[EM] Meet Warren Smith

Tue Jan 16 00:58:48 PST 2007

I'm posting some comments from Warren Smith, and my replies. I emphasize 
that Warren's angry tone was there from the start, entirely unprovoked. I 
had never criticized Warren.

This discussion is on-topic, because it's about the electoral system, even 
if it reveals someone's anger. How we discuss, and what motivates us, are 
relevant on this list. I believe that it's a good thing to find out whom 
you've been correspoinding with.

Warren said:

Re your formula involving a^a and b^b, let me just simply say:
I don't believe you.  Specifically, whatever integral it is
you claim to have done, I claim you did it wrong and got a totally
bogus, albeit cute looking, formula out.

Re "cycles", I did not disparage that idea, in fact I thought it was
interesting, and credited you with it.  (Basically this entire thing
was me trying to make sense of what you'd been saying, failing, and
then
redoing it myself trying to preserve the good or interesting ideas
you had, but getting rid of the wrong execution of them, unclear stuff, 
etc.)

I reply:

But Warren never showed the "failing" or the "wrong execution, etc.

Warren continues:

Specifically, you executed wrong in the following ways:
1) you did your integral or integrals wrong, getting bogus formulas.

I reply:

Warren seems to believe that sufficiently frequent repetition is a 
substitute for substantiation of his claims.

Warren continues:

2) your probabilistic model was inappropriate (uniform distribution);

Bias-Free, even with its uniform distribution assumption, is less biased 
than Webster.

Warren continues:

exponential distribution is appropriate if any simple one is.
This is well known to mathematicians for various reasons probably
unfamiliar to you, but I'll probably try to write them down later.
(Dirichlet...  Poisson...  Binomial Combinatorics...  to mouth some
buzzwords in a manner which is probably meaningless, but it is
too much of a pain for me to explain it right, right  now, and I'm
not sure even if I did so, that you would appreciate it...)

I reply:

...or maybe yoou're just parroting some terminology from a class in which 
you got a "D" or worse, or from other source, on topics about which you're 
as clueless as you are about the derivation of Bias-Free.

In any case, it's better to not make so many unsupported statements.

Your misunderstandings and confusions about the other things you've 
discussed make it unlikely that you were doing other than parroting and 
posturing, in the above quote.

Warren continues:

Incidentally, computer simulation by a modification of Bishop's sim program
http://rangevoting.org/BishopSim.html
indicates my new method is indeed an improvement over Webster and
all the classical divisor methods.   A computer simulation of your
method would indicate it was a total failure, because your formulas
are bogus.

I reply:

Bias-Free has been applied to censuses, and, in 2 apportionments, it agreed 
with Webster, then wilth Hill. On the bias signed-number-line, Bias-Free is 
between Hill and Webster. In the long run it will be less biased than 
eilther.

Mike Ossipoff

_________________________________________________________________
Your Hotmail address already works to sign into Windows Live Messenger! Get 
it now 
http://clk.atdmt.com/MSN/go/msnnkwme0020000001msn/direct/01/?href=http://get.live.com/messenger/overview