[EM] Thoughts on Burial

[Juho]

On Jul 23, 2010, at 4:47 PM, Jameson Quinn wrote:

> Do you assume (in the scenario where strategic cycles are common)  
> that the voters would be coordinated (by parties or by media) or if  
> the voters would make the strategic decisions themselves (maybe  
> based on general poll and theoretical strategy related information  
> from media but not voting directly as recommended by some central  
> entity)?
>
> I believe that a truncation-type strategy is simple enough that it  
> will occur to voters on their own. Both before the vote, and in the  
> context of post-facto analysis of "why did my candidate lose", when  
> people are very prone to find relatively sophisticated  
> rationalizations. In the latter context, while it would obviously  
> not affect the outcome of the given election, it would be toxic to  
> the system and encourage division and silly strategy in later  
> elections. A good system, one in which its defenders can make the  
> strongest possible statement that "no, that's impossible", is  
> therefore desirable.

In this sense plurality and TTR are good systems since they don't  
leave too much detailed information around for the media and voters to  
wonder and speculate afterwards. In Condorcet methods one could or  
could not have the actual votes available for investigation (in  
addition to the matrix). The possibility of strategic manipulation (if  
one group of voters could change their voting behaviour while other  
voters would stay as they are) in the last election may not yet make  
strategic voting rational in the next election. This irritation (when  
voters notice the potential strategies afterwards) may lead to more  
strategic votes in the next election but since working implementations  
of the strategies may not be easy to find, because different voters  
may apply different strategies and because it is not easy to separate  
strategic votes from the sincere ones when analysing the results of  
the election afterwards, it is also possible that people start once  
again trusting the system and consider strategic votes just as random  
noise or as potentially counterproductive activity. One can thus not  
say "no, that's impossible" but maybe one can say "you are free to  
try" :-).

> (BTW, I think that the media is pretty hopeless in terms of helping  
> voters understand strategy; if we can't agree on things on this list  
> of knowledgeable people, how can you expect a clear, unanimous  
> message to emerge from the soup of interest and stupidity that is  
> the punditocracy? Even if the good analysis is a plurality at 40% of  
> the media, while 60% is three-different-kinds-of-wrong, then a naive  
> voter gets nothing much from that.)

Yes, this speaks in favour of Condorcet methods in which strategies  
are typically very complex to identify and apply.

> Do you assume that the strategies are rational or irrational?
>
> There will certainly be both kinds, but any rational ones will have  
> a long-term competitive advantage and come to be most common.

If there are rational implementable strategies.

Juho