Democratic Underground Latest Greatest Lobby Journals Search Options Help Login
Google

Audits in Primary Elections?

Printer-friendly format Printer-friendly format
Printer-friendly format Email this thread to a friend
Printer-friendly format Bookmark this thread
Home » Discuss » Topic Forums » Election Reform Donate to DU
 
garybeck Donating Member (1000+ posts) Send PM | Profile | Ignore Thu Aug-05-10 07:27 PM
Original message
Audits in Primary Elections?
Hey there buddy... got a question for ya. I am wondering if you know if any of the states that have adopted audits are doing it for their primary elections?

We are coming up on a primary election for Governor in here in VT. The SoS is running for governor. We don't have mandatory audits. In the last few elections they have done some limited (lame) audits on the general elections, just at the choice/discretion/whim of the SoS. Now she's running for governor and there are 4 others running against her in the democratic primary. Guess what, she says we don't need any audits in the primary.

Conflict of interest?
Double standard?
Stupid?
Illogical?

yes yes yes yes.

but I'm wondering if other states who have instituted mandatory audits, if they apply to primary elections or only general elections. this information will help me beat the drum here in VT.

any info is appreciated.

thanks
gary
Refresh | +3 Recommendations Printer Friendly | Permalink | Reply | Top
Stevepol Donating Member (1000+ posts) Send PM | Profile | Ignore Thu Aug-05-10 08:19 PM
Response to Original message
1. How many states have mandatory audits? And what are they? 1%?
Printer Friendly | Permalink | Reply | Top
 
garybeck Donating Member (1000+ posts) Send PM | Profile | Ignore Thu Aug-05-10 08:43 PM
Response to Reply #1
2. that doesn't answer my question but...

the dark green states have mandatory audits
you can look up the particulars on verifiedvoting.org
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Thu Aug-05-10 10:25 PM
Response to Reply #2
3. Caveat Emptor
With one exception, the "Manual Audits" employed in the nation will have an extremely low chance of catching an election flipping error. PERIOD. They simply don't call for ENOUGH ballots to be counted in close elections (the time you'd need an audit most :crazy: ).

The one exception is New Jersey, where they have an audit law that has an extremely high chance of finding such error--'cept one thing. NJ uses paperless DRE's. :grr:

Anyone looking solely at Verified Voting's main page would have absolutely no idea that the audits are that useless. And I have absolutely no idea why that is.

Verified Voting: Mandatory Manual Audits of Voter-Verified Paper Records
http://www.verifiedvoting.org /

Printer Friendly | Permalink | Reply | Top
 
garybeck Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 12:25 AM
Response to Reply #3
4. I agree with what you're saying but that doesn't answer my question
My question is not about the worthiness of audits. that's another discussion. my question is whether the states that have mandatory audits, if they implement them in the primary elections.

while i agree with your basic response, I would add, that in my opinion ANY audit is better than NO audit. Even a 1% audit is better than no audit.

either way, i'm just tring to find out if these states do audits during the primaries or not.

thanks
Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 07:57 AM
Response to Reply #3
8. I don't think that's "PERIOD" at all
One can't determine the "chance of catching an election flipping error" without making some assumptions about the nature of the error. If one assumes that the errors entail switching relatively small numbers of votes in contests that were close to begin with, then the auditing problem is very hard -- but in that case, election-flipping probably isn't common. (It may depend on what kinds of elections you are thinking about.) Very loosely speaking, the more common election-flipping errors are, the more useful even a crude audit protocol is for detecting or deterring them.

For statewide contests, California's 1% manual tally is kind of huge. (For smaller contests, nothing huge about it -- but it still isn't useless.) Of course Vermont is much smaller, and a 1% audit there would be barely a sneeze.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 11:08 AM
Response to Reply #8
10. You say 1% is huge in CA, without suggesting a reported margin.
Edited on Fri Aug-06-10 11:17 AM by Wilms
So how wide a race before you claim 1% is huge? A 5, or 10% margin? How 'bout for a 1% margin? Still "huge"?

Statistically speaking, what confidence percentage do these other states' audits provide in close statewide elections?

For even more fun, tell our readers the chance of any of these catching jack-shit in a <statewide contest.

OTOH, if an audit designed to verify a vote can't do that it is "u s e l e s s" except for those paid to BS about them.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 11:41 AM
Response to Reply #10
11. jeepers
Wilms, why don't you hand-count all the contests on all the ballots from over 600 precincts yourself and then see whether you feel like contesting that the sample is "kind of huge"? Are you seriously suggesting that if one of the contests is close enough, then the sample becomes smaller?

"Statistically speaking," you would have to define what you mean by "confidence" and "close" (at the very least) before your question could be answered.

As I already explained, there is no way to calculate "the chance of any of these catching jack-shit in a <statewide contest." If you think I'm being paid to BS about that, I guess I might as well seek out TIA for some more thoughtful discussion.
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 12:51 PM
Response to Reply #11
14. Lots of problems with audits though. (And I don't get paid to deal with them either!)
Edited on Fri Aug-06-10 01:21 PM by Bill Bored
Improper sampling methods, improper escalation methods, improper random selection methods, lack of expertise or any due diligence in the legislative and rule-making processes, excessive complexity, etc.

Even if a statewide audit is large enough to find serious problems somewhere in the state when the outcome of a statewide election is wrong (say in a county or two), this rarely, if ever, results in expansion of the audit where problems happen NOT to be found. Since the per-county samples are small (i.e., fewer audit units) compared to the statewide sample, this is bad design. It assumes that problems are limited to the county or counties in which problems happen to be discovered. Maybe this is what Wilms is referring to.
Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 10:02 AM
Response to Reply #14
23. yeah, county-by-county escalation makes little sense
The best thing I can say about that is that sooner or later (assuming it isn't fixed), there will be a case where it's obvious that more auditing or a recount is necessary -- and then the need will be harder to ignore.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 10:38 AM
Response to Reply #23
25. *cough*
I think you have a point.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 01:05 PM
Response to Reply #11
16. Don't take it personnel.
If I thought YOU were paid, I would have said so.

And no, I didn't suggest the sample be smaller in a close race, I'm trying to point out that a fixed sample provides less and less confidence as the margin shrinks.

So tell me. What do you think it would take to get a a >90% confidence in one of the state's state-wide elections where the margin is, say, 2%?

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 08:38 PM
Response to Reply #16
20. hang on, which state?
The percentage would have to be a lot larger in VT than CA.

I certainly agree that a fixed sample provides less and less confidence as the margin shrinks. It's the "useless" that I take strong exception to. I think it's like saying that a medical treatment is "useless" unless it has a guaranteed high probability of curing every case of disease.

(I see you started a new thread on FL -- I haven't read it yet, but FL's audit verges on useless.)
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 09:22 PM
Response to Reply #20
21. What do we call medical treatments with low probability of helping?
And do we see published probability numbers with these audit results? (That would get ugly very quickly.)

My issue is the voters being told that the elections are audited leaving them to assume there is a high confidence in the outcome when there isn't. Actually, harmful might be a description for that. Particularly in close races.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 08:12 AM
Response to Reply #21
22. hmmmm
Edited on Sat Aug-07-10 08:13 AM by OnTheOtherHand
I don't concede the apparent premise that audits have a low probability of helping. You seem to be taking the effectiveness of an audit protocol in a worst-case (or almost worst-case) scenario and treating it as the first and last word on the protocol. If we assessed medical treatments in that way, we'd consider almost all of them useless. (I don't know if you're actually doing that, or whether we're just talking past each other.)

I think published probability numbers are a good idea, although they would be very widely misconstrued. If I reckon that a particular audit gives 10% "confidence," that basically means that if the contest was actually tied, and if miscounts were packed into the fewest possible units without anyone otherwise noticing, then 10% of audit samples would contain more miscounts than the observed audit sample contained. In practice, even an audit that yields 10% "confidence" under that definition can provide useful additional evidence about the election outcome. What it can't do is to provide strong evidence on its own -- and, yes, it's important for people to know that.

Of course, for people who regard the scanner counts as inherently utterly unbelievable, the audit results may be construed as the only evidence about the election outcome. That isn't my view.

My issue is the voters being told that the elections are audited leaving them to assume there is a high confidence in the outcome when there isn't. Actually, harmful might be a description for that. Particularly in close races.

That's a legitimate concern, but not the only one. One of my concerns is that if we insist that New York's 3% audit is useless or worse than useless, someone will believe us and gut it as a cost-cutting measure -- at which point all the "confidence" numbers go to zero. Some people may think that's just fine, but from my standpoint, it would sacrifice a very large amount of information about how the election systems are functioning.

ETA: I still wonder which state you had in mind in your earlier question.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 10:37 AM
Response to Reply #22
24. Seems like you're overemphasizing that an audit with low confidence could still reveal a problem.
I think everyone gets that. A smaller chance...but a chance nonetheless.

In particular, I'm referring to states in general. :) Pick a few if you want. Cherry pick as you might.

But what not everyone gets is that most of these states audits...in a given race...will yield a low confidence which means it has a much smaller chance of catching a problem. That's a danger. Pointing out the "uselessness" of that arrangement isn't dangerous at all. In fact, the opposite is true. I didn't say "audits are useless". I said the ones we have are because in the closest of races (when you need them most) they don't provide a high level of confidence.

And the idea that the audits could be gutted because someone points out the limitations is a bit...well. I think there's a MUCH greater danger of voters thinking all is well because they voted on a paper ballot. (Case closed.)

I think you generally agree with my concern even while you otherhand it to death. But I'm ok with that.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 07:40 PM
Response to Reply #24
27. I'm hardly making that point at all
You seem to be essentially disregarding all the contests where the confidence level wouldn't be low -- never mind the fact that the "confidence level" is generally much smaller than the actual chance of detecting a problem.

"...most of these states audits...in a given race...will yield a low confidence..."

In what "given race"? I daren't engage your points when I can't tell what they are. You seem to think you've made a cogent argument that actually existing audits are useless, but I can't find it. Maybe we need to start from scratch.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 07:53 PM
Response to Reply #27
29. OK, then. Can you name any races with a 2% margin where an audit achieved >90% confidence?
Edited on Sat Aug-07-10 07:56 PM by Wilms
Also, I'll assume you'd like to see some improvements in auditing. What would they be?
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 11:24 PM
Response to Reply #29
30. How 'bout a 2% margin in a US House, State Senate or State House (or Assembly) race?
Edited on Sat Aug-07-10 11:32 PM by Bill Bored
Just to make it interesting. :evilgrin:
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-08-10 02:01 AM
Response to Reply #30
31. Yep. If guess if we exclude President, US Senate, and Governor, it's really game over. n/t
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-08-10 03:06 AM
Response to Reply #31
34. Well there is the argument that if errors happened on EVERY machine, ANY audit would find some.
Edited on Sun Aug-08-10 03:25 AM by Bill Bored
But that's a really stupid way to try to steal an election.

That said, a lot of states don't audit non-statewide races at all!
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 01:37 PM
Response to Reply #22
26. NY's audit has already been gutted. The SBOE's version was dumb but the election law's is dumber!
Edited on Sat Aug-07-10 02:04 PM by Bill Bored
In 2009, the SBOE required an audit of at least 1 machine of each ballot style, which at least tested...1 machine of each ballot style. There's some logic to that, but it didn't do much for close contests. Advocates said as much, but county election officials took a different view. They said they should only have to count 3%, even if all contests or ballot styles are NOT audited, because that's what the election law says.

Instead of consulting with experts to find a better way (due diligence and all that), the SBOE gutted their own audit and reverted to the 3% in the election law.

Some have responded rather sheepishly with: "Oh well, at least every race has a CHANCE of being audited."

I think the problem with some advocates is that they keep trying to see the glass as half-full, no matter how empty it is. And they don't want to admit that getting rid of lever machines in New York (or elsewhere) was a HUGE mistake.

For some, not doing the audit math is enough to keep them in denial.

Others don't believe in stolen elections anyway, and simply trust that the better side of human nature will prevail, even as they make the tools available to facilitate rigged elections.
Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-07-10 07:42 PM
Response to Reply #26
28. as long as we're waxing metaphorical
I think your overwhelming problem is that you can't see the glass, so you have no clue how full it is. It's too bad.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-08-10 02:24 AM
Response to Reply #28
32. Being a bit harsh?
Apart from Bill venting, he made a substantial charge much larger than his snipe or the reply. :shrug:
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-08-10 03:00 AM
Response to Reply #28
33. Well if that's the best you can come up with...
Edited on Sun Aug-08-10 03:04 AM by Bill Bored
...the problem is no one can see the glass; it's just ones and zeros.

You might think it's cool that NY is doing something better than some other state like Georgia or South Carolina. But there are a lot of things going down in NY that stink to high heaven right now, or at best are the result of some serious incompetence.

There's no way anyone could have possibly anticipated that overvoting problem, is there? :sarcasm: That's not all we should be concerned about.
Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 07:32 AM
Response to Reply #2
7. AFAIK the best resource for this kind of question
is Citizens for Election Integrity Minnesota's State Audit Legislation Reference Guide. Unfortunately it doesn't clearly distinguish which states' audits apply in primaries -- but it links to applicable state laws, so it usually isn't too hard to figure out.

I don't know just how many states audit primaries, but I know that Connecticut's audit applies to primaries (as does NY's, as Bill Bored already mentioned). New Mexico's definitely doesn't. California's 1% manual tally applies to all elections in which voting machines are used. In Ohio, Jennifer Brunner made a similar decision not to audit the primary in which she was running for the Senate nomination (she lost). (Ohio has no audit law; I think there's a court order that requires some sort of auditing in major general elections.)
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 01:17 AM
Response to Original message
5. NY does. That's close enough to VT. nt
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 02:51 AM
Response to Reply #5
6. But do you know, generally, if primaries are included, or not? n/t
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 12:38 PM
Response to Reply #6
13. No, I don't. Some states definitely include them. NY, NC, maybe others. nt
Edited on Fri Aug-06-10 12:54 PM by Bill Bored
Printer Friendly | Permalink | Reply | Top
 
garybeck Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 10:27 AM
Response to Reply #5
9. do you know if they audit their primary elections? n/t
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 12:35 PM
Response to Reply #9
12. YES, as I said, NY audits primary elections. I'm not sure what the other states do. nt
Edited on Fri Aug-06-10 01:02 PM by Bill Bored
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 01:01 PM
Response to Reply #9
15. Here's the NY audit law:
9-211. Audit of voter verifiable audit records. 1. Within fifteen
days after each general or special election, and within seven days after
every primary or village election conducted by the board of elections,
the board of elections or a bipartisan committee appointed by such board
shall manually audit the voter verifiable audit records from three
percent of voting machines or systems within the jurisdiction of such
board. Voting machines or systems shall be selected for audit through a
random, manual process. At least five days prior to the time fixed for
such selection process, the board of elections shall send notice by
first class mail to each candidate, political party and independent body
entitled to have had watchers present at the polls in any election
district in such board's jurisdiction. Such notice shall state the time
and place fixed for such random selection process. The audit shall be
conducted in the same manner, to the extent applicable, as a canvass of
paper ballots. Each candidate, political party or independent body entitled
to appoint watchers to attend at a polling place shall be entitled
to appoint such number of watchers to observe the audit.

2. The manual audit tallies for each voting machine or system shall be
compared to the tallies recorded by such voting machine or system, and a
report shall be made of such comparison which shall be filed in the
office of the state board of elections.

3. The state board of elections shall, in accordance with subdivision
four of section 3-100 of this chapter, promulgate regulations establishing
a uniform statewide standard to be used by boards of elections to
determine when a discrepancy between the manual audit tallies and the
voting machine or system tallies shall require a further voter verifiable
record audit of additional voting machines or systems or a complete
manual audit of all machines or systems within the jurisdiction of a
board of elections. Any board of elections shall be empowered to order
that any such audit shall be conducted whenever any such discrepancy
exists.

4. If a complete audit shall be conducted, the results of such audit
shall be used by the canvassing board in making the statement of canvass
and determinations of persons elected and propositions rejected or
approved. The results of a partial voter verifiable record audit shall
not be used in lieu of voting machine or system tallies.

5. Notwithstanding subdivision four of this section, if a voting
machine or system is found to have failed to record votes in a manner
indicating an operational failure, the board of canvassers shall use the
voter verifiable audit records to determine the votes cast on such
machine or system, provided such records were not also impaired by the
operational failure of the voting machine or system.
Printer Friendly | Permalink | Reply | Top
 
garybeck Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 01:42 PM
Response to Reply #15
17. thanks very much!!!! n/t
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 01:53 PM
Response to Reply #17
18. You're welcome. You might also look at what CT does.
It's another nearby state and close in size to VT. Also uses the Diebold eqpt. Contact this guy:
http://www.ctvoterscount.org/
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-06-10 02:25 PM
Response to Reply #18
19. Thank Bill.
:thumbsup:
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Aug-14-10 02:05 PM
Response to Original message
35. Kicking for audits with high confidence levels. Anyone got one? n/t
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-15-10 12:53 AM
Response to Reply #35
36. I've seen some where they audited up to about 75% of the vote to get to about 75% "confidence."
Of course there could be other ways of doing such calculations and the 75/75 thing was just a coincidence. Here's a link:
http://www.usenix.org/event/evtwote09/tech/full_papers/...
See the table on the top of page 11.

As previously mentioned, smaller relative sample sizes would be required for statewide races.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-15-10 01:18 AM
Response to Reply #36
37. Well, I kind of figured that was the case, and the reason OTOH didn't respond.
Edited on Sun Aug-15-10 01:23 AM by Wilms
I think once he understood what I was asking for--audits that yield high levels of confidence in close non-state-wide elections--he realised there wasn't going to be much in the way of encouraging news.

While I agree the near universally flawed audits occasionally conducted could act to deter a thief, they'd have no influence on Murphy's law as it relates to electronic hardware, software, and ballot programming.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-15-10 01:59 PM
Response to Reply #37
38. you were?
Because what you wrote was, "What do you think it would take to get a a >90% confidence in one of the state's state-wide elections where the margin is, say, 2%?" I never did establish what you meant by "the state[]," but I thought at least "state-wide" was clear enough.

Certainly 3% audits can easily fail to detect some outcome-changing errors in small contests -- although if such errors are common, some of them will be detected. If we want more rigorous audits, we can pay more for them. Or not. People actually get to discuss the issue and choose. At least, that's how it ought to work. Arguing about it on DU doesn't make much difference, obviously.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Aug-15-10 02:47 PM
Response to Reply #38
39. You're having a real tough time with this. Aren't you?
No matter. Anyone paying attention will be hard-pressed to find a good reason to have faith in these audits even IF we "argue" about it on DU.

You've offered no example of a close election where an audit yielded a >90% confidence in the outcome.

Ya don't need to be a statistician to understand that.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Aug-16-10 11:03 AM
Response to Reply #39
40. there go the goalposts again
From "statewide" to "non-statewide" to "close." It just seems remarkably sloppy to me. And I'm not asking anyone to have "faith" in any kind of audit. Yes, I have a hard time conversing with people who seem like they are most comfortable talking to themselves, especially because that isn't what I expect from you.

In California, it works out that if 1% of precincts contain miscount, a 1% simple random sample has about a 92% chance of including at least one miscounted precinct. With sufficiently conservative assumptions, one could shoehorn maybe a 4% miscount into 1% of precincts -- but it would stick out like a sore thumb. Even 0.4% would be hard.

I find that math impossible to reconcile with what you wrote in post 3. You don't seem to be defending post 3, just busting my chops. OK, good luck with that.
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Aug-16-10 08:13 PM
Response to Reply #40
41. How about running the same numbers for VT since that's what this thread was supposed to be about? nt
Edited on Mon Aug-16-10 08:18 PM by Bill Bored
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 12:02 AM
Response to Reply #41
43. Don't waste his time unless it's a real audit in VT.
Is it risk-limiting? Does it call for escalation? Are these procedures actual law, and followed?

These are all the things I want to know about audits.

So before you bug OTOH, let's get an answer to the above questions.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 08:31 AM
Response to Reply #41
44. not even
Wilms said something kind of crazy and I politely pointed it out, although I don't think he gets it. What the thread was supposed to be about got covered in the first 24 hours; the rest has been time-wasting.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 10:10 AM
Response to Reply #44
47. Inconvenient truths prevent a direct reply?

"What the thread was supposed to be about got covered in the first 24 hours; the rest has been time-wasting."...is a cop out.

Take the marbles out of your mouth and answer the question...or don't, but spare us your indignation.

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 06:21 PM
Response to Reply #47
48. yadda yadda yadda
Wilms, I really don't know what game you're playing that you pretend not to be playing. All I know is that I'm replying to yet another crappy, off-topic post.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 06:47 PM
Response to Reply #48
49. You didn't answer a single question. Pathetic.
Don't give me any of this "off-topic" nonsense. We're talking about audits. It's an election reform forum.

Audits in this country are unable to confirm most close elections. Deal with that, or go back to whatever this thread pulled you away from.

I'm calling out audits. Don't like it? Then hand-count, or get a lever machine, or just keep letting the vast majority of election results be determined by computer. But stop BSing the forum members about the reliability of those results.

Funny you should be defending ineffective protocols for determining election outcomes. Ya got a little TIA thing going there, don't you?

Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Aug-16-10 11:59 PM
Response to Reply #40
42. Wanting to know the confidence level of various elections isn't moving the "goal post".
And I'm not here to get into a pissing match, which that term alone suggests.

I want to know what elections I can have a high level of confidence in.

"From "statewide" to "non-statewide" to "close."". Exactly. I WANT TO KNOW WHO WON. Not unreasonable on it's face particularly with respect to discussing audits. Don't call that sloppy.

If you want to talk about it, I'd like that very much. And I'm not talking theoretical, I mean practically. So we need to know, for example, which states audits don't call for escalations so we know which ones might catch a problem but not necessarily put you in a place where you can do anything about it. Is that not right?

So in the California example, we might be ok for state-wide, but not congressional on down. Correct?

Are there any other states that would perform similarly?

Printer Friendly | Permalink | Reply | Top
 
OnTheOtherHand Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 08:59 AM
Response to Reply #42
45. I think your actual posts have been sloppy
When you change the terms of discussion from post to post, that is sloppy. Trying to figure out what a 1% audit can and can't do is a very different enterprise than defending the assertion that it is "useless." I don't know if you're actually defending that assertion, because you are changing the terms of discussion so that it is very hard to tell what you are saying.

Yes, I think in the CA example "we might be ok for state-wide..." is not a bad summary of that part (brief, of course) -- the manual tally isn't well designed for the purpose of correcting wrong outcomes, but if someone is wondering, say, if Prop 8 was straight, it gives a lot of information. And, yes, it's not very convincing for smaller contests. The ability to get partial recounts of a few precincts/batches (without paying for all the absentee ballots to be hand-sorted, for instance) would help a lot in getting a handle on how much error there might be in any one batch. (Unfortunately, short of laying out a Bayesian framework, there's no good way to quantify how useful it is.)

A 3% audit in New York ought to be about as useful as a 1% audit in California -- more or less, depending on the details of what gets audited. It's more like a smoke alarm than an actual risk-limiting audit -- as we've discussed, escalation on the county level doesn't make sense -- but as a smoke alarm it could be pretty good if done well. A 5% partial recount in Ohio is kind of similar, but only applies to one contest. If Florida picked a contest and audited it statewide, 1% would be kind of similar too. Connecticut's 10% audit isn't as good because it's 10% of a much smaller number.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 10:07 AM
Response to Reply #45
46. In other words, none of these audits provide high confidence levels. Correct?
Edited on Tue Aug-17-10 10:10 AM by Wilms
I don't won't to put words in your mouth, but while insisting that I'm moving the goal post, you've not really come out and directly answered.

Answer this question, please: If an anomaly is found in CA, does the audit provide...by law...for escalation?

I'm not much interested in an audit that "gives a lot of information" or act as "smoke alarms". Nor should you be citing "ifs", "buts", and "coulds", as you are. I've made it sufficiently clear what I am asking for. The difficulty in answering my question is not my fault. I want a high-level of confidence in election outcomes, particularly (pardon me) when it's a close race. If it doesn't provide that, the audit is..."useless" for the purpose of correcting the outcome.

It's apparent that an audit that doesn't provide for escalation, or the ability to correct the outcome, can provide a high level of confidence that the election result is WRONG. But that's obviously not what I'm interested in.

OTOH it death, if you will. Insult me a few more times, if you like. Insist that asking for high confidence levels in various election situations is "moving the goal post". Tell me how unimportant my posts are compared with people you deem important. The facts remain...there are no audits you've mentioned that will meet that criteria with the possible exception of "one contest" in OH.

Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 10:43 PM
Response to Reply #46
51. Ohio didn't provide for a proper escalation procedure, so it's no exception. nt
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-20-10 04:33 AM
Response to Reply #51
53. Does NJ's audit law provide for escalation?
No paper to audit there yet, of course. But that audit has a pretty good housekeeping seal, for state-wide anyway.

Given a state-wide paper ballot system, how would NJ's audit perform with regard to my (sloppy, goal-post moving :eyes: ) query about the state of (Un)Verified Voting?

Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Fri Aug-20-10 12:04 PM
Response to Reply #53
54. By law, NJ's audit must meet certain requirements that would require escalation:
Edited on Fri Aug-20-10 12:09 PM by Bill Bored
(1) Any procedure designed, adopted, and implemented by the audit team shall be
implemented to ensure with at least 99% statistical power that for each federal, gubernatorial
or other Statewide election held in the State, a 100% manual recount of the voter-verifiable
paper records would not alter the electoral outcome reported by the audit. For each election
held for State office, other than Governor and Lieutenant Governor, and for county and
municipal elections held in 100 or more election districts, any procedure designed, adopted,
and implemented by the audit team shall be implemented to ensure with at least 90%
statistical power that a 100% manual recount of the voter-verifiable paper records would not
alter the electoral outcome reported by the audit. Such procedures designed, adopted, and
implemented by the audit team to achieve statistical power shall be based upon scientifically
reasonable assumptions, with respect to each audited election, including but not limited to:
the possibility that within any election district up to 20% of the total votes cast may have
been counted for a candidate or ballot position other than the one intended by the voters; and
that the number of votes cast per election district will vary. Such procedures and
assumptions shall be published prior to any given election, and the public shall have the
opportunity to comment thereon.

(2) Any procedure designed, adopted, and implemented by the audit team for each county
and municipal election held in fewer than 100 election districts, but more than a single
election district, shall be conducted in at least two election districts.

(6) ...the audit team shall
have the authority to cause audits to be conducted of any election district or audit unit which
has not been randomly selected for auditing in which a majority of the audit team determines
from the un-audited election results, past election results, or other data that the votes are
likely to have been miscounted.

(9) If the Attorney General, based on a recommendation of a majority of the professional
audit team, determines that any of the hand-to-eye counts conducted under this section show
cause for concern about the accuracy of the results of any election in the State, or in a county
or a municipality, or with respect to a particular election, the independent audit team shall
oversee, supervise, and cause to be conducted hand-to-eye counts under this section in such
additional election districts or audit units as the Attorney General considers appropriate to
resolve any such concerns. The Attorney General shall issue previous to any election the
criteria to be employed to determine whether the hand-to-eye counts show concern about the
accuracy of the election results in order to trigger further hand-to-eye counts. Such criteria
shall be published prior to use in any given election, and the public shall have the
opportunity to comment thereon. Notwithstanding the requirements previously set forth in
this paragraph, additional hand-to-eye counts shall be conducted if in the initial audit
conducted pursuant to the procedures set forth in this subsection, any discrepancy or
discrepancies attributable to the electronic counting system would alter the vote share of any
candidate or ballot position by one tenth of one percent or more of the hand counted votes in
the sample. Under such circumstances, the audit of the election shall be expanded using the
same number of election districts and when possible, audit units, as the initial audit and shall
be conducted under the same procedures used to conduct the initial audit, provided, however,
that if the initial audit comprises more than one half the total number of election districts and
audit units in the election, the expanded audit shall be a full hand-to-eye count of the
remaining un-audited election districts and audit units. Further hand-to-eye counts shall be
conducted if any discrepancy or discrepancies attributable to the electronic counting system
detected by the initial or subsequent expanded audit indicates a substantial possibility that a
complete hand-to-eye recount would alter the outcome of the audited election.
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Aug-23-10 02:03 AM
Response to Reply #54
55. OK. That's how it looked to me.
That's why I wrote what I did in reply #3.

OTOH reports being confounded by the post, but couldn't acknowledge the NJ reference it contained. And this, despite the fact he lent considerable support to it's passage, as I recall. Doing so, however, would beg comparison with other states.

Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Aug-23-10 07:10 PM
Response to Reply #55
56. NJ's audit is the only one with a large chance of finding and correcting serious miscounts in all
Edited on Mon Aug-23-10 07:23 PM by Bill Bored
Federal and State elections in a state, including all elections for members of the State Legislature -- not just statewide elections.

It is instructive to point out that the reason there was an audit bill in NJ was because of a lawsuit that resulted in a court decision to require VVPATs or VVPBs and audits. The legislature and governor enacted the audit law, but then they allowed the paper requirement to become hopelessly delayed by changing the deadline for compliance in the VV-paper law.

An appeal will likely be filed, but the lawsuit has been in the courts for YEARS with some of the best experts in the country from Princeton testifying in favor of the verification requirements (and demonstrating how to hack the machines too).

So even though NJ has a serious audit law, not only is it unenforceable, but without litigation, it would not have existed and is not likely to become enforceable in the future.

What this means for election integrity advocates is:

1. We probably won't get anywhere without a lawsuit (including lawyers willing to bring it on behalf of voters).
2. Even if we win in the courts, there are tactics that can be used to delay the changes for years.
3. Despite the help of dedicated lawyers and academics working for free, as well as organized lobbying at the state level, and some legislators who are willing to do the right thing, the chances of the NJ audit law, or anything close to it, being implemented or repeated elsewhere are slim to none.

When we talk about groups like Verified Voting, etc. we have a situation where some folks are trying to fight the good fight against what appear to be overwhelming odds. It's no wonder they may sometimes try to paint a rosier picture than what actually exists. The question is what to do about it. Being apologists for the status quo doesn't seem like much of an answer to me -- UNLESS the status quo is a non-computerized voting system!
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-31-10 11:08 PM
Response to Reply #56
57. Thanks Bill.
This is in line with my notorious reply #3.

Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 10:18 PM
Response to Reply #45
50. So how 'bout Vermont then, since it's garybeck's thread?
Edited on Tue Aug-17-10 10:44 PM by Bill Bored
Hi Gary! :hi:
Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Tue Aug-17-10 10:48 PM
Response to Reply #45
52. Of course Florida does not "pick a contest and audit it statewide." nt
Printer Friendly | Permalink | Reply | Top
 
Wilms Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Nov-01-10 09:12 AM
Response to Original message
58. OK. Jersey (and did someone say NC?) have statistical audits.
Any others...

Lot's of tight races are going to be determined by computers tomorrow.

Printer Friendly | Permalink | Reply | Top
 
Bill Bored Donating Member (1000+ posts) Send PM | Profile | Ignore Mon Nov-01-10 07:25 PM
Response to Reply #58
59. New Mexico. But don't count Jersey because they don't have paper ballots. nt
Printer Friendly | Permalink | Reply | Top
 
DU AdBot (1000+ posts) Click to send private message to this author Click to view 
this author's profile Click to add 
this author to your buddy list Click to add 
this author to your Ignore list Tue Sep 16th 2014, 11:11 AM
Response to Original message
Advertisements [?]
 Top

Home » Discuss » Topic Forums » Election Reform Donate to DU

Powered by DCForum+ Version 1.1 Copyright 1997-2002 DCScripts.com
Software has been extensively modified by the DU administrators


Important Notices: By participating on this discussion board, visitors agree to abide by the rules outlined on our Rules page. Messages posted on the Democratic Underground Discussion Forums are the opinions of the individuals who post them, and do not necessarily represent the opinions of Democratic Underground, LLC.

Home  |  Discussion Forums  |  Journals |  Store  |  Donate

About DU  |  Contact Us  |  Privacy Policy

Got a message for Democratic Underground? Click here to send us a message.

© 2001 - 2011 Democratic Underground, LLC