[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: More on HB1296



Margit,

 

DRE vendors can refurbish their vote recording machines by adding a scanner and a printer, and by removing the vote recording and vote counting functions. 

 

The User Interface to the voter would be mostly transferable, and instead of recording the votes in memory, digitally, the machine would print the votes on the paper ballot.

 

I have not seen the Paul and Scott plan for sampling.

 

It is my hope that a transparent counting system will be adopted that will make it possible, and relatively easy, to reveal all errors.  Verifiability of each step in the counting process is the only way I have found to do this. 

 

Even in hand counting, there are multiple steps to determine the total:

 

Step 0 is the partitioning of the task into repeatable batches of uniquely identified ballots.

Step 1 is the interpretation of each vote on a ballot. 

Step 2 is deciding how to count an ambiguous vote – by the duplication board.

Step 3 is the recording of each vote as a hash mark on an accumulation page – by batch.  

Step 4 is the summarization of the hash marks by batch.

Step 5 is the summarization of the summaries for the election.

 

An independent team can repeat all or a portion of the steps for all or a portion of the batches and arrive at a count for a ballot, a batch and/or the election.  This might be done as a sampling device to monitor the original and each recount, and of course it would be done to perform a recount. 

 

A third team can compare the results of two independent counts and declare that the ballot or the batch or the election counts match or do not.

 

Al

 

 

 

 


From: Margitjo@xxxxxxx [mailto:Margitjo@xxxxxxx]
Sent: Wednesday, February 11, 2004 6:40 PM
To: cvv-discuss@xxxxxxxxxxxxxxxxx
Subject: More on HB1296

 

Al, et al,   :)
       Re HB1296, I feel uncomfortable abandoning the counties that now have  DRE's, which were, at last count, Arapahoe, Denver and El Paso, with a total of 1,150 precincts.  Are there, or will there be in the near future, no systems that would produce a verifiable receipt from a DRE that gave a completely legible record  of how someone voted, on paper that made counting relatively easy?
       Isn't one of the major problems with any system that counts votes electronically how to determine whether there might be "errors" in the machine counts so we can do a paper recount?  Paul and Scott have suggested some statistically-valid sampling of paper, which if it revealed discrepancies, would trigger a full paper recount.  Shouldn't we encourage Alice Madden to incorporate this step?
Margit