# [EM] RE : Naive question about Range Voting -- why 0-99 and not 0-100?

Abd ul-Rahman Lomax abd at lomaxdesign.com
Mon Nov 20 19:27:08 PST 2006

At 03:44 PM 11/20/2006, Monkey Puzzle wrote:
>As he states in the paper, most people seem to understand 0-100 just
>fine, but a few do not -- they think that all the scores MUST sum to
>100.

The only reason I can see for 0-99 vs 0-100 is that it adds another
digit. This increases ballot complexity by 50%, quite a bit for just
one point more in granularity.

Personally, I think 0-99 is way overkill. I would instead shoot for
0-10. People understand it quite well, and human judgement of value
is simply not reducible to exact numbers, 0-10 is, I think, quite
adequate. I think of a 9 as being only slightly less than 10.

I've proposed that overvoting be allowed in Range scores. That is, if
a ballot is marked with more than one score for a candidate, the
resulting score is the average of what is marked.

This allows, actually, 0-10 to easily express increments of 0.5.
Can't decide between 9 and 10? Mark both.

This however, would require more sophistication in vote counting, it
does not have the advantage of standard no-overvoting Range of being
able to use existing equipment, more or less.

sometimes with, say, lever equipment. It's going to be an argument
used against Range. 0-10 adds one more mark to 10, i.e., it is eleven
positions. Not bad for the intuitive gain.)

To me, the *real* plan is to push Approval Voting, which is Range 2.
(Some argue that Range will, in practice, reduce to Approval anyway,
with people voting the extremes; while certainly some people will do
this, I don't think this is an argument against Range in itself.)

One there is Approval -- which is really very simple to implement,
the ballots don't change except for a few words of instruction --
then people will want to express more sophisticated votes. I've
argued that the first improvement over standard Approval is a
Favorite option. It would not be used, necessarily, in determining
the winner, but would allow better analysis of results, plus it could
be used for campaign finance distributions when they are based on
vote counts. (Otherwise Approval votes for a major party candidate
may be unreasonably reduced because the voter doesn't want to donate
to that party. Plus determining how to distribute funds, without a
Favorite, is in itself a thorny problem.

How's this for a slogan: