[EM] Kevin, 0330 GMG, 20 March, '05

[MIKE OSSIPOFF]

Kevin--

You quoted me:


>What you need, then, is something quite different from WDSC. You need to 
>define the No Silliness Criterion. Go for it. In the meantime, though, I'll 
>settle for methods that don't require those voters to reverse a preference.

You say:

I don't need to define a criterion to explain why I don't like another
criterion.

I reply:

Don't be so defensive.

I didn't say that you need to define a criterion to explain why you don't 
like another criterion. I said, however, that you need to define a criterion 
that protects us from methods that could require voting in a silly way. I 
said that because you keep lamenting that I haven't written that criterion, 
that my critera aren't that criterion that you want. Therefore, since you 
want such a criterion so badly, you need to write one.

As I said, though, that would require you to tell us a definition of voting 
in a silly way. So far you haven't done that. Kevin, you're not making much 
progress with your new criterion if you don't know how you want to define 
silly voting. Voting alternately ">" and "=" in your ranking? Just guessing.

You continued:

I'm surprised you don't know that I like SDSC. Only I use Steve Eppley's
votes-only wording under the name "Minimal Defense" since he has offered
such a wording.

I reply:

I don't criticize Steve's criteria. Of course they're different from mine. 
They aren't just re-statements of my criteria, equivalent but differing only 
in wording. They're different criteria. I've consistently said that I make 
no claim that my criteria are the only ones that could be written to measure 
for the goal of minimizing defensive strategy need.

You continued:

Why do I want a votes-only wording of this? Because it's very easy to use.
Votes-only Minimal Defense says certain candidates can't win given some
situation. I can program this into a computer. But I can't easily program a
computer to search for how a majority "needs to vote."

I reply:

I have no idea what is easier for you to program, but whether or not they're 
what you most like to program, the majority defensive strategy criteria are 
unambiguous in all situations about whether a method passes them or fails 
them, and they apply uniformly to all methods.

Neither you, Markus, nor Paul has posted an example in which one of the 
majority defensive strategy criteria SFC, GSFC, WDSC, or SDSC, is ambiguous 
about whether a method passes or fails.

A need for certain wording improvements in FBC as pointed out, and the 
improvements were made. Such suggestions are always welcome.

For instance, in recent postings I've shown that Margins fails WDSC, 
Approval passes WDSC, and that BeatpathWinner passes WDSC and SFC.

More such demonstrations will be posted.


You said:

"The majority might need to vote in morse code" is not a very meaningful
guarantee.

I reply:

It certainly is not a meaningful guarantee. In fact it isn't a guarantee at 
all. It's an expression of the non-offering of a guarantee that you want, 
and that must be very important to you.

Youir point, presumably, is that WDSC doesn't guarantee that a majority who 
prefer X to Y won't need to vote in Morse code in order to ensure that Y 
won't win.

No, but why should it? WDSC guarantees that those voters won't need to 
reverse a preference.

Perhaps you'd like my criteria to require that a method guarantee everything 
that you'd like in a method, but that's not how criteria work: Different 
criteria require different guarantees. That's why we have various criteria, 
instead of just one.

It works like this: If there's something that seems important to me, that a 
method should guarantee, then I write a criterion requiring that guarantee. 
Then, if there's something different that you want a method to guarantee, 
you (but not I) write a criterion to require that guarantee.

Millions of voters reverse their preferences in our elections, out of 
perceived strategic need, thereby concealing what they want. I suggest that 
that isn't good for democracy. You must not think that's important, because 
you say that it isn't useful to  state conditions under which it can be 
guaranteed that it won't happen. To each their own; There's no reason why 
what I consider important should be important to you. But you must 
understand that the reverse is true also.

As I said, it's unreasonable to expect one criterion to guarantee 
everything. That's why we have more than one criterion.

To you it would be terrible if some as-yet unknown method required that 
majority to alternate ">"  and "=", or to spell out "WDSC" in Morse code 
with ">" and "=".  Maybe. I can't say that that is something that has 
worried anyone but you. I'm not denying that that is important to you. Your 
feelings about what is very important to you are valid, and they are _not_ 
wrong.

Maybe you, or some other relentless crusader-for-justice will write a 
criterion that requries that methods protect us against those serious 
dangers. Maybe you, or that other crusader, will even find a method that 
fails that criterion, or explain to us what you mean by a "silly way of 
voting".

You say that you don't have anything that you want me to do, and yet you're 
complaining because of something that I haven't done: I haven't written a 
criterion that says that a majority who prefer X to Y should have a way of 
voting that will ensure that Y won't win, without alternately voting ">" and 
""=" or voting in Morse code.

And, it's at that point in this cyclic repetition that I again suggest that 
if you want such a criterion, then you need to write it yourself.

Mike Ossipoff

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/