[EM] Re: IRV-P. Another name for Condorcet?

[Brian Olson]

I also like descriptive naming.

For what we know as "Condorcet's Method", how about "Virtual Round 
Robin"? I've used the round-robin tournament system in describing it to 
people who hadn't heard of it before and are new to election methods. 
"round robin" implies the all-vs-all nature of the system. I guess the 
tricky point is noting how the "virtual round robin tournament" is 
within a single voter's ballot, and then all those "tournaments" are 
added up on top of each other.

Hmm, change the order of the description and "virtual round robin" 
still works. Instead of "A beats B within voter X's ballot", "i voters 
prefer A to B while j voters prefer B to A. If i is greater than j then 
A beats B". Does that sound ok? Maybe by thinking procedurally based on 
how this method is actually implemented in the computer I dug myself a 
hole I didn't need to.

Also, "IRV-P" annoys me. But maybe that's just because IRV annoys me. 
:-P

Brian Olson
http://bolson.org/

On Sep 16, 2004, at 3:26 PM, Jobst Heitzig wrote:

> (ii) such names don't tell anything about the thing.

> So I appreciate this discussion which will perhaps lead us to some more
> descriptive terminology. Perhaps "IRV-P" is not the worst proposal.
> After all, a traditional runoff usually is a pairwise comparison, so in
> that respect the term fits the pairwise methods quite well. However, a
> runoff also is something applied *after* some first tally, whereas most
> pairwise methods consider all pairs at the same time at the start, sort
> them in any way and only then apply some iterative procedure. So in 
> that
> respect, the term "runoff" seems misleading. Perhaps one could replace
> "instant" by "simultaneous" or "concurrent", or some synonym of these
> which begins with an "I" :-) One could also use "Independent Runoffs
> Voting -- Pairs" in order to indicate that the runoffs are not in
> sequence but at the same time between all pairs...