[EM] Last call for edits to consensus statement (Jameson Quinn)

[Stéphane Rouillon]

  The need for a computerized counting system depends not only on the 
maximum number
of allowed ranks but on the number of candidates too. That does not 
imply the need
of a centralized to do all the job. Local precincts can do some part of 
the job before the
central gathers all the information to declare winners.

Especially in the case of range (median) voting, even without 
restricting the number of grades or values,
the central job is a very easy task if you know what you are doing. I 
could do the centralized job alone within
3 hours for some US states if precincts do the job I ask.

On 2011-09-11 16:00, Kathy Dopp wrote:
>> From: Jameson Quinn<jameson.quinn at gmail.com>
>>
>> New proposed section:
>>
>> Vote-counting details
>> Most of our supported methods will require no updates to voting equipment,
>> and all can be counted at the precinct level.
> False claim. All of them require at least software updates and fully
> ranking more than 3 candidates would also require hardware updates.
> Range voting *may* require hardware updates too.
>
> Always best to stay away from making too-strong claims if one wants to
> be believed.
>
>> The supported methods will require little or no changes to voting equipment.
>> Approval voting, majority judgment, and in some cases range voting can all
>> be counted on any existing equipment, by treating each candidate as a
>> separate ?election?.
> What do you mean by "treating each candidate as a
>   separate ?election?"?
>
>   Be more specific.
>
>> For instance, for range voting, the equipment could count how many people
>> gave each rating to candidate A, from a simple array of choices such as 0,
>> 1, 50, 98, 99, or 100. Most choices are bunched near the ends of the scale,
> OK. I see *if* range voting is restricted to filling out bubble
> choices of a restricted number of ratings.  Is this truly range
> voting? Or more like rating on surveys where 5 choices are available.
> This might still require new ballot scanning systems than those in use
> today.  Where are you obtaining your information on voting systems?
> > From a computer scientist who is a voting system expert? Directly from
> the voting vendors?
>
>
>
>> as this helps range voters get near to the maximum power from their ballot.
>> Once the number of voters choosing each rating is known, simple arithmetic
>> gives the overall score for the candidate.
>>
>> Counting our supported methods will be as trustworthy as counting plurality.
>> Computers will help give results faster, but all results could be checked by
>> hand. Publicly-available tables of results could be detailed enough to allow
>> such checking, without compromising the secrecy of individual ballots or
>> taking more than a few pages or kilobytes of data. Precinct-level counting
>> and reporting helps allow for good security measures such as paper ballots
>> and sampled auditing. All the methods will result in a low rate of spoiled
>> or invalid ballots, perhaps lower than plurality.
> I believe these new additions above are way too detailed for this type
> of statement, and thus prone to error or disagreement.  You are also
> going to turn off the large contingent of election advocates who want
> to be able to hand count all ballots publicly if you focus on the use
> of computers in counting alternative voting methods.
>
> I believe that the number of claims for additional benefits of
> alternative methods should be toned down too.  What scientific
> evidence do you have for lowering the rate of spoiled ballots with
> range voting for instance? Probably very little.
>
> Maybe it's time to stop working on the statement, except for removing
> a few things, such as the discussions of SODA and IRV and any other
> unnecessary details that people may not agree upon.
>