<br><br><div class="gmail_quote">2011/7/13 <span dir="ltr"><<a href="mailto:fsimmons@pcc.edu">fsimmons@pcc.edu</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<br>
<br>
----- Original Message -----<br>
From: Kristofer Munsterhjelm<br>
> <a href="mailto:fsimmons@pcc.edu">fsimmons@pcc.edu</a> wrote:<br>
...<br>
<div class="im">> There may also be another scenario where Majority Judgement (or<br>
> median<br>
> ratings, for that matter) would do better than ranked methods.<br>
> If it's<br>
> possible for the voters to agree on what, say, "Good" means<br>
> (comparability of utilities), then MJ might extract usable<br>
> cardinal<br>
> information from the voters, while the strategy resistance makes<br>
> the<br>
> cardinal information much less prone to the sort of Approval-<br>
> reduction<br>
> that you would see in Range. If one holds certain assumptions<br>
> that make<br>
> cardinal methods useful at all, then MJ could well be strategy<br>
> resistant<br>
> enough that it would do better than Range*.<br>
><br>
> B&L spends quite a bit of their paper on the claim that the<br>
> voters *do*<br>
> agree on what the different categories mean, and so that there<br>
> is<br>
> comparability so that the cardinal information can be used.<br>
<br>
</div>Instead of asking voters for "utility" values, ask them to rate the candidates on a scale of zero to 100%,<br>
where rating candidate X at 37% means that you think that 37% of the time candidate X would vote the<br>
same way that you would vote if you were there representing yourself.<br>
</blockquote></div><br><div>I'd certainly be able to understand and use that system "honestly" if I wanted, and for well-informed voters it seems reasonable. But I doubt that your average "independent" voter, who looks on voting as a chore, would like such a system. </div>
<div><br></div><div>I think B+L's method of using a commonly-understood grading method of a society would be better for most people. In France, that means 0-20 or "reject" to "excellent"; in the US, that would be the letter grades A-F (without the "E" unless you went to Harvard) or their commonly-understood corresponding numbers (90-100 = A, 80-89=B, etc).</div>
<div><br></div><div>(If you use letters, the MJ tiebreaker system assigns a + or - by itself, although A+ and F- would be impossible. In AT-TV, this +/- would correspond to order of election at a given rank.)</div><div><br>
</div><div><meta http-equiv="content-type" content="text/html; charset=utf-8">0-5 stars is also commonly used on the web, but it doesn't have any anchoring meaning like the letter grades.</div><div><br></div><div>JQ</div>
<div><br></div><div>ps. Looking at current US politicians, I can well imagine that my "honest" letter grades for both frontrunners would frequently be below the winning median, which I imagine would be a C+ or B- in most cases, although B+L's poll during the presidential primaries found that Obama would have actually made it to the equivalent of B+. </div>
<div><br></div><div>In some cases, I would "strategically" inflate my preferred viable candidate to a B. But on certain issues, I would be happy for the opportunity to say to my preferred viable candidate: if you betray my ideals on this important issue, I will in good conscience simply vote honestly. I will still prefer you to your principal opponent, but this preference will do you no good, as it will still be pulling down your median score.</div>