<div>Reweighted Range Voting <a href="http://rangevoting.org/RRV.html">http://rangevoting.org/RRV.html</a> does not check every possible combination of candidates. However there may be a way to determine the optimal candidate quickly.</div>
<div> </div>
<div>Set X and set Y are adjacent if it is possible to create one group by changing a single candidate in the other. …in other words, all the members are identical but one.</div>
<div> </div>
<div>Set X is a local maximum if the utility of every adjacent set is less than Set X’s utility.</div>
<div> </div>
<div>The utility function is rather simple.</div>
<div> </div>
<blockquote style="MARGIN-RIGHT: 0px" dir="ltr">
<div>for each voter, the utility is ln(1+score_sum/max)</div></blockquote>
<blockquote style="MARGIN-RIGHT: 0px" dir="ltr">
<div>with score_sum being the score they gave each candidate individually and max being the maximum rating allowable for a single candidate.</div>
<div> </div>
<div>This is taken from the D'hondt divisors 1+1/2+1/3..., but integrated rather than summed.</div>
<div> </div>
<div>presumably ln(1+2*score_sum/max) would work as well. </div></blockquote>
<div> </div>
<div>From a few tests I’ve run, it seems as if there’s never more than one local maximum. Naturally, this single local maximum would be the optimal candidate set.</div>
<div> </div>
<div>This suggests that a simple iterative procedure will determine the optimal candidate set without examining all of them. (Perhaps using Reweighted Range Voting or Naive Multiwinner Range as a starting point)</div>
<div> </div>
<div>I, however, lack the expertise to prove whether it is possible for multiple local maxima to occur. I was wondering if anyone could.</div>
<div> </div>
<div>This method is called Proportional Range Voting due to its resemblance to Proportional Approval Voting <a href="http://www.knowledgerush.com/kr/encyclopedia/Proportional_approval_voting/">http://www.knowledgerush.com/kr/encyclopedia/Proportional_approval_voting/</a></div>
<div> </div>
<div> </div>