BBV: opinion of AccuPoll machines?

[Eric J in MN]

I think that plain paper ballots are the best, and optical-scan ballots are second best, but among electronic voting machines, the AccuPoll seems ok based on their website description.

Anyone have an opinion of AccuPoll?

http://www.accupoll.com/

NOTE: my objection to most electronic voting machines is that even if they print paper ballots, they generally require the voter to insert a card, and if the machine doesn't display all the races it should for that voter, complications ensue.

With plain paper ballots or optical-scan ballots, if all the races you should be able to vote in aren't listed, you can tell the polling place worker before casting your vote and receive the proper forms.

[Eric J in MN]

1. kick

kick

[AP]

2. Which is the machine that is all hardware and no software?

It's an optical scan machine which kicks out under/over votes, retains the ballot and counts ballots on the spot, but has no software.

That's SO OBVIOUSLY the way to go. That's even safter than the mechanical counters they use in NY.

[Eric J in MN]

3. The AutoMark from Vogue Election Systems

http://www.vogueelection.com/

The AutoMark from Vogue Election Systems lets the blind use optical-scan ballots, and checks anyone's optical scan ballots for undervotes and overvotes.

I agree that the AutoMark is better than an electronic voting machine.

I was just wondering if anyone had an opinion of AccuPoll.

[Don_G]

4. It Looks Good On The Surface

Linux based open source software with a paper receipt for the voter, but I'll have to research the officers of the corporation first.

[RedEagle]

5. Accupoll Appears to be Pretty Good

Due to the needs of finanacing, if you dig enough, you'll find people you don't like behind most systems out there, even if they appear to be the good guys.

Look who controls and runs it.

Accupoll, as near as I can tell, is based on Linux, HTML, and Java. I believe they are open source, or more so than anyone else.

They print a voter verified paper ballot and are supposed to have a bar code pen for the visually disabled to read the paper ballot back that way, in addition to an audio review. However, the DOJ has written that the audio review does the same thing.

Everyone needs to keep in mind that optical scan is not the entire answer- it provides voter verified paper ballots but we need to robustly audit the results of the COMPUTER count of those ballots, by hand audits.

I was told by one former election official that with the optical scan system, if a race was within about a 2% margin, it was out of range of the machines ability to accurately count. In other words, the machines could not be considered to be accurate up to 100%, there was a margin for machine error. I think this official was exemplory in recognizing and dealing with this.

We don't have any data, even after all these years, as to accuracy of machines- and that data should not come from testing in some lab. It has to come from auditing out in the field over a period of years.

That said, a DRE that produces a voter verified paper ballot should be the ONLY acceptable DRE. Accupoll would meet that criteria.

That is a different argument from DRE or not to DRE. No system should be used without robust, random, hand auditing.