[EM] strategy and method complexity and the advantage of minmax methods

[Juho Laatu]

On Jun 9, 2005, at 07:45, Abd ulRahman Lomax wrote:

> Which is better, a winner supported by half the voters plus one, or a 
> winner with the largest approval rating?

> What is needed, I'd say, is consensus, not regarding the "best 
> method," but simply upon the characteristics and likely -- or 
> preferably demonstrated -- outcomes of the application of each method. 
> Choosing election methods then becomes a deliberative process based on 
> solid information and broadly accepted opinion.

I agree that there may be different best methods for different 
purposes. If one first picks the targets for the election, e.g. to 
elect the most approvable candidate, then one should in principle be 
able to pick the most accurate voting method. In addition to that one 
may have different needs with respect to defense against strategic 
voting. Also strategic voting risks are different in different 
elections (e.g. large public elections, repetitive small elections on 
same candidates where some set of members of this mailing list vote). 
Also elections with no intention to vote strategically are possible. 
"Best" should thus be seen in a framework set by the utility, strategic 
voting elimination and other requirements.

> One of the fundamental problems with developing consensus in a mailing 
> list with no adjunct polling process that includes all interested (or 
> however else membership is defined) is that the appearance of 
> consensus in that environment can be quite false.

> Enter our superhero, Delegable Proxy....)

I agree with your analysis, and yes, proxy like or other methods 
discussed on this list might (in principle at least) be a tool to 
derive some more representative results from the discussions that often 
seem to drift in random directions and lead to random ballot style 
conclusions. I guess this mailing list is the correct place to start 
the new era and revolution and pick new rules for extracting better 
than ever seen results from this type of free discussion fora. Maybe 
you already have a proxy based complete proposal on the new rules of 
behaviour in your back pocket? I note that Jobst Heitzig was quick and 
already created one such new tool 
(http://wiki.electorama.com/wiki/Essential_Questions). I can't yet tell 
if proxies or other methods are the best way forward. But I'm sure that 
finding new efficient discussion tools could be a step forward for the 
mankind (small or big, I don't know yet :-).

Best Regards,
Juho