[EM] replies to Ossipoff re Range Voting; explanation of latest RV results

[Abd ul-Rahman Lomax]

At 12:41 PM 2/10/2007, Warren Smith wrote:
> >WDS: In IEVS, presently, equal rankings are forbidden in rank-order methods.
> >MO: which (like Warren's other assumptions) makes the results meaningless.
>
>--WDS: While I agree it would be nice if IEVS did equal rankings, 
>and I plan to make
>a future version do that,
>(a) I do not agree I ever made any "assumption" here.
>I simply described the status of IEVS.  I did not "make an assumption."
>(b) I do not agree every result in the universe that concerns rank 
>order voting methods
>is "meaningless."

I have a policy of not replying directly to Ossipoff, there is a 
history of endless debate that turns over details of "you said," and 
"I said," endless argument that goes nowhere.

Here Warren noted in his post that IEVS did not presently allow equal 
rankings. He was listing this as a shortcoming of IEVS. Ossipoff 
apparently turned this into an assumption that there was something 
defective about equal rankings. Warren is correct. He simply 
described the status of IEVS, which has not yet been programmed to 
allow equal rankings in ranked methods.

The charge that his results are therefore "meaningless" is, well, 
silly. Many implementations of ranked methods don't allow equal 
ranking, in the real world. Yes, as Ossipoff points out, most of us 
would prefer equal ranking (which actually turns ranked methods into 
something closer to Range, or at least to Approval). But programming 
equal ranking is trickier, if you are using issue space analysis to 
determine votes. At what level of preference do you decide to rank 
equally? Or what other factors influence the use of equal ranking? It 
is actually a *lot* more complex.

And, yes, it is necessary for Warren's results to have wider 
application. But they are not at all "meaningless" as they stand.