[Election-Methods] Another Lottery Method for the Record

[Jonathan Lundell]

On Jun 30, 2008, at 12:58 PM, Fred Gohlke wrote:

> I've thought about your proposal since I first read it the other  
> day. I'm unclear about the human dynamics.  The proposal, as later  
> adjusted, seems to have a negative tinge; it relies on eliminating  
> people judged undesirable rather than elevating those judged  
> desirable.
>
> Am I wrong in imagining that accentuating the negative assumes those  
> not eliminated must be desirable?
>
> Presumably, after the undesirables are eliminated, those randomly  
> selected to make succeeding decisions will base them on the greater  
> or lesser amounts of desirability of those remaining until, in the  
> final stages, only the most desirable remain.
>
> Do I have the fundamental idea right?  Does accentuating the  
> negative provide an advantage?

Something like this, it seems to me, depends for its desirability on  
the agreed intent of the group. This mechanism is good for finding a  
centrist, compromise candidate. But one could also argue that in the  
4D-2R example, in some circumstances the D's are entitled to choose  
their own idea of a best candidate.

Take the US Senate, for example. Should the Senate be forced to elect  
Joe Lieberman (say) as the Senate leader because he's in some sense in  
the middle of the spectrum? Bad example, maybe, because he might be  
knocked out earlier, but the point remains. Similarly in a  
parliamentary system--there are good reasons for the PM to be the  
majority party leader, and not the most bland compromise candidate in  
the group.

Nonetheless, I can see the advantage of this method in a highly  
factionalized group that is nonetheless motivated to choose a mutually  
acceptable leader and do business.