Re: Fw: Use of audit/copies in event of loss of original ballots Riverside County

From: Alan Dechert <alan_at_openvotingconsortium_dot_org>
Date: Mon Jul 05 2004 - 21:34:35 CDT

This tentative order carries some very sloppy reasoning and omissions. The
description of touchscreen is dumb. For example, footnote 1 (pg 2) says,
"In touchscreen (DRE) systems...." Touchscreen has nothing to do with DRE.
DRE means the vote is directly recorded electronically. This is independent
of the actual input method. Some DREs use a dial. Some DREs use a keypad.
Then is says "The voter then touches a yellow 'Cast Vote' cue...." I don't
think "yellow" has anything to do with DRE definition.

DRE is a pretty simple concept. It's amazing that lawyers and lawmakers
have such trouble with this.

Footnote 4 (pg 9) has this:

> The Court is not satisfied that there is any scientifically
> tested or peer-reviewed research to support the notion
> that most undervotes are unintentional or due to system
> error. Comparing only residual vote rates is not a
> scientifically valid method of comparing voting system
> performance.
>
I totally agree with this. However, it's ironic that the court is saying
this in rebutting Henry Brady's declaration. As far as I can tell, the
CalTech/MIT project invented the notion of "residual vote" as a "A Yardstick
for Reliability."

http://www.vote.caltech.edu/Reports/report1.pdf

See page 7 of this report for "Residual Votes: A Yardstick for Reliability."
I tend to agree with Henry Brady (and Caltech/MIT) insofar as "residual
votes" can be used in measuring voting system performance. But I think we
need to look at other factors that could cause undervotes (residual votes
include overvotes but these are relatively rare compared to undervotes). We
should not assume that all, or even most, undervotes are unintended due to
voting system pecularities. The Court is claiming that that's what Brady
(and others) are saying. I think it's enough to say that "residual votes"
could be a yardstick (not the only yardstick) while a significant percentage
(not necessarily "most") undervotes are due to the voting machine. I think
a lot more study is needed to establish the relationship with any degree of
accuracy. It's wrong to say the residual vote rate has nothing to do with
the voting machine, but it's also wrong to say that using only residual vote
rates will give you an accurate measure of overall voting machine
performance.

So, in my opinion, The Court is beating a straw man. They are faking an
excuse to disregard Brady's declaration.

CalTech/MIT invented the residual vote measure because, imo, they were a
little bit lazy digging into the data and trying to figure out what it says.
Brady adopted the terminology, in part, in his 2001 study, Counting All The
Votes (http://ucdata.berkeley.edu/new_web/countingallthevotes.pdf ), because
he intended to show that CalTech/MIT was mistaken in saying DREs were giving
voters more problems than other systems (Sequoia put up some of the money
for the study, but that did not influence Brady's thinking--he was already
suspicious of CalTech/MIT's findings). Roy Saltman concurred with Brady on
this point in the letter he wrote 5 Dec 2001.
http://www.openvotingconsortium.org/ContentFrame/Docs/RoySaltman/LetterToUCB/SaltmanToUCB.asp

The Court is exhibiting poor comprehension of who said what and why.

> Michelle Townsend (sp?) has quit... further Riverside (in name and
> through her deposition) was a part of the Benavidez v. Shelley
> litigation... that was recently denied a TRO or a PI (lawgeek-speak
> for a temporary restraining order and preliminary injunction
> respectively). Here's the TRO order hot off the presses:
>
> http://www.eff.org/e-vote/benavidez_v_shelley_tro.pdf
>
> Joe
>
> PS: Note that the court chose to ignore the Brady deposition in it's
entirety.
>
Noted.

Alan D.

==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Sat Jul 31 23:17:14 2004

This archive was generated by hypermail 2.1.8 : Sat Jul 31 2004 - 23:17:15 CDT