Verified Voting: Developing Standards for Election Data

Developing Standards for Election Data

Thursday, December 17, 2009

By John McCarthy, Verified Voting

One of the challenges faced by advocates of election audits and transparency is that current voting systems each record and store election file data in unique ways. This is no surprise given that vendors have long claimed that their systems are proprietary. But the current model of storing election data in ways that prevent easy sharing and analysis is proving difficult for election officials, statisticians, election integrity advocates, and even voting systems vendors. Because of these problems, serious discussion is taking place about what can be done about standardizing election data.

Often, within a single state there are many different voting systems from multiple vendors. At the same time, many elections, including most federal and statewide races, cross election jurisdictions so that votes for the same race are reported in different ways, depending on the system type used in each district. Even a single polling place may have different types of equipment – an optical scanner and a touch screen device for accessible voting for example – which report results in incompatible ways but which must be combined after the polls close.

On election night, results from different systems are sent to a central location where the results need to accurately collected, totaled, and reported to the media and the public. But when election officials have to deal with data in several different formats these tasks become difficult if not impossible. The problem is further exacerbated by several types of voting systems which export data only in PDF format. This format, well know to computer users everywhere, is designed to be read by humans but is not easily imported into spreadsheets to total detailed election results.

snip

http://blog.verifiedvoting.org/2009/12/17/269

more "efficient" and thus easier to hack, more undetectably and at greater speeds.

The solution is to GET RID OF 'TRADE SECRET' VOTE COUNTING!

Having said that, the only value I see is making it easier to conduct a proper audit...if they'll actually do one.

Verified Voting and plenty of activists work within the constraints of e-voting. I have many disagreements with that, but I generally see them as good guys trying to do the right thing.

And getting rid of "Trade Secret Vote Counting" really is only so much help. Is the alternative e-vote counting on open software that crooks have time to study and compromise; open software that I have no way of knowing if it's really installed on a given machine??

No thanks. Hand Count...and/or audit the hell out of it by hand counting.

One reason states don't like to audit elections is because they can't. The counties have too much control of the election results. If you can't get the results, you can't audit them to see if they're true. This is OK for lever machines because the results in each county are not very hackable. It takes many of these machines to rig a county-wide tally.

But as soon as you add computers to the mix, you have to have some centralized reporting of results. (Not necessarily centralized PROGRAMMING, which is fraught with the kind of risk you're talking about -- just centralized reporting, which should of course be checked against local reports too.)

Once you have all the data, it's easier to hand count enough to see if it can be true, or if someone's lying. The less data you have, the more hand counting you have to do, all things being equal. And unfortunately no one wants to hand count one ballot more than is absolutely necessary. (Sad but true.)

So getting the data as quickly as possible is a reasonable way to determine how to audit the election results. There are other ways, but they might be more hit and miss and they might cost more.

Getting rid of secret software is not the answer because any software can be hacked.

On the other hand, delaying the effective auditing of elections until every last bit of election data is available is silly. So we have to do as much as we can with the available data, unless we don't use computers in the first place.

code that anyone may review--and they also handcount a whopping 55% of the votes, as a check on machine fraud (more than five times the minimum needed to detect fraud).

THAT is the failsafe if you are going to use software to count votes--a very substantial audit, at minimum 10%. And nobody in the US is doing this. In fact, about half the systems in the US don't have any paper TO count. It's mindboggling.

In addition to having a transparent system, on its face, Venezuela's elections have been certified by every major international election monitoring group as transparent, fair and aboveboard. That's why they have universal health care and we don't.

Number one on the list of things to reform is getting rid of the secret programming.

As was the case in Sarasota in 2006, when researchers wanted to look at the code to find out what happened, a judge told them NO.

Un believable!! We paid for it and it counted our votes, but we can't break it down to find out why it made so many mistakes?!?!?

The secret code is the main tool to stealing elections and the secret needs to be done away with post haste. Any code used to count our votes needs to be able to be examined at any time. Sure any code can be hacked, but at least we would be able to trace the hack and fix it, not so now with the secret code.

This would take away each vendor having their own individual "proprietary" reporting system that can't talk to any other system and requires a copy of their own proprietary election management system to get data out.

Will make things more transparent and make it easier for us (citizens) to track and keep tabs on things, not harder.

software from the process and use lever machines or hand-counting at the precinct level with public oversight, the problem gets solved and the result is far more accurate, economical, and better for the environment (where do they dump the old DREs and optiscans anyway? And how many barrels of oil does it take to make a DRE or an optiscan?) The 'sort and stack method' for hand counted paper ballots has a 100% audit built into the process and it all happens on election night after the polls close and before the ballots are removed from the precincts.

The fact that the DU software can make two posts out of one
is evidence of the mistake of using software to count votes.

It is how Bush won in 2004.... the software used by the vote
counting computers double posted votes for Bush. Same as DU has been doing.