>> From: Richard Moore <[EMAIL PROTECTED]> >> Subject: Re: [EM] Some brief campaign argument >> Anthony Simmons wrote: >> > I wonder if other methods have this same characteristic even >> > if they don't seem to. IRV, for example, allows you to >> > specify a complete ranking, but during the actual counting, >> > at any given time, you are either voting for a candidate or >> > you aren't. While this looks like a ranking on paper, it >> > seems to function more like Sequential Approval. Much like >> > Approval, except that instead of approving the lot all at >> > once, you approve and disapprove in turns. >> I don't buy the comparison. A more apt description would >> be Sequential Plurality. Good point. Though it seems the distinction between Sequential Approval and Sequential Plurality is rather vague. After all, if Plurality is modified so that you're permitted to vote for more than one candidate, then you have Approval. So, if you're allowed to vote for more than one, but not simultaneously, which is it? Fortunately, I paid no attention to that, since it was the "Sequential" part that I was talking about. >> Approval works the way it does because all expressed >> preferences are counted in parallel. Break that >> parallelism up and you've fundamentally altered the >> method. And on the other hand, Plurality works the way it does because you only get to vote for one person. Alter that ... I mean, let's face it, IRV is not Plurality and it's not Approval. But intuitively, IRV does sound more like Plurality. I suppose that's why it's popular. >> Richard
