Signal vs Noise in sci.stat.math

I've been busy today, getting some errands done before I go to the
Caribbean
to show some clueless NAG that I'll be the Captain of a Princess
cruiseship
again, with unlimited free use of the highspeed internet on the ship.
;->

Both NAG and Russell Mars had been making NOTHING but NOISE --
both of them. (see my follow-up to Russell to come, and the fact that
he had killfiled most of mine and Afonso's posts).

This group has been griping about the NOISE when the loudest gripers
are the ones who visits sci.stat.math the LEAST and know the LEAST
about what I post and what Luis Afonso post. Most of these gripers
had
their heads in the sand ... and all they know is to gripe BLINDLY, and
complain about the low signal to noise.

How the hell would you know what signals there are and what NOISE
there are without READING at least a small SAMPLE of it?

And if you killfiled indiscriminantly both Afonso and Reef Fish, what
sample of the population do you have?

Here are the TOP 10 posters this month:

176 Reef Fish
90 lic...@xxxxxxxxxxx Luis A. Afonso.
-----------------------------------------------------
Reef Fish's posts are all SIGNAL -- whether you recognize it or not.
Afonsos are 100% NOISE. And if you don't know much about
Statistics, you wouldn't understand the Statistical SUBSTANCE in
Reef Fish's posts.
-----------------------------------------------------

The ostriches with their heads in the sand are missing 266 of the
posts this month, while the rest total about 133, exactly HALF of
the Top 2. Let me give you a hint which are NOISE and which
are SIGNAL of the next Top 8:

31 jtom...@xxxxxxxxxxxxx Tom Tomsky: ALL SIGNAL
20 doe_s...@xxxxxxxxxxx David Winsemius: 80% NOISE
19 Nag Naga1010 100%
NOISE
18 Ethan.Johnsons@xxxxxxxxx New in Sept insufficient data
appears to be
mostly SIGNAL
17 bob.oh...@xxxxxxxxxxxxxxxxx Anon Bob O'Hara: ALL NOISE
One of Top 3 worst in statistical substance in
sci.stat.math
10 jkawc...@xxxxxxxx Janusz Kawczak : 90% NOISE
9 sjhersc...@xxxxxxxxxxxx Stephen Herschkorn: 95% NOISE
9 coreywh...@xxxxxxxxx CoreyWhite: vacuous in Statistics
from alt.magick, misc.writing,
rec.arts.sf.writing
rec.arts.sf.written, rec.puzzles,
sci.physics,
alt.conspiracy ... were his LATEST 8 posts.

So, for the 8 above, the SIGNAL/NOISE ratio would be approx.

(31 + 4 + 1 + 1)/(16 +19 + 17 + 9 + 9 ) = 37/ 70

still far above the estimate by the CLUELESS of 5% signal/noise.
Jack Tomsky alone account for nearly half of the signal, when you
remove Reef Fish and Afonso from the population count!


Robert Dodier, who hardly EVER contributed anything to sci.stat.math
griped about the noise and the low signal to noise ratio.

So, to make his presence known, he posted the most INAPPROPRIATE
and ignorant post in "HELP: Non-linear regression problem" for TWO
reasons:

1. I had help Paul THREE times explaining something very simple,
about his questions on p-values. On the FOURTH time, I simply
had to tell Paul perhaps someone else could help, because most
people who asked his questions should have understood my
response the FIRST time.

2. So, Robert Dodier stepped in with this:
"YOu can ignore the p-values, ... they are a distraction
anyway"...
and start telling Paul how he should be a Bayesian ... including
"you might consider that several leading Bayesians are/were
physicists,. .."

That did it! I challenged Dodier

RF> So that's who Robert Dodier is, the self-proclaimed Physicist
RF> Bayesian.

RF> Have you EVER assessed a bivariate prior distribution
RF> that reflected your own prior belief. If so, just ONCE in
RF> your life even, tell us HOW you did it, and I'll tell you
RF> either how you were callous or how you are NOT a
RF> Bayesian.

RF> This is my challenge to see if you known ANYTHING
RF> about Bayesian statistics.

RF> This is 100% serious and absolutely substantive in
RF> Bayesian statistics.

RF> Here is your chance to show if you are just blowing smoke
RF> or if you actually know something

This was nearly 24 hours ago. Silence from this NOISE maker
Robert Dodier. My challenge stands. He can take ALL the
time he wants. He showed every sign that it's all HOT AIR.
And he tried to palm himself off as a "Physicist Bayesian".

This is the kind of QUACKS that routinely get unchallenged in
sci.stat.math before I started in 2005.


Then there is john2 in the thread "variation & mean level" giving
Ethan the most uneducated, nonsensical, and GARBAGE answer,
after I had already answered Nathan that his question was one
of "heteroscedasticity" vs "homoscedasticity".

John2's response (which is so OBVIOUSLY wrong) stood there
for over 13 hours now ... with NO ONE even noticed his
malpractice! Just watch my follow-up to his when I get around
to it.


Folks, if you want to be critical about the way I treat Quacks, and
especially the obnoxious Quacks and abrasive Quacks, such as
Richard Ulrich, Luis Afonso, Anon Bob O'Hara, and the rest of
them down the line, first try to understand how THEY provoked
my "not-so-polite" responses, and then YOU have to know
something about Statistics!

Most of the critics of me are BLIND critics -- not having seen much
of what I posted in their SUBSTANCE, and when they did, they
didn't recognize the Quackery of Richard Ulrich, and sided with
Richard and flamed ME. That's exactly the problem with the
sci.stat.math newsgroup.

I don't mind being criticized for my statistical substance EVER, and
I don't even mind to be criticized about my "tone" of being abrupt
(as spyrian did), but at least syprian has read SOME of my posts
and he recognized that I am NOT the kind of poster most other
blind critics said I am.

In less than a year of posting in sci.stat.math, I had already topped
the ALL TIME list (the posting ID died quite sometime ago, from
being sabbotaged by posters from a different group). I am fairly
sure I've posted at least TWICE the all time Top number.

Call that a rough 3,000 posts in sci.stat.math. Go to the advanced
search in Google and search for ANY keyword within the subject of
Statistics, you'll probably find SOME contribution from me on the
subject (you can discount all my responses to Afonso -- even
THERE, quite a bit of statistical SUBSTANCE were contributed,
such as in Hypothesis Testing, p-value, probability, and other
100% topics of statistical substance.)

THEN, you can take issue with ANY of my statistical contribution,
if you know how. You are even welcome to tell me that I treated
anyone too harshly, IF you read enough of what the OTHER person
had said or done, to have been the quid pro quo recipients of my
wrath.

Without doing your homework, and without knowledge, any critic
is just as ignorant as those who I call Quacks and malpractice
snake-oil salesmen in this group. The characterization, in every
case, was based on, in my judgment, supporting FACTS and
evidence beyond a reasonable doubt!

-- Reef Fish Bob.

.

Relevant Pages
Loading