From: charlie strauss <cems_at_earthlink_dot_net>
Date: Fri Nov 12 2004 - 14:30:12 CST

okay alan... here's some text to chew on. You may not like this since its not sensational.

Close elections, unexpected voting patterns, armies of statisticians...Who can really tell the difference between and error and an unexpected outcome. "Dewey beats Truman" will always be with us. The crux of the issue is that there will be alays be close elections and we have to be able to tell the difference between unexpected voting patterns and actual errors. We cant simply make error-free machines--those will never exist. But we can make a process that can withstand errors and work well. How? you need two ingedients; first you must be able to spot errors, and second you need to be able to correct them. The final requirment is the most important of all: it must be transparent to every voter how it does this.

 Unlike other schemes using complex data handling and voter receipts the OVC system meets all three of these needs. A voter verified paper trail provides and independent audit trail that allows electronic errors to be both spotted and corrected. It combines the security advantages of paper and electronic systems while retaining their individual advantages. And it's all transparent to the voter and the election officials how it works.

At present we really have no checks on either ordinary paper ballots or electroinc voting machines. With paper we at least can recount if something goes wrong--as happened in Daytona FL this election. OVC provides two audit paths, one electorinc and the other on paper, with each having independent failure modes.

The press has been too quick to sieze the false story that the great fears of electronic election catastrophe never materialized. Yet not a single one of the election vulnerabilites has been addressed. This is not unlike the critisism recently leveled at NASA over the Columbia disaster. With each successive successful launch, known threats were further ignored even though their chance of happening was unchanged. The situation is somewhat worse with pure electronic vote recording. When the Columbia exploded we knew it happened, but we dont actually know if electroinc voting is working well since there's no way to check.

and so on...

that bit about bush winning by large margins as the lead just stinks to me....

-----Original Message-----
From: Alan Dechert <>
Sent: Nov 12, 2004 1:11 PM

> Alan I hate it.
Thanks, Charlie. Specific wording would be appreciated.

Alan D.

> The message for me this election was
> 1) it was close
> 2) there were known errors
> 3) there were certainly unknown too
> 4) But we cant tell the difference between an error and a surprising
> 5) there are always going to be errors, and close outcomes
> 6) the goal shoul dnot be to make an error-proof machine, but rather a
robust system
> 7) this requires making errors detacable
> 8) the requires making error differentiable for odd statistics
> 9) this require being able to correct errors when they occur
> 10) voter confidence requires tranaparentcy in all of the above.
> Frankly it makes no difference to me that there were errors in this
election or if we averted a disaster just with a large margin. The looming
potential of fraud and errors is problem enough.
> I also like Avi Ruben's take that the press set up this hypothetical
train wreck by hyping the potential for errors as a high likelihood of a
disaster. Now the story is that this was another millenium bug overhyped by
the verified voting henny pennies. I think you should address this by
stressing the potential for problems remains unchanged. Microsoft has had
secuirty holes lurking in designs that only come to light years later.
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Tue Nov 30 23:17:29 2004

This archive was generated by hypermail 2.1.8 : Tue Nov 30 2004 - 23:17:44 CST