[EM] The Range Voting example

[MIKE OSSIPOFF]

Jim--

Thanks for the interesting and problematic example.

The real problem there is the co-operation/defection problem that has been 
discussed on EM a few times. It's usually discussed in terms of Approval, 
though the problem can  happen in Condorcet as an order-reversal problem (if 
the ATLO option isn't available).

That's the first time I've heard it as an RV scenario.

Of course, though exact voting predictions won't be available, it's quite 
possible that people will know that A and B are the frontrunners.

You wrote:

The real strategic idiots are the voters who like A and B best. If they
just kept their high rankings of each other, A or B would still win.
Only, how high should they rank their second choice? Too much, and their
second choice wins over their first choice, and they need the other guys
to rate their second choice higher or their last choice (G) wins. This
is starting to sound very much like the awful game that gets played with
some strategic Borda voting.

I reply:

Sure, the A and B voters have to choose whether to help eachother or compete 
with eachother, and if one competes and one helps, then that faction is 
being had. If they both compete, they both lose.

That's the co-operation/defection problem, as it occurs in voting systems.

I don't think that there's any solution to that problem. It just has to be 
accepted as something that one may have to deal with in Approval, RV, or 
maybe even Condorcet (if the ATLO option isn't available).

I don't consider it to be a serious problem. If the A voters co-operate and 
the B voters defect and take advantage of the A voters, the result isn't a 
majority rule violation.

And if there's a significant difference in A and B, then it will be clear 
that one of them is more of a compromise than the other is, and that one 
will have the advantage in the game of chicken, as the rightful compromise. 
And that more middle candidate's supporters will have less to lose if the 
one of the other, less desirable, candidates wins.

Or maybe one of {A,B} looks like a better votegetter than the other. If A 
looks stronger, then it would be more natural and expected for the B voters 
to give 10 points to A.

The fact that A & B depend, for their lead, on both getting votes from both 
candidate's supporters is what makes the problem. But if they share so many 
votes, how different can they be? And therefore how much difference can it 
make which one wins?

If I were an A voter, I'd give 10 points to both (or vote for them both in 
Approval). If the B voters really want to win so badly, then let them have 
their win, if A and B are as similar as is implied by all the shared votes.

Other solutions:

One faction, the one that has the advantage of somewhat greater size, or 
more middle-ness as a compromise, or a sounder moral/ethical position, could 
publicly announce that of course everyone knows that that 
co-operation/defection problem exists, and that they therefore are 
announcing that they will not vote for the rival.

Or one or both factions could announce publicly that if the other faction 
defects, then that will be the last time they give a vote to that faction. 
That would be a stong discouragement to defection.

Aside from all that, as I said, if they're as similar as the shared votes 
suggest, I"d just vote for both.

Mike Ossipoff

_________________________________________________________________
On the road to retirement? Check out MSN Life Events for advice on how to 
get there! http://lifeevents.msn.com/category.aspx?cid=Retirement