Religion, Politics and World Events They make great dinner conversation, don't you think? plus Political Film

Diebold Machines More Accurate Than Human Vote Counters

Old 10-21-04, 07:00 PM
  #1  
DVD Talk Hero
Thread Starter
 
Join Date: Jul 2001
Location: MI
Posts: 25,054
Diebold Machines More Accurate Than Human Vote Counters

The test is only 50 votes, broken into groups of ten. I can't believe the human counters failed to agree 4 of 5 times; that is just pathetic. (It also probably says the Florida recounting dweebs staring at chads weren't worth shit).

Where is the John Henry spirit here? I'm sure the machine could beat me to 5 million, but I could give it a run for its money to fifty.

http://www.foxnews.com/story/0,2933,136259,00.html
Touch Screen Tally Beats Human Count
Thursday, October 21, 2004
By Joseph Bacchus
STORIES
GLEN BURNIE, Md. The state's new touch-screen voting machines passed a test Thursday by the Maryland State Board of Elections (search), as man and machine went head-to-head to see who could tally votes more accurately. The machine won.

The Diebold voting machines (search) will be used in precincts across Maryland in the Nov. 2 presidential election. For months they have been a center of contention in the state, with some citizen groups saying they are susceptible to error and fraud, which they worry could mean the same sort of vote-counting confusion Florida saw in the 2000 election.

Nikki Trella, the board's election reform director, said the successful "parallel testing" demonstration proved that the voting machines are reliable.

Parallel testing is a sort of mock vote where a machine and a group of people count the same stack of sample ballots, and then the results are compared for accuracy.

The test was performed using a voting machine to be stationed at Bates Middle School in Annapolis. The machine was randomly chosen from a warehouse by Rose Brooks, legislative assistant to Anne Arundel County Council Chairman C. Edward Middlebrooks, a Republican representing part of Glen Burnie, near Baltimore.


The voting machine included all the final software needed for Election Day, according to Trella.

In the test, election workers passed out 50 mock ballots, which were filled out and returned. Then two people read off the votes as two more tallied them by hand and another put them into the voting machine. Afterward, the hand-count total was compared with a printout of the machine total and they matched.

Despite final success, the test did point out a few flaws ... but in the people, not the machine.

A quick tally was done after every 10 votes to see if both human vote recorders had matching figures. There was a miscount in four of the five checks. However, recounts solved the problem quickly each time.

"When you put humans in (to the vote-counting process), that's what you can get," Trella said of the mistakes.

Trella said parallel testing will also be performed on four secured machines at the State Board of Elections on Election Day.

She said this is to assuage worries that a computer virus could be hiding in the machines waiting to activate until Election Day.

Robert Ferraro, co-founder of the touch-screen voting machine opposition group TrueVoteMD (search), attended the demonstration. He said he wonders what happens if those Election Day tests show a problem.

"My question is if you do a parallel test on Election Day and there is a mistake, then what?" he said.

Linda Lamone, Maryland elections' chief, said she has complete faith in the voting machines' accuracy, but if there was an error then her staff would audit the machines to pinpoint the problem.

Ferraro said he is still not convinced the machines are secure.

"The testing is good, but it doesn't guarantee accuracy," he said. "The best test would be a paper audit trail on Election Day."

The lack of any paper trail to confirm votes has been one of TrueVoteMD's major complaints about the machines. In September, the Maryland Court of Appeals rejected its argument, ruling the state did not have to provide paper ballots or a paper record for the machines.

The demonstration came a day before another meeting of TrueVoteMD and the State Board of Elections, this time in U.S. District Court in Baltimore. A judge is deciding whether TrueVoteMD poll watchers may break the normal 100-foot limit that partisan activists must maintain from polling places.

Ferraro said the state's position is an infringement of the First Amendment (search) guarantee of free speech, and the 14th Amendment guarantee of equal protection.

"We're not even going to be as intrusive as exit-pollers," he said, citing a group allowed within the 100-foot limit.

Lamone backs the decision, saying TrueVoteMD poll watchers could interfere with voters.

"It's intended to create a bubble of safety - a haven to protect voters from harassment from anybody," she said of the limit. "I don't want those people in there harassing the voters of Maryland."
OldDude is offline  
Old 10-21-04, 07:27 PM
  #2  
DVD Talk Hero
 
Join Date: Sep 2003
Location: Portland, OR
Posts: 25,062
Hey, Fox News missed the point again!

These voting machines are unverifiable. There is no way to know if the votes put into them are the votes that are recorded without a secondary paper record. A small test under controlled circumstances doesn't prove anything.

She said this is to assuage worries that a computer virus could be hiding in the machines waiting to activate until Election Day.
Any intelligent person looking to change votes wouldn't introduce such a virus until shortly before Election Day to avoid "tests" such as this.
Tracer Bullet is offline  
Old 10-21-04, 07:37 PM
  #3  
DVD Talk Hero
Thread Starter
 
Join Date: Jul 2001
Location: MI
Posts: 25,054
Typical of short human attention spans, you appear to have not read to the end of the article (which is why humans can't count votes for shit)
Trella said parallel testing will also be performed on four secured machines at the State Board of Elections on Election Day.

She said this is to assuage worries that a computer virus could be hiding in the machines waiting to activate until Election Day.
As I have pointed out before, the old-fashioned lever voting machines did not have paper trails either. While they were a PITA to mechanically program for each election, they had an excellent reputation for accuracy.
(I grant software leaves more opportunity for "unexpected results" and requires more careful checking)
OldDude is offline  
Old 10-21-04, 07:46 PM
  #4  
DVD Talk Hero
 
Join Date: Sep 2003
Location: Portland, OR
Posts: 25,062
Originally posted by OldDude
As I have pointed out before, the old-fashioned lever voting machines did not have paper trails either. While they were a PITA to mechanically program for each election, they had an excellent reputation for accuracy.
(I grant software leaves more opportunity for "unexpected results" and requires more careful checking)
That's true, but replacing one insecure system with another doesn't seem like progress to me.
Tracer Bullet is offline  
Old 10-21-04, 07:52 PM
  #5  
DVD Talk God
 
kvrdave's Avatar
 
Join Date: Aug 1999
Location: Pacific NW
Posts: 86,201
Originally posted by TracerBullet
That's true, but replacing one insecure system with another doesn't seem like progress to me.
A perfect system doesn't exist. This is better than the previous system. That seems like a good thing to me.
kvrdave is offline  
Old 10-21-04, 07:58 PM
  #6  
DVD Talk Hero
 
Join Date: Sep 2003
Location: Portland, OR
Posts: 25,062
Originally posted by kvrdave
A perfect system doesn't exist. This is better than the previous system. That seems like a good thing to me.
Electronic voting systems are much easier to tamper with than lever systems, though. And it wouldn't have taken much effort to design these machines to print a paper duplicate of each vote, show it to the voter, and store it.
Tracer Bullet is offline  
Old 10-21-04, 08:20 PM
  #7  
DVD Talk Hall of Fame
 
Join Date: Apr 2002
Location: Philadelphia
Posts: 8,572
I went to an "E-Voting" lecture today since I am a computer science undergrad. The woman speaking has a CS Ph.D. and is currently working at Harvard. She has become something of an authority on electronic voting after writing about the challenges in her dissertation a couple of days prior to the 2000 election.

Basically, according to her its impossible or near impossible currently to implement reliable e-voting. This is because you have to implement a system for each state and then each different municipality (if need be) since voting laws are different in each state. Furthermore, the voting program would need to be able to keep track of votes while not keeping track of who voted, since voting is anonymous. This presents a huge problem in auditing and ensuring a human voted and not a script.

The solution to this she said was more transparency from the e-voting machine makers who are currently being secretive and not complying to several government data security and integrity standards. This standards are only mandatory for DoD systems, but have been implemented on their own in the health care industry to acheive HIPPA compliance. She feels this standards along with a more open view to the public would ensure confidence. Also, if the public could help to improve some problems would be alleivated. She noted things such as Diebold using a z80 microprocessor which does not hold any definition of what is the data segment and what is the code segment, essentially allowing mallicious code to be executed off the tape storing votes under the guise of being "data".

Finally, we have to compete with sociological issues such as people still selling votes and even fraud by the people running the polling place. Something that technology alone cannot fix.

Fascinating stuff .
Brain Stew is offline  
Old 10-21-04, 09:29 PM
  #8  
Admin-Thanos
 
VinVega's Avatar
 
Join Date: Nov 2000
Location: Caught between the moon and NYC
Posts: 31,584
Johns Hopkins University professor Aviel Rubin, who last year published a study of portions of the Diebold software code, says the quality of that code was below minimum standards for a production system. Rubin's report cites a lack of industry-standard change-control processes and documentation, as well as specific technical weaknesses.

Jonathan Gossels, founder of SystemExperts Corp. in Sudbury, Mass., says his review of the Diebold code showed that it was "amateurish" in its design. More important, the amount of code that has been studied and found wanting "is only the tip of the iceberg" of the millions of lines of C++ and Microsoft Windows-based code that powers the Diebold touch-screen systems and back-end management servers, says Gossels.
Computerworld full story
VinVega is offline  
Old 10-21-04, 09:40 PM
  #9  
DVD Talk God
 
kvrdave's Avatar
 
Join Date: Aug 1999
Location: Pacific NW
Posts: 86,201
Originally posted by TracerBullet
Electronic voting systems are much easier to tamper with than lever systems, though. And it wouldn't have taken much effort to design these machines to print a paper duplicate of each vote, show it to the voter, and store it.
If a lever system did not make a backup, I don't know why it would seem necessary to have one in a computer system, so long as the system let you know that it had recorded your vote.
kvrdave is offline  
Old 10-21-04, 09:47 PM
  #10  
DVD Talk Hall of Fame
 
Join Date: Apr 2002
Location: Philadelphia
Posts: 8,572
I put up a long and well thought out post. Someone please validate me .
Brain Stew is offline  
Old 10-21-04, 09:56 PM
  #11  
DVD Talk Legend
 
Join Date: Jul 2001
Posts: 11,747
Originally posted by Brain Stew
I put up a long and well thought out post. Someone please validate me .
Sorry, we only validate with a purchase.
dork is offline  
Old 10-21-04, 10:41 PM
  #12  
Admin-Thanos
 
VinVega's Avatar
 
Join Date: Nov 2000
Location: Caught between the moon and NYC
Posts: 31,584
Originally posted by kvrdave
If a lever system did not make a backup, I don't know why it would seem necessary to have one in a computer system, so long as the system let you know that it had recorded your vote.
The lever system was all paper ballots printer by the lever machines. Paper trail is the key. You know how many people signed their names to vote. You should have the same number of paper ballots. It's not that complicated. Personally, I feel that it's a lot harder to mess with paper ballots, than it is to hack a system and screw it up. You don't have to even be physically there when your code executes itself, whereas you have to physically mess with the paper and get past all the scrutiny of the election watchers.
VinVega is offline  
Old 10-21-04, 10:41 PM
  #13  
DVD Talk Hero
 
Join Date: Sep 2003
Location: Portland, OR
Posts: 25,062
Originally posted by kvrdave
If a lever system did not make a backup, I don't know why it would seem necessary to have one in a computer system, so long as the system let you know that it had recorded your vote.
Two reasons:

1) It's much harder for someone to tamper with a lever machine, as they would have to physically monkey around with the equipment. With a computerized voting process, someone would simply have to surrepticiously introduce code to tamper with the results.

2) We have the capability for a computerized voting machine to print out a paper record and store it. The fact that lever machines don't isn't really a reason against the introduction of such a practice.
Tracer Bullet is offline  
Old 10-21-04, 11:06 PM
  #14  
DVD Talk Hero
Thread Starter
 
Join Date: Jul 2001
Location: MI
Posts: 25,054
Originally posted by VinVega
The lever system was all paper ballots printer by the lever machines. Paper trail is the key. You know how many people signed their names to vote. You should have the same number of paper ballots.
The lever system had no paper ballot. The little levers controlled cams and the big lever that opened the curtains plus the cams caused the counts to increment on selected mechanical counters. The machine was locked while polls were open so observers could not see the counts advancing, as it would show how people were voting. When the polls closed, the machine was opened and the votes were tallied from the counters. The total votes should have been equal to the number of people who signed voter slips.

So no one understands lever machines which were used for decades but understand touch screen well enough to demonize it.
OldDude is offline  
Old 10-21-04, 11:23 PM
  #15  
Admin-Thanos
 
VinVega's Avatar
 
Join Date: Nov 2000
Location: Caught between the moon and NYC
Posts: 31,584
Why was there a guy inserting a card before I voted and then removing it after I voted then on the lever machines?
VinVega is offline  
Old 10-21-04, 11:29 PM
  #16  
DVD Talk Hero
Thread Starter
 
Join Date: Jul 2001
Location: MI
Posts: 25,054
I don't know. When we had lever machines (we phased them out a few years ago) no one did this. We use optical scan now, and the ballot is counted immediately. If it is screwed up, it is rejected and you can get a new one. That seems to be the best of the available systems.

If optical scan ballots are simply stored for later count, that system is worse than punch card, yet it has the precious paper trail. The whole process has to be viewed as a system, and one ballot type can go from best to worst depending on other elements in the system.

Maybe immediately scanned optical ballot should be the standard. The problem is states, and I guess counties have the authority to decide what to buy. Lets be sure and take it away from them and decide in Washington.
OldDude is offline  
Old 10-22-04, 01:56 AM
  #17  
Banned
 
Join Date: May 2004
Location: Under the Stars
Posts: 1,330
Originally posted by OldDude
Lets be sure and take it away from them and decide in Washington.
Good lord. Does the spin never end!?

National elections are critical to our democracy. There should be a national standard. That may not mean directing every state to use the exact same methods but there should be some baseline requirements that every state has to meet. I would think that both parties would support fair and accurate voting. Hardly seems that way sometimes though.
Captain Pike is offline  
Old 10-22-04, 09:58 AM
  #18  
DVD Talk Limited Edition
 
Nazgul's Avatar
 
Join Date: Jan 2001
Location: Jayhawk Central, Kansas
Posts: 7,125
Originally posted by Captain Pike
Good lord. Does the spin never end!?

National elections are critical to our democracy. There should be a national standard. That may not mean directing every state to use the exact same methods but there should be some baseline requirements that every state has to meet. I would think that both parties would support fair and accurate voting. Hardly seems that way sometimes though.
Should we take away State/Local power and federalize the election ballots/process?
Nazgul is offline  
Old 10-22-04, 10:12 AM
  #19  
DVD Talk Hero
 
Join Date: Nov 1999
Posts: 36,981
so people want a computer that you vote at and it prints out a ballot and you check it and then put it in a locked box?


what's the point of the computer? is it just to make sure there are no "chads" or whatever? why not do opitcal scanner where you check them right there like OldDude said? That is cheaper to do.
Venusian is offline  
Old 10-22-04, 10:13 AM
  #20  
DVD Talk Hero
 
Join Date: Nov 1999
Posts: 36,981
Originally posted by VinVega
Why was there a guy inserting a card before I voted and then removing it after I voted then on the lever machines?
are you sure that wasn't a punch card ballot?
Venusian is offline  

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Thread Tools
Search this Thread

Archive Advertising Cookie Policy Privacy Statement Terms of Service

Copyright 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.