[EM] An IRVing response

[Richard Moore]

Douglas Greene wrote:
> Message: 2
>   Date: Fri, 03 Aug 2001 04:03:58 -0500
>   From: Randy Kunkee Subject: Re: The Problem(s) with 
> ...
> The claim that IRV votes cannot be summed is flat out
> wrong.

Technically, this critic is correct, if one accepts a weaker 
definition of summability that does not include the 
constraint that the information can be summed *in a compact 
matrix form*.

However, there are two disadvantages to IRV's weaker form of 
summability (in addition to the more complicated vote 
counting procedures) that are seldom discussed (though they 
have been mentioned on this list).

1. Reporting of results is complicated. Many people would 
like to see the election totals published in the newspaper. 
But you can't just publish the pairwise or preference 
matrices and expect readers to be able to see how the 
results follow from that. You have to publish the number of 
ABCD votes, the number of ACBD votes, the number of ADBC 
votes, etc. With more than a few candidates the list gets 
quite long.

2. As Forest Simmons pointed out on this list a few months 
back, it is possible for a candidate to win every precinct 
in IRV and still lose the election.

We much prefer a method that meets the strong definition of 
summability.

Richard