[EM] General PR question (from Andy Jennings in 2011)

[Richard Fobes]

On 10/3/2014 4:26 AM, Toby Pereira wrote:
> Richard, there may have been come confusion because my examples have
> been assuming an approval ballot, so A, C and C, A are the same.
> Obviously if these are ranked ballots this changes things.

Thank you for clarifying.

Can we avoid this ambiguity for the future?

One possibility is to use an "&" (ampersand) between two choices that 
are ranked at the same preference level.  For example, approval ballots 
would look like this:

10 voters: A & B
10 voters: A & C

When ranked ballots are used, there can be a mixture of same-level and 
different-level preferences:

10 voters: A, C & B
10 voters: C & A, B

Another possibility is to omit any symbol between same-level choices:

10 voters: A  B
10 voters: A  C

or:

10 voters: A, C B
10 voters: C A, B

Obviously I'm used to a comma indicating a change in preference level. 
Am I the only person who interprets ballots this way?

BTW, in my opinion a voting method that does not allow a voter to rank 
two choices at the same preference level is not a good voting method.

 > ...
 > I also tend to think that the argument for score or approval ballots
 > gets stronger for proportional representation than for single-winners.

I suppose that by limiting the preferences that voters can express, an 
approval approach might appear to achieve better proportionality.

Yet I think that if full voter-preference information is suppressed, we 
are doing what single-mark ballots now do: hide the unfairness by not 
collecting all the relevant information.

Richard Fobes