the concern is moot.

Second, I pointed out that receipts are useless for the reasons mentioned above.

I mention prattle because it is a silly argument in light of absentee and postal ballots. I have also made a point that it is far cheaper to get an insider to rig the voting machines invisibly than to buy votes. Each vote you buy is a person who can get you convicted of a felony. Buy a thousand votes and you get a thousand chances of getting caught.

Buy an assistant election diretor and you have one person to keep track of.

[eomer]

142. Thanks for the clarification.

I don't see the issues I raised as moot. They are real issues with real consequences.

Absentee ballots are necessary because they allow people to vote who wouldn't be able to otherwise. But they bring difficulties and those difficulties must be addressed as well as they can.

The use of voter-retained receipts would bring some of the same difficulties and should not be implemented unless there is a compelling case with substantial benefits to justify it. I don't see the the compelling case or the substantial benefits so far but am willing to listen if there are. If there aren't, then I don't agree with the position that the concerns are moot. Just because we're stuck with those problems in one segment of the voting does not mean we should open ourselves to them in the rest of the election.

I take your point about vote buying being a less efficient and more risky method of fraud but would say that this argument goes only so far. Vote buying, after all, does occur. You seem to be arguing that it wouldn't. But it does.

[Kelvin Mace]

145. I am certain it does occur

but it is not the problem it was in the past. No voter wants to go to jail for $20 and the likelyhood of getting caught is pretty high.

My opinion is that of all the threats to the voting process, vote-buying is pretty low on the list.

But, that is just my opinion.

[papau]

90. You vette via appointment of poll workers by the judge that runs the local

court system -

damn sight better system - and the one that was in place in many locals 50 years ago - than the current system of political appointees and cronies that are with one campaign or another who "volunteer" plus overtime for some election department workers who got their job via political favors in the first place.

Or use vetted professions like the use of bank tellers in Mexico

[Kelvin Mace]

128. Right

And we trust these judges because...

And using bank tellers who work for pro-GOP corporations will establish certain neutrality how?

[Febble]

22. Yes

But we never need very many colours!

We have an occasional referendum on some grand constitutional issue (membership of the EC for example), we have elections for local government, regional government (Scotland and Wales), European Parliament, and for Westminster (national government). But it is rare to have more than two on the same day (and more common to have one).

We don't vote for judges (it's supposed to be a strength of our system that our judiciary are not beholden to the electorate), and we certainly don't vote for all those downticket posts you guys do. And we wouldn't dream of having a plebiscite on gay marriage. Jeez.

Things is, Brits are lazy. Counting votes by hand might be a good excuse for not doing it very often, and leaving the decisions about how the country is run to someone else.

(I'm only half joking. I do have an enormous admiration for American energy and passion!)

[kster]

26. How do your exit polls come out in the end?

[Febble]

36. Well, they are a bit different

to yours, as they are just designed to give an estimate of which party is likely to have the most seats in the new government during the short period between close of poll and the first results. After that, as with yours, the estimates are based on the incoming vote counts as well. But they mostly poll swing seats - the exit polls are not supposed to be a random sample of the electorate, and the catch phrase of Peter Snow, the BBC guy that does the projections is "it's all a bit of fun".

In 1992, the pre-election polls had Labour slightly ahead, and the exit polls didn't give the Tories an absolute majority (my memory is that they gave Labour the most seats, but I might have remembered wrong). In the end the Tories had a majority of 23. There was a lot of fuss about it, as there was real optimism that finally the Thatcher years were at an end. But no-one doubted the count - everyone assumed the polls were off. Ever since, polls have factored in "Shy Tories" to try and give them a better estimate.

[kster]

115. You knew about the "Shy tories"

and you didn't tell our guy about it? I wonder why Mitofsky didn't factor in the rbr.

[papau]

89. The procedure is a simple do the National first and report, go back and

do the down ballot the next day.

Ballots in the 60's were not any easier than today - and a 3 day wait for down ballot results was not considered a disaster.

[Febble]

109. Well, that makes sense to me

as does abandoning the electoral college vote - there would be far less incentive for fraud, because you couldn't target it at critical states.

In other words, take the presidential vote out of the system completely, run it on hand counted paper ballots, and declare the winner of the overall vote the winning ticket. Have it on a different day, even.

I'm sure there are complicated constitutional reasons why you can't. I can dream, though. Oh for a world in which the leader of the most powerful nation was elected by sensible people!

[Awsi Dooger]

110. It can be argued there's more fraud potential minus the electoral college

All that middle American red, precinct after precinct. Right now no incentive to steal since those states are locked. But make every ballot of equal weight and suddenly snatching a handful of votes here and there can really add up. I'm not saying it would happen but I guarantee this forum and the liberal blogs in general would insist it happened, in the event of a close Democratic loss. Ohio? We'd have claims of Blackwell-type activity coast to coast, along with "I saw vote switching in my precinct!"

What amazes me is we can isolate the likely states yet still fail. Florida was considered the vital state in '00, then Ohio immediately identified for '04. I suggest we focus on preventing Virginia suppression and shenanigans right now.

[Febble]

113. hmmm.....

Well, I suppose you could argue that there's more that would be worth stealing. But I'm assuming that stealing is HARD WORK, and that the more you have to steal to win, the more likely you are to get caught. In effect, at present, a lot of the votes come ready-stolen (blue votes in red states).

I suppose it all depends on how hard votes are to steal, and I remain unconvinced that stealing millions of votes is easy (or even possible). If it were that easy, I think it would have happened in 2004, and, for reasons I've given elsewhere, I don't think it happened in 2004.

But I do agree about Virginia. Well, I agree about suppression all round.

[Awsi Dooger]

114. What's the common argument around here, only 4 votes per precinct?

Now project that nationally and you can tell what the claims would be. If we're debating a state with a 118,000 vote official gap, among 50 states I'm sure the cries of foul would be widespread and high decibel even if the margin were 2 or 3 million, already evident via '04.

Words like stealing and switching are thrown around here like checking your email. I agree it's hard work and probably impossible, although I've only voted in major cities like Miami and Las Vegas. Whether tiny rural precincts are monitored and secure, I don't have a clue. In fact, I can't envision how they are set up or operate.

I loved that ready stolen description. Never thought of it that way before.

A national vote would certainly require Democrats, actually both sides, to compete in 50 states. Who would gain more votes from that I'm not sure, but it would confuse the national parties initially and dramatically alter the dynamic. You wouldn't have saturated commercial coverage in the swing states alone, nor dozens of trips by the nominees to only those states. I guess it would mandate relentless emphasis on California and New York and the other high populus states, with much more diversity and scrutiny of campaign strategy.

Overall, I think we're in a slightly negative situation right now in terms of how frequently we would win the electoral college in a 50/50 popular vote with normal distribution, but if states like Virginia, Colorado, Nevada and New Mexico continue to drift our way as predicted, that changes in a hurry. BTW, I completely agree Kerry was probably the rightful winner of New Mexico, the Florida 2004 in terms of will of the people not reflected by the state's electors. Paper ballots in New Mexico this year.

[papau]

87. Hi Febble - levers are infamous for high error rate due to wear on the

mechanical parts of the machine - but the distribution around exit poll data "truth" I thought was classic bell curve.

Did the lever data show more bias than DRE?

Since I come from a background of watching/finding out about Dem votes tossed in the river with bogus GOP votes stuffed and counted - all with paper in 1960 in Illinois - I am not more concerned about DRE/digital voting systems being "more" corruptible - indeed I see them as just being easier to corrupt - less effort required and fewer fingerprints left - if the urge hits the GOP folks at that poll.

If the urge hits the GOP folks, there will be corruption with or without DRE/digital voting systems.

[papau]

93. Just realized that given the rather small number of samples my

question on biasing of lever machine error would be impossible to answer.

But I am still curious about how you would "fix" those machines to favor a given candidate - the random changes as to name placement by precinct required by many jurisdictions make deliberate wearing out of a lever not effective if it is a bias in favor of a given person that you want.

[Febble]

96. Well, that's true when it comes to

paper ballot precincts, as there were only 40 of them in the sample, and 33 of those were in rural precincts (5 in suburban precincts, 2 in small towns, and none in urban places with more than 50,000 population). That was why I had to consider size of place when comparing them to precincts using the other kinds of technology.

But I also had to pool across states. It is possible that the bias associated with levers was not due to anything to do with the levers, but to something else at state level that lever precincts happened to have in common (like being in NY, for example) and that it is the "something elses" that are responsible for the discrepancies. An alternative is that Kerry voters were more likely than Bush voters to register presidential undervotes on lever machines. I don't know enough about the technology to be able to guess how or why this might happen.

The trouble is that this all goes back to the issue of whether the exit polls can be considered any kind of audit - and the answer must be "only with difficulty" - because it wasn't a random sample (which underscores the need for any audit protocol to specify very clearly how the random selection is made). What is certainly clear from the exit poll data is that the discrepancy was not greater in DRE precincts than in precincts using other voting technology, which is one of the findings that leads me to consider it very unlikely that the election was stolen on DREs (voter suppression and lies being my top two candidates). And it is certain that without any one of: voter suppression, punchcard technology, badly designed ballots, in 2000, Gore would have been the incumbent in 2004, and probably would have won.

But I should also add that that leaves aside the issue of New Mexico, where it seems as though Hispanic votes in particular were lost on pushbutton DREs (not touchscreens). My hunch is that Kerry won NM. But I didn't get that from the exit poll data!

(edited for clarity - probably still not clear)

[Febble]

94. Yes

lever precincts had a larger mean exit poll discrepancy than precincts using any other voting method. To do the comparison properly, though, I separated rural, suburban and smaller urban (pop<50,000) from larger urban precincts, as paper ballots were only used in smaller places (mostly rural) in the NEP sample. When I looked at the discrepancies in these smaller places there was no significant difference between precincts using the different technologies. However, for urban places, the discrepancy was greatest in lever precincts, followed by punchcard precincts, followed by optical scanners, then DREs. Lever precincts had significantly greater discrepancies than either DREs or optical scanners.

I agree with your last sentence. Corruption is as old as democracy, and it is perfectly easy to corrupt an election conducted with hand-counted paper ballots (an attempt was made in the UK in 2004). No voting method is immune from corruption, though some counting methods may be more accurate than others, and some are more reliable than others.

IMO, the two key issues are chain-of-custody and random audit, whatever method is used, although to be able to do the second you have to have some kind of ballot that can be independently counted.

And none of that will deal with voter suppression, or indeed, with candidates that tell lies.

[Wilms]

4. Then let each voter use an OpScan to check--not cast--the ballot.

Awkward, but doable.

[Kelvin Mace]

9. But we can't TRUST the machines, remember?

Diebold will deliberately program these machines in certain precincts to reject certain ballots causing disenfranchisement.

Also, just because a machine accepted the ballot doesn't mean the vote will be counted later. The vote is NOT tallied until the hand count, and the ballot WILL be contested by one side or the other if it is anything but absolutely pristine (and given the wingers, even then).

With OpScan, your vote is talled when cast, unless rejected for under/over vote. If rejected, you may correct it, then submit it again and get it counted.

With a hand count, your vote is not tallied until you are no longer present to explain your intent and is thus subject to interpretation by humans with political agendas.

Not being snarky, just putting forth the arguments I keep hearing about not trusting machines. :)

[Febble]

11. Well, I think there are at least three

actions done to a ballot that are worth separating:

1. marking it
   
2. casting it
   
3. counting it

With paperless DREs they are all the same action.

With hand-counted paper ballots all three are entirely separate.

With optical scanners, from what you say, casting part of the process marking it - in that if you try to cast an under- or over-vote, you have the opportunity to re-mark. This seems like an advantage.

[Kelvin Mace]

13. That's the way I see it.

[Wilms]

14. Perhaps I wasn't clear.

I am suggesting the possibility--not even advocating--of providing a PBOS that a voter can use, if desired, to aid them in checking their handmarked ballot for under/overs, not for actually casting/tallying the ballot. The machine would be used only for checking.

Didn't take your comment as snarky.

Regardless of methodology, my guess is random auditing is an important component.

[Febble]

18. I think random audits are KEY

which is why I think HR550 is so important (you can't do an independent audit if you don't have an independent record).

And a secure chain of custody is critical too.

Our UK elections are good, but there are plenty of ways in which they could be (and are) corrupted, without a machine in sight. Stealing postal ballots is one.

[liam_laddie]

43. Objections to Holt's HR 550...

My comment may belong in another thread, but Holt's bill sets the
audit percentage at, what, 2%? This is less than half a minimum
need and 10% (5x Holt) is more like a sensible check. Would it
really be random? Many Ohio counties in 2004 certainly violated the
"random" aspect; all under the direction of the notorious Blackwell...

I believe this low audit percentage is the main objection to 550's
passage "as is"; it casts into law an insufficient fraud check and thus
it might be hard to amend the percentage later, particularly anticipating
rethug arguments..."you libruls ain't never satisfied, are ya?"

Our version of postal ballots is the mail-in absentee; do I know if my
ballot was counted, or even received? Well after the election, the BoE
posts a registration list indicating whether I cast a ballot (in OH, no
party affiliation is recorded, or actual vote choices, of course.)

[Febble]

46. On what basis

do you consider 2% inadequate?

The percentage is irrelevant to its adequacy anyway. What matters is sample size. A random sample (which Holt mandates) of precincts will give you a lot of statistical power.

How much do you want? What are your criteria?

[liam_laddie]

49. Not criteria...(I'm not a scientist-stat person)

but my mistrust that the random requirement would be met. For that matter,
a 10% requirement could be rigged too. Just the experience here in OH;
who can we trust to enforce the 2% random? The elections process here
is now simply too politicized to fully trust every county BoE. Recent rules
passed in Ohio (HB-3, etc.) are blatant attempts to suppress the vote.
Let's see how Mexico comes through its crisis.
BTW, I read that the UK is uncomfortably hot these days...hope you have
some cold ones handy...
:toast:

[Febble]

53. We just had a big thunderstorm

and it's cooled down a bit (one of my cats just came in drenched).

But yes, we've got a good supply of beer in the fridge. :toast: (yes, we do drink cold beer.)

I agree about the random bit. Statistical probabilities always assume random samples, and it is quite incredibly difficult to get a random sample (and that of course goes for exit polls).

I'm more concerned with how the random sample is specified than with how large the sample is, actually. If you know your sample is random, you can be very precise about your confidence limits. If you don't know your sample is random, your confidence limits are meaningless, as in the case of the Ohio "recount".

[Kelvin Mace]

61. You need to trust mathematics, and distrust your instincts

10% is excessive.

[Kelvin Mace]

59. Who says 10% is needed?

Also, there is NOTHING in Holt stopping a state from requiring a higher percentage.

[liam_laddie]

98. My concern is...

that BoE's would never exceed the minimum law-set percentage. A higher sample
would mean more expense piled on top of the ongoing operating costs of the electronic
systems, which include on-going contracts with the systems vendors - maintenance,
software updates, "fixes" (wink, wink.) It is highly doubtful that 2% would ever be exceeded,
even though the law would not prohibit a higher sample. Some BoE's act as if budget
decisions are more important than delivering a fully trustworthy and accurate tally.

As I understand it, there are other parts of HR 550 which are troublesome *because*
it sets into law some requirements about which there is now more informed discussion
and why more input is needed. I see no other reason for reich-wing support of 550 except
that they want it on the books before more revisions are made which would harm their
control of election results...limit the damage from 550, so to speak.

I will try to find a list of objectionable aspects in the current 550 and post a link.

[Febble]

99. As I understand it, HR550

requires a minimum of 2% of precincts to be audited, with a minimum of 1 from each county, and each precinct to have an equal chance of being selected.

Taken literally, this would, in general, result in more than 2% of precincts being selected, but I am unclear about the current version of the wording. To comply literally, at least 1/N precincts would have to be selected from each county, where N is the number of precincts in the county with the fewest number of precincts, and you'd often end up with more than 2%.

But don't think this is what is intended, at least from the wording I've seen, which implies that precincts have to have an equal chance of being selected except that at least one has to be selected from each county. In which case, one precinct from every county would have to be selected, and if this was less than 2% of the state, the shortfall would have to be randomly selected from the remainder.

I'm not sure that the "at least one per county" is such a good idea, as one per county is not enough to audit a county and will tend to lower the probability of finding miscounts concentrated in one part of the state (because it puts a lower ceiling on number of precincts from any one county likely to be selected). I also think it would be much better to mandate a minimum sample size rather than a percentage.

But the good news is that if there are around 10,000 precincts in a state, that will give a sample of about 200 precincts, which should give you a high probability of detecting fraud in as small a proportion of 4% of precincts, although for smaller states, the statistical power would be less. And if you seriously want to influence an election without being caught, you probably need to do it in more than 4% of precincts (if you'd tried to switch, 50,000 votes in Ohio in 2004, by corrupting only 4% of the precincts, each corrupt precinct would have had to have flipped over 100 Kerry votes, which would have stuck out like a sore thumb anyway.

So the 2% doesn't bother me as such. My two cents is that I don't see the point of the 1 per count restriction (which seems to weaken the audit, not strengthen it) and that I think the legislation should mandate a minimum sample size, not a percentage.

And at state level, there should be within-county audit protocols as well. I'd like to see some counties selected either at random, or on the basis of deviation from past results, for a much more intensive audit (or complete recount).

[BeFree]

100. The only solution

Audit every machine. All 10,000+ machines should have audit protocols in place before the sucker is ever plugged in.

Any machine should be designed to be easily audited. And the machine needs to be the same type make and model across the nation, with the same source code and have ballot definition files that can be written in the offices of each jurisdiction.

If we are going to use machines, every local election office must be able to do any software work needed, or they are to be considered as incompetent to handle the machines.

[Kelvin Mace]

104. This bears repeating

2% IS THE MINMUM, NOT THE MAXIMUM!!!!!!

[Kelvin Mace]

105. The ony objection I have seen 550

is the 2% minimum.

Because of this, some folks are prepared to throw out the entire law.

HR 550 has many of the same provisions as our law in NC. In some areas 550 is tougher than our law. We also sample 2%, yet Diebold left the state rather than comply with our law.

Our law passed both houses of the legislature unanimously.

[Kelvin Mace]

19. I agree that your suggestion is rational

and addresses the point I raise. However the bar has been set by the hand-count crowd: No machines.

[papau]

91. I agree - at least 5% of precincts and 5% of votes - and truly random n/t

n/t

[GeorgeGist]

17. Are you serious?

setting up a system whereby paper ballots cannotbe over/under voted is not a difficult engineering challenge.

[Kelvin Mace]

21. Then tell me how

[Jane Austin]

31. Not so long ago, in my Austin precinct

the votes were counted all day long.

Many precincts never get more than 500 votes on a given day. That's fewer than 50 votes an hour to count. Even if it were 200 votes an hour, that's just not that many.

When the polls close, post that vote count on a bulletin board right there. Then you send your ballot box to the central tabulator, and if the count isn't pretty close to what you counted - - in public - - at the precinct level, you raise Cain.

One thing I'd like to see is some citizen impediments to these certification votes in state legislatures. Why do they ignore all the complaints and certify anyway?

[Kelvin Mace]

65. Because they don't believe in questioning their own

elections.

500 votes may be the norm in Austin, but it varies wildly nationwide. I believe the average we have been using is about 1200 ballots per precinct.

You are assuming a ten hour day after the twelve hour voting day? The ballots have to be counted once for EACH race on the ballot, by EACH party rep. An election with 25 races in my state (which requires 4 people to count) mean each ballot will be counted 100 times by hand in order to tally the races.

That's a Hell of a lot of handling of a ballot.

[papau]

92. Over 4 or 5 days, counting 25 races is not a big deal - and National is

done first.

Is speed more important than confidence in the result?

[Kelvin Mace]

101. It is not just speed

though the longer a race goes on, the greater the chance of chicanery, it is cost. Man power intesive systems will cost a LOT of money and the overall accuracy will go down, while contention over elections goes up.

[papau]

107. what backs up those assertions? The longer counting goes on does

not seem to a positive correlation to todays GOP cheating, nor do short vote counting times correlate with fewer dollars spent on finalizing the election returns, and the overall accuracy goes down statement needs a plausible mechanism that depends on time before the idea even gets to being a theory.

Sounds a bit like a power point slide presentation by a salesmen for a voting machine seller - not that you are such - but it does sound like the presentations I have heard over the years.

[Jane Austin]

116. And well worth it, to preserve democracy.

n/t

[Kelvin Mace]

118. As the more the ballot is handled

the greater the chance for error, lost ballots and cheating, nationwide hand counting will NOT preserve democracy.

[Jane Austin]

141. I guess we disagree about that. n/t

[Kelvin Mace]

146. Sorry we disgaree

but it is established fact.

The more steps and complexities you add to a system, the more likely the system is to fail. This has been a major principle in engineering since, oh, I would guess the builders of the pyramids.

Each time a ballot is handled for counting is a step in the process. The more steps in the process, the more cahnces for something to go wrong.

This is why jet engines are far more reliable than car engines (by an order of magnitude or better). Jet engines have fewer parts.

[diva77]

72. Please explain this to me: if you're in favor of audits that make use of

hand counting paper ballots (or perhaps you in particular are not; but others are) yet you are not in favor of 'total hcpb', then please tell me:

1) how do the security requirements for 'total hcpb' vs. 'hcpb for audits only' differ?

2) where do you draw the numerical line between believing that 'total hcpb' is an impossibility vs. 'hcpb for audits only' is ok ? How many ballots are too many to count by hand?

[BeFree]

82. Here you go: a compromise

Audit the paper ballots that will be op-scanned.

2% or 10%..... whatever a commission of statisticians come up with.

That way just a few (2% too 10%) of ballots need to be hand counted.

Then run all the the ballots through a scanner and compare the totals.

If there is agreement, fine.

If not, hand count all the ballots.

[diva77]

83. thanks for response, but I'd really like someone to answer my questions

specifically. BTW, in many counties, the RoV's staff is given the machine result BEFORE the hand-counted audit is done -- then they keep adjusting the handcount 'til it comes out "correct" - yeh, machines make things easier

[Kelvin Mace]

102. Random sampling

for audits is done by trained people in a few precincts which makes it easier to control the conditions and keep the public eye focused. When you handcount everywhere, you lose control of the conditions, the quality of the people doing the audit WILL fall, and the focus of the public diffused.

My father-in-law runs a machine shop. The quality of his parts is critical. A recent orders he shipped invloved seven million parts. Now, which makes more sense:

1) Selecting a statistically significant number of parts at random and having his highly trained QC employees (that is their only job) EXTENSIVELY test the parts for adherence to standards, or

2) Hiring hundreds of people, giving them crash courses in QC, and having them examine all seven million parts?

It is easier to do the former, and mathematics will yield a result that is 99% accurate, than to do the latter which will actually lower quality, take longer, and cost more.

For the record, the company has been in business over 30 years, and has a QC rating of 99.96%

[BeFree]

121. Dumb

What if, as is the case with votes, he had many machine shops? Would he audit just one line? No, he would audit each and every line in each and every shop.

Too, what happens when the audits find bad parts? They shut the line down and fix the problem. Not so with cast votes.

Your example can be further shown as to how dumb it is by thinking that if the mechanics examine just a few of the parts making machinery, one can surmise that all the machines are working properly. That is not the case in any shop, and it shouldn't be the case with our voting machinery.

Every line and machine in every shop needs to be examined.
Every vote machine in every precinct needs to be audited.

[Kelvin Mace]

122. Actually, since you haven't figured it out

you don't produce seven million parts on a single machine. This was done by a dozen machines on three separate plants. So it works even then.

Parts are randomly sampled DURING manufaturing, not after, so that corrections are done in real time.

Sheesh!

[BeFree]

124. Apples and oranges is dumb

Votes are not sampled during manufacturing so corrections cannot be done in real time.

Look, you are trying to sell a pig in a poke. Nobody likes the voting machines and I doubt you have changed anyone's mind, except for a newbie, maybe.

Your personal attacks on me show that you have to stoop to prove what points you are trying to make.... why do you go so low if you think your case is so high? It really is not very pretty.

[Kelvin Mace]

126. I anxiously await to see the law that you

and your supporters come up with.

Can't wait to see who you get to introduce it to congress.

[BeFree]

127. You

Well get you to push it, because you know it is the right thing to do, eh? Then you can be famous and tell everyone you saved the world!

[Kelvin Mace]

132. I never claimed I save the world

that would be Bev Harris.

I did at least work on a bill that actually passed. And you have done... what, precisely.

[BeFree]

135. I've worked a number of bills

But I don't brag about that. That's in the past. It is all about the future for me.

But I have seen so many big egos come and go, it bothers me little that they tend to take credit for things that many hidden hands helped make come about.

[Kelvin Mace]

144. What bills, involving what legislation?

How many passed? What was your contribution?

You have already admitted no expertise on the computer side, how much expertise do you have on the democracy side?

[BeFree]

151. Odd

I find this fixation on me to be quite odd.

Frankly I find it as an affirmation that I am barking up the right tree.

It matters not what I have done, it matters what I will do.

What I will do is continue to ask questions of the authorities, continue to seek the best of voting systems and seek the best from the laws and the people that have influence upon those systems and laws.

I like my anonymity, and know that any attempt to boost my ego by posting personal information here will put me in the cross hairs of the ChoicePoints of the world. I detest the ChoicePoints and will do nothing to help such corporations to endanger your's or my freedom.

Asking me what I have done is like asking what has DU done.

So what has DU done?

[Kelvin Mace]

152. Uh, I'm am, not "fixated" on you

You are asking ME lots of question and turnabout is fair play.

So, based on this answer, you have done nothing but bloviate on the issue which you have little experice in and you will continue to do so in the future.

OK, just so I understand and will stop wasting my time.

[yourout]

95. Use some form of scanner to help with Under/Over votes and count by hand.

See how easy that is.

[Kelvin Mace]

103. Ah, no

The purists have stated machines can't be trusted. You can't have machines involved.

Also, even if you do, the fact that a scanner said your ballot was OK, is no guarantee that the handcount will say it is OK. The scanner may register an oval with an x in it as a vote, but the guy couting the votes says it doesn't count becuase the oval must be "filled in" not "x'ed".

Rules is rules, he will say, espeically if it deprives his guy of a vote.

[timewellspent]

106. You are right Kelvin

I have witnessed a hand count in Ohio and that very thing happened. A few voter's seem to have a problem with the concept of filling in the oval. Let's say they fill it in and change their mind. They may X out the first choice and mark the new vote by filling in another oval. The Precinct optical scan will catch that and tell them it is an overvote.

People seem to think that hand counting is the answer. Let me ask you this question. If the ballot is counted at the precinct level, while the voter is there, how would it be better to place them in a ballot box to be counted later? If that happens, how does the voter know his vote is counted?

The undervote was programmed to alert the voter in Ohio during the May Primary (in optical scan counties). It helped the voter because sometimes there is an issue on the back of the ballot that they might have missed. The machine helps the voter to determine if they have correctly filled out the ballot. I think that is the basic intent of HAVA. To alert the voter that they might have made a mistake. That is why punchcards got outed.

Just some food for thought.

[Kelvin Mace]

119. I keep tryingt to point this out

but at times I feel I am talking to a wall. Since what I say contradicts the conventional truthiness of the HCPB crowd, they are NOT going to accept the fact that a properly secured and monitored OpScan system will render a fairer and more accurate count than a manual system.

[BeFree]

130. Where is this fantasy machine?

"....a properly secured and monitored OpScan system..."

Who makes such a thing?

And what expert will stake their reputation on it?

[Kelvin Mace]

131. Asked and answered

[BeFree]

133. No, you haven't.

[Kelvin Mace]

134. Asked and answered

I can't help your faulty memory.

[BeFree]

136. Where? Is that so hard to provide a link?

Where is this fantasy machine you keep talking about?

[Kelvin Mace]

143. This is a problem I run into quite often

with Bev supporters. They never want to make use of the search function and expect other people to wait on them hand and foot.

Asked and answered.

[diva77]

108. No 2 scanners are calibrated exactly alike, so your argument falls apart

here. Also,scanners are capable of creating votes that were never cast. This has been documented from studies of the 2005 ballots in CA.

You adher to belief that people are incapable of counting, but then you say that 'experts' from the RoVs offices can be brought in. By having it done by pollworkers at polls (or pollcounters, or whatever they are called) this would serve as a further check and balance on results, rather than RoV employees that may not be allowed to question what is going on for fear that they may lose their jobs.

But still, my questions remain unanswered:

1) how do the security requirements for 'total hcpb' vs. 'hcpb for audits only' differ?

2) where do you draw the numerical line between believing that 'total hcpb' is an impossibility vs. 'hcpb for audits only' is ok ? How many ballots are too many to count by hand?


We used to have hcpb; why can't we have 'em again and just beef up the security and chain of custody requirements, and enforcement thereof ?

[Kelvin Mace]

117. I answered the questions

Total HCPB involves MASSIVE mobilizations of resources and people. I gave you an example as to why a limited count usings a mathematically valid sample and trained people is better than counting everything with mediocre or poory trained people.

While no two scanners calibrate identically, they can be calibrated to withing 99%+ of each other.

To determine a "line" you would have to do a study measuring the deterioration of accuracy as the number of ballots and counters increases. I am sure a statistician can provide some modeling data, I know MIT did some studies on hand-counting and will have to see if the study is on the web.

We used to have hcpb; why can't we have 'em again and just beef up the security and chain of custody requirements, and enforcement thereof ?

We used to have a smaller country, with simpler ballots and fewer voters. As populations grew, it became necessary to automate the counting process. Automation was also done to cut down on cheating.

As to adding more security and more enforcement, that translates to more people and we are already up to about a quarter of million to conduct a nationwide handcount of 100,000.000+ ballots.

[diva77]

140. I have heard that hcpb costs, I believe, 33% of what machines cost. Where

did you get info. citing that "it became necessary to automate the counting process" ? As it is now, we have machines in every precinct that we have no idea what type of "easter eggs" they may be harboring, and tabulators that may be harboring those eggs as well. If we're going to be realistic about security, training, maintenance, etc etc then we need MASSIVE mobilizations of resources and people for securing machine-based elections and still, with that, we can never be sure whether our votes are being counted as cast. We have lots of people unfamiliar with computer technology handling the machines. You might consider them to be "mediocre and poorly trained." Most people can understand and handle and count a paper ballot.

I am not against machines simply for the sake of being against machines. I am against citizens having to bear the burden of proof that the elections are transparent and secure. The burden of proof is BACKWARDS. So far I have no proof that elections with the machines are/can be secure, transparent and accurate and that in the long term we can bear the costs of replacing/maintaining/trusting these machines.

[WillYourVoteBCounted]

148. first paperless voting systems have to be banned

Half of the country has been voting on paperless machines
for 40-60 years, and
half of the country has been voting on hand counted paper ballots
until about 1980.

So voters have to be taught to expect a paper ballot of any sort.

When canvassing a large NC county with paperless voting, many of the
people answering the door expressed high confidence in the
paperless touchscreens. Really. This county had used
paperless DREs for 20 years!

If we can get paper back into the equation,
then perhaps then the next step can be -
gee, why are we making this so complicated,can we do it better by hand?
Especially after having mandatory hand to eye audits and recounts of
elections by the paper record.

Of course, much has to change -

More elections, less contests on the ballots,
allow poll workers to work in shifts,
tighter security and chain of custody

The issues that raised suspicion with HCPB would have to be dealt with.

The issues that affect logicistics and efficacy would have to be dealt with.

[Kelvin Mace]

149. Your local library has dozens of books

on the evolution of voting and why machines were brought into play. The purpose of automation was to speed up a process that was slowing down with population growth and a desire to remove human bias from the process. Like many things, it started out with good intetions...

As to "easter eggs" hiding in computers, how do you hide such things in computers when:

1) You are required by law to disclose the code to election officials.

2) Your CEO must sign a sworn affidavit that the code you submitted is the code running on the machine.

3) Using code other than the one provided on the machines will result in your CEO being charged with perjury, and the persons responsible for the rogue facing felony charges.

4) The software's accuracy and reliability is checked after then election by random sampling and hand-eye counts. If the paper count deviates from the digital count, the paper count is used. Deviation above a statistcally expected norm will result in wider audits of more ballots.

These are the protecions my state enjoys under it's S.223 law in application and practice. HR-550 offers many similar protections.

[Wilms]

150. An increase of population increases the pool of people to count ballots.

Not to discount the management of such enterprises.

Also, does ES&S operate in your state? If so, why can they operate uncertified software each and every time their equipment is deployed?

ES&S Programming Is Unverifiable
http://www.votetrustusa.org/index.php?option=com_content&task=view&id=1475&Itemid=51

[peacetheonlyway]

112. simple...

make the UNDERVOTE a candidate and explicitly highlighted in each race...

for example:

*** MARK ONE AND ONLY ONCE CHOICE FOR EACH RACE
PRESIDENT
____Bush ____Kerry ____No Vote WriteIn________________

GOVERNOR
____Perdue ____Taylor ____No Vote Write-In_______________


we still can't help stupid people marking 2 times on the same race but a left to right (not up to down like most computer screen or paper ballots in past) will eliminate the confusion.
This makes the undervote explicit. believe it or not many elderly folks hated Kerry and Bush (surprise, surprise) so the undervote is not always a result of bad ballots or bad technology but often the intention of the voter and this allows it to be EXPLICIT.

The other way to solve stuffing the ballot box is a $1.50 turndial lock. 2 of them.
Basically one lock the combination of which is known by the election director, the other combination is held by a State appointed citizen election activist group. There are over 200 election fraud groups and sometimes 3 or 4 per state, so this should be easy...

this will make the UNDERVOTE explicit and if there is nothing marked on the line it gets listed as an NO Vote.

folks many of these problems are easily solved with paper ballots...

[Kelvin Mace]

120. How does this help in the case where

the voter misses the race for some reason.

Your suggestion simply helps us spot intentional undervotes, not accidental undervotes.

[peacetheonlyway]

138. the design itself

removes the 'unintentional undervote' as people who explicitly did not want bush or kerry were lumped in the same boat as folks who accidentally did not prick the ballot hard enough.

going back to a simple pencil paper and fill in the bubble ballot will likely remove the bulk of undervotes because it will be clear how the person marked their ballot. if you see no mark at all on the line you assume the user did not want to vote in the race but likely you will get folks filling in the NOVOTE giving you explicit understanding of waht is a chosen undervote versus what is a badly marked ballot... FYI a true democracy should allow for dumbass people who don't mark their ballot correctly.. at that point the GOAL becomes readable, easily counted ballots designed in a much better way than current ballots...

undervotes are almost entirely a function of the machines that had you prick your ballot with a little stick or lever machines that had evil republicans nailfile down the levers so that no democrats were ever selected (we have proof of that in a little place called licking county Ohio)

[Kelvin Mace]

147. And how do we determine the intent of

"dumbass" people who don't mark their ballot correctly?

Also, since the voter has already left the precinct hours ago, and can no longer defend his ballot, how do we protect against "stray marks" that somehow get on the ballot wgioe the vote is being counted? (If you don't think that I can secretly spoil a ballot in plain sight, you are seriously kidding yourself).

With OpScan, the vote recorded (or rejected for correction) WHILE THE VOTER IS STANDING THERE.

With your system, the vote is recorded hours, days, perhaps weeks late.

BTW, how many elections have you worked? In what capacity? What method do they use in your county?

[Awsi Dooger]

137. You don't want to do that

It gives far too much weight to "no vote". If voters are suddenly provided that as an eyesight option, the percentage of voters choosing it could soar, probably to our detriment.

We have that option by law here in Nevada, for statewide and presidential races. Check out this link and scroll down to Nevada. In the 2002 governors race, "None of these" came in third with 5% and almost 24,000 votes, defeating four names on the ballot: http://www.cnn.com/ELECTION/2002/pages/governor/index.html

In 2004 it wasn't as bad as I expected, but "None of these" got about 3700 votes in the presidential race.

I believe in GOP suppression but not vote switching. However, for those worried about fraud, I don't know why we want to provide another place to stash votes, ones that may have been intended for our side.

[WillYourVoteBCounted]

139. Election Integrity and Cigarettes

While I agree with much of the points made here in this thread, and
also realize that there are no perfect answers, I wanted to add my 2 cents.

First of all, it took us a long time to get into this mess.

We won't escape it quickly

In 1980, according to the Cal Tech MIT 2001 study.

40% of the country was using hand counted paper ballots,
36.4% of the country was using lever machines.

This all started to change as people sought ways to deal with
election problems (fraud, errors, labor, etc).

It is easy to see how we ended up with an explosion of DRES.
How hard would it be to trust paperless voting when you have
had paperless voting anyway?

So, we have a long walk to go to getting people used to paper again.

We are over half way there (27 states )



Just getting paper ballots in combination with computerized voting
has been a major revolution.

It may be as difficult as taking on the tobacco industry.

Tobacco Industry's Economic and Political Influence

AHA Advocacy Position
To help reduce the economic and financial influence of the tobacco industry, the American Heart Association supports efforts to level the playing field for health advocates through political action committee (PAC) and campaign finance reform, gift and honoraria limits, so that members of Congress can make decisions on health matters based on a review of the facts.

Background

Historically, tobacco companies have used their economic power to wield considerable influence on the political process. http://www.americanheart.org/presenter.jhtml?identifier=11224

We are asking people to change their way of thinking, and
that is hard to do.

Cal/Tech MIT tried to measure the accuracy or quality of voting systems

Examining undervotes for president in years 1988 to 2000
1.6 percent had over vote or no vote with optical scanners
1.8 percent had over vote or no vote using paper ballots or lever machines
2.0+ percent had over vote or no vote with punch cards or electronics (DREshttp://www.cs.duke.edu/~justin/voting/docs/CalTech_MIT_Report_Version2.pdf

Take a look at how each technology performed in North Carolina in 2004:

Race_______________________DRE__Lever_OpScan_Paper_Punch_Card_Other
PRESIDENT_________________2.05__1.27__2.19___0.67__1.94__2.09
US SENATE_________________3.49__4.69__2.50___4.75__2.41__3.42
GOVERNOR__________________3.03__4.93__2.00___2.45__1.96__2.91
LIEUTENANT GOVERNOR_______5.70__9.90__4.26___9.15__4.25__5.14
ATTORNEY GENERAL__________6.61__12.94_5.43___8.53__5.90__6.75
COMMISSIONER OF INSURANCE_6.50__12.08_5.09___7.37__6.20__6.07
SECRETARY OF STATE________7.48__10.51_6.09___9.95__7.12__7.49

http://www.cs.duke.edu/~justin/voting/totals.html

Hand counted paper ballots had the lowest undervote rate for President.
However, the drop off rate(higher undervotes) got worse for lower contests.
While many of us don't care about that - guess what - the folks who decide
how we will vote - many of them are affected by the drop off rate.
Whether we agree with this or not, it is a political reality.

There were only 3 counties in North Carolina that did hand counting,
and the counties only had about 5,000 - 7,000 registered voters living there.
I called the directors of these counties, to ask them how it went in 2004.
They told me that they had confidence in the HCPB, that if they had teams whose
counts did not agree that they could always find the discrepancy.
But they also told me that they had to count through the night and into the next day,
finishing about noon. That is a long night.

And - you might say -its worth it.
Well, to you it is, but to the election officials and polticians -
knowing that they will be working all day, all night and part of the next day -
it is going to take alot to engage their support.
And they do hold tremendous sway with lawmakers.
It would take a major upheaval that the entire population of this country
was aware of, to bring about the type of radical change that some folks
are looking for.

Or it could take some patience.

Some folks fear that if HR 550 or any federal law passed that restricted
how voting machines were used, set audit requirements, made disclosure of source code
mandatory, and elimination of wireless voting machines -
that we would in effect be saying - voting machines are o k.

The fact is, a law like HR 550 - a law like NC's Public Confidence in Elections -
makes it hard to be a voting machine vendor. HR 550 more so, because it is
tougher and it means radical change in electronic voting, enough to most likely
bankrupt many companies, and cause major metamorphasis of any that plan
to remain. NC's law ensured that counties could choose paper ballots if they wanted.

We're only now reaching mainstream media on this!

Lou Dobbs is a first step into the mainstream or regular media,
and he is just awakening a small portion of the sleeping public.
http://www.votetrustusa.org/index.php?option=com_content&task=view&id=1406&Itemid=27

Its a very hard sell.

Look at how hard the battle against tobacco has been



And that has taken decades!

Everyone has seen TV ads about how cigarettes cause cancer!

Have you seen any adds that say electronic voting loses votes or
can be rigged?

So anything that sets the bar high is a good step, whether you are for paper ballots
scanned by optical scanners, or if you are for hand counted paper ballots.