In Layman's terms, Please, Explain to me what a "Random Audit means"

[kster]

because I believe the only way to secure, without a doubt, that the Opticscan machines, vote count was correct, is to Hand Count All The Optiscan Paper Ballots By Hand, and compare the two counts.

Some people believe that a Random Audit will do the same thing. I need to know why, a Random Audit of the Optiscan Paper ballots, would be as secure as a Full Hand Count of All the Optiscaned Paper Ballots.


What am I missing? If you are for Random, Hand Counted Paper Ballot Audit of the Optiscaned Paper Ballots to secure our elections, then , DEFINITELY, You should be for a 100% Optiscan Paper Ballot Hand Counted "Audit" to secure our elections.

Am I missing something here? :)

[WCGreen]

1. Random Audit, at least when I was on the Board of Elections,

meant selecting a ramdom number of precincts and then doing a hand count...

[kster]

3. Right, so Audit is another word for "Hand Counted Paper Ballots"

because that is what they are doing Hand Counting SOME of the Paper Ballots or "AUDITING" , The Audit as they say is actually a Hand Count of the Optiscaned Paper Ballots, The Election Reformers want to count ALL of the Paper Ballots. The people getting voted in by these crooked machines want to limit us (the Election Reformers) to counting only 2% of these Optiscaned Paper Ballots,

Why do you think that is?

[WCGreen]

7. Because it's a statistically valid sample....

But that doesn't mean they will get it right...

[Bill Bored]

2. Here's the difference.

A properly conducted random audit can confirm that the outcome of the election is correct.

Only HCPB can confirm that the actual vote count is correct.

However, audits being proposed in federal legislation do not even confirm that the outcome is correct with high confidence in many cases.

That unfortunately is the reality in which we find ourselves.

[BeFree]

4. Define, if you will

"A properly conducted random audit"

IOW, what are the proper steps, amounts of counts, and other particulars of a random audit that would be certain to confirm the outcome of any election. Never have seen that definition or the particulars listed and it'd be nice to see before I die, eh? Thanks ahead of time, BB, if indeed you wish to undertake such a task.

[kster]

5. If you are for a 2% audit , then you should be, big time for a 100 % audit

thats what the Election Reformers want. If the Politicians want to make a law to Hand Count 2% for an AUDIT and the Election Reformers want to count 100% for an AUDIT.

What is the problem? What is better 2% for accuracy, or 100%?

[OnTheOtherHand]

6. well, what you say is untrue

Many election reformers do not support a 100% hand count in all cases. And I think you know this already. If you're going to insist that Andy Stephenson wasn't an election reformer because he didn't agree with you on hand counts, well, that's your choice, and I think it is a revealing one.

[kster]

9. OK, I'll try it this way, do you agree

that we should randomly audit/hand count some of the ballots to see if the machine count was correct? If so, what % of the ballots should be randomly audited/hand counted?

[OnTheOtherHand]

10. well, hang on

First it would be useful to know whether you are committed to civil conversation. I'll try it this way: do you agree that it's improper to say that "the Election Reformers want" a 100% audit when obviously that isn't true of all election reformers? If you don't, then why do you think it is acceptable? If you do, then will you apologize for your earlier post?

Nevertheless, in case any lurkers are trying to understand this issue, I will try to answer your questions in the meantime.

Yes, I agree that we should randomly hand count some ballots to check the machine count. As several of us have explained repeatedly, if the object is to confirm the outcome of the election, then there is no set percentage that will work in all cases. In a large statewide race that isn't very close, a 1% audit or less may be plenty. In a close congressional race (or a very close statewide race), up to a 100% audit may be necessary.

I would support auditng some minimum percentage or number of ballots even if a smaller sample would be enough to verify the outcome. I don't have a strong opinion about what that minimum should be.

[FogerRox]

8. Is it like when I buy a cup of coffee

With a ten dollar bill, and I only count the pennies to see if I was given the right change back?

/snark

[kster]

11. So, You Hand Count the Copper Pennies?

I wish I would have thought of that, why do you audit the change back? Are you auditing the register machine COUNT, or are you auditing the employees, COUNT of the change? Wouldn't it be more secure to Hand Count all of the change you got back? :D :D

[Bill Bored]

12. Not really. It's more like this:

Assume this is a Starbucks where each cup of coffee costs several dollars.

In a year, you buy 1000 cups of coffee;
you pay for every cup with the exact change in pennies,
except for 10 cups where you shortchange the seller by giving her less than the correct number of pennies.

How many times would the seller have to count ALL the pennies you give her to find out that you've been cheating?

That's the problem.

With HCPB, the seller would hand count every penny every time.

With an audit, the seller would randomly count only enough times to find out that you have been cheating. At that point, a full hand count, or a larger audit would be triggered. But if the seller didn't find the undercounts, she could safely assume that you haven't been cheating, or that you haven't been cheating enough to affect her bottom line.

The answer to the above problem BTW is 368 random counts out of a possible 1000, or a 36.8% hand count. This provides 99% confidence that you weren't cheating. But in reality, the seller would have a 99% chance of finding at least one of your attempts to underpay.

[FogerRox]

13. yeah, what he said.

[Bill Bored]

14. Yeah, so what this MEANS is....

...that in a close race with 1000 precincts, by counting far fewer ballots than 100% with the audit, you could still detect that someone might be screwing around. Of course it's more complicated with actual precincts (suppose you bought a different SIZE latte each time you went to the Starbucks?) but none of this is rocket science to those who have studied the problem seriously.

[BeFree]

15. Real easy

Take the paper ballots and make two piles (or three depending on how many candidates are running) place votes for candidate A in one pile and B in the other.

Look at the size of the stacks and compare to the machine count. If around 50-50 hand count all the ballots.

Too simple? Yeah.

[kster]

16. Or get rid of the machines and buy an Auto stacking system

[BeFree]

17. That would work

But take Florida, for instance. The new governor there has proposed spending $32M for machines. Now that's some serious pork and will bring home the bacon.

The Auto stack idea would cost the machine heads that influence, keep corruption to a minimum, and be so simple even dunderheads like me could make heads and tails out of the process. The last thing they want is to have us looking into this mess and keeping it honest.

[kster]

19. With 32 million, you can probably buy ballot sorter/stackers

for the whole country, and have money left over to buy pizzas and soft drinks for all the people that will be Hand Counting the Paper Ballots.

B-)

[Febble]

18. kster, could you explain

how your system would work for an election with 20 or so races? Do you envisage a separate ballot and a separate box for each race?