[Election-Methods] RE : Is "sincere" voting in Range suboptimal?

[Abd ul-Rahman Lomax]

Well, now I see what happened. I did *not* 
ridicule Venzke. Read it carefully. I said that 
Venzke had confirmed the obvious. Since the 
obvious sometimes leads to error, it is an 
appearance only, confirmation is quite useful. I 
was just pointing out that it *was* obvious. I 
apologize for the ambiguity in my message that 
allowed a tone of ridicule to appear.

Confirming the obvious is a valuable activity, it 
keeps us from being stuck in what may even be ancient errors.

Further, while we do have Warren's simulations, 
and some theoretical work as well, many don't 
trust those, for reasons which must be 
frustrating to Warren, since they boil down to 
"You've done it wrong," or "You are biased," but 
don't point out exact errors, and, in some cases, 
the alleged "errors" are only simpifying 
assumptions that Warren is explicit about. For 
example, it's possible that the utility 
distributions Warren uses don't match reality. 
However, it is far from clear that this has any 
significant effect on what the simulations show.

It's true, Condorcet methods work better in real 
elections, we can expect, than we would expect 
from, say, random assignments of utilities, 
making cycles rarer than they would otherwise be. 
However, I think the distributions Warren uses, 
which are *not* even random distributions, if I'm 
correct, are likely to be reasonable, with more 
accurate distributions not significantly shifting 
the results. How voter utilities will be 
distributed, in real elections, will be a constantly shifting thing.

No more new text below.

At 09:51 AM 7/24/2007, Kevin Venzke wrote:
>Hi,
>
>--- Abd ul-Rahman Lomax <abd at lomaxdesign.com> a écrit :
> > >  When I produced simulation results to
> > >show that in the zero-info case one improves one's expectation under
> > >Range by voting at the extremes, you essentially ridiculed me for
> > bothering
> > >to show something so obvious.
> >
> > I'd be interested to see how I made such an
> > error. Perhaps I'd been influenced by the barrage
> > of experts claiming this.... and I'd like to see those simulations again.
>
>In November you wrote:
>
> >> This study seems to
> >> replicate the common criticism of Range Voting, which is that voters
> >> may attempt to maximize their personal expected benefit by voting
> >> strategically, in this case by voting Approval style, with a
> >> particular strategy:
> >>
> >> >These results suggest to me that in the zero-info case, if there are
> >> >thought to be more than about 2 other voters, Range should be voted
> >> >as in Approval.
> >>
> >> This is utterly unsurprising. Venzke has simply confirmed the
> >> obvious.
>
>Warren implemented his own version afterwards; I suggest his results if
>you're really interested. http://rangevoting.org/RVstrat3.html
>
>Kevin Venzke