Ping cc: some questions you never answered

[Snit]

You have not answered these questions:

1) What is an inflection point on a curve?

2) How does the distance from the mean to the inflection
points on a bell curve relate to a standard deviation...
and how does this mesh with your claim that the distance
from the mean is irrelevant to the standard deviation?

3) You repeatedly claim I used both Excel and Numbers
incorrectly to come up with the linear trend lines I
showed you. What do you think is the right procedure and
what do you think I did wrong? Be roughly as specific in
your response as I was in showing you what I did to get
the trend line in Excel.

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Keep in mind: I have already given you the answer to each of those
questions.




--
🙈🙉🙊

[Torre Starnes]

Snit....Please stop....
You are reminding me of high school statistics.
A course I hated.

[Snit]

On 5/30/12 2:53 PM, in article 1p0sww76lg46m$.1nn5e8iu45g12$.d...@40tude.net,

LOL! Really, the answers are not that hard.

1) An inflection point is a point when the curve changes
concavity (going from concave to convex or vice versa).
Technically this is where it is exactly perpendicular
(tangent) to the X axis (if that is complex language,
where the curve is perfectly vertical)

A good explanation of this can be seen here:
<http://goo.gl/7iOHF> - and on a bell curve here:
<http://goo.gl/BA6WP>.

2) For a normal distribution (a bell curve), each inflection
point is one standard deviation away from the mean. That
is what the standard deviation is - that distance.

If you know this, you can easily see that some depictions
of a standard deviation are wrong. I showed several
examples, the most extreme of which is at the top of this
set of images: <http://tmp.gallopinginsanity.com/sd.png>.
I also show the approximate inflection points.... easy to
see visually (at least where they roughly are).

3) There is *nothing* I did wrong in coming up with the
linear trend lines. I showed cc info directly from MS and
Apple which described the process - and in my video I show
it is really very easy to do. And both Excel and Numbers
get the graphs correct (as shown in my above links from the post you
responded to):



And that is the end of your math lesson for the day. :)

If you want, next time we can look at some of cc's comments which clearly
are contradictory to the above claims... he was completely clueless and is
now, for some reason, running from the topic as fast as he can. :)



--
🙈🙉🙊

[Steve Carroll]

On May 30, 6:17 pm, Snit <use...@gallopinginsanity.com> wrote:

(snip)

> "Torre Starnes" <torre.star...@gmail.org> wrote:
> > On Wed, 30 May 2012 13:19:59 -0700, Snit wrote:

(snip)

> > Snit....Please stop....
> > You are reminding me of high school statistics.
> > A course I hated.

(snip grandstanding, red herrings and other unrelated horsesh*t)

>     3) There is *nothing* I did wrong in coming up with the
>        linear trend lines.

Other than to have written the following initial statement (that lead
up to the creation of your trend line fiasco in another thread) you
mean.

"The correlation fits with my prediction. I know, I know, the word
"correlation" has more than five letters and will confuse you, but go
beg someone else to help you understand what it means." - Snit

https://groups.google.com/group/comp.os.linux.advocacy/msg/e9e30f0d7bf7b139

And then this thread, that ties your correlation/prediction statement
above together with a "trend line" and does so "very, very" clearly
as you explain what the "context" is and how you expressed to keep it
"clear":

"When I looked at your data I found, of course, that there is still an
upward trend line.

Just as I predicted there should be." - Snit

https://groups.google.com/group/comp.os.linux.advocacy/msg/9f888a7a06d0a234

So, let's do as you suggest in the thread I just quoted, " let us keep
the context clear", most notably, *all* along the way you kept talking
about "correlation" and your "predictions" as you entered into this
"trend line" BS. Now let's get to the facts:

Not only have you failed to show your alleged "correlation" between an
"increase in usage" of Linux and the addressing of its "usability
issues", you've also failed to show how the trend lines you created
were predictive in *any* way (they clearly are not and, if you knew
what cc was talking about, you'd know why).

(cue up Snit arguing that his trend lines were not supplied in the
"context" (that he needed to make "clear") of showing his alleged
"correlation" and alleged "prediction", despite the fact that the
threads I just pointed to above **clearly** show otherwise). Oops!!

Snit, you *really* need to learn that people aren't as stupid as you
need them to be.

[Clogwog]

"Snit" <use...@gallopinginsanity.com> schreef in bericht
news:CBEBCEFF.2221%use...@gallopinginsanity.com...

> You have not answered these questions:

IMHO, he doesn't owe you an answer.

This discussion, a very boring one, is leading nowhere, please try to accept
that cc has his arguments to disagree with you.

[Snit]

On 5/31/12 11:50 AM, in article jq8ei7$1o9$1...@dont-email.me, "Clogwog"

<clo...@anon.eu> wrote:

> "Snit" <use...@gallopinginsanity.com> schreef in bericht
> news:CBEBCEFF.2221%use...@gallopinginsanity.com...
>> You have not answered these questions:
>
> IMHO, he doesn't owe you an answer.

Well, this is Usenet where none of us really are obligated to post at all...
but after all of his name calling and lying about me, I do think it would be
the right thing for him to do to acknowledge his errors - and the fact is he
is, or at least was, not only completely clueless about the following (that
is fine, we are all ignorant in different areas), he pretended to be
knowledgeable *and* based on his fantasy of that he sank to name calling and
lying about me.

If he was a moral person, which he is not, he would apologize for his
behavior or at least admit to his ignorance on the following and his
accusations against based on his ignorance. But he will not. He simply is
not a moral person.

>> 1) What is an inflection point on a curve?
>>
>> 2) How does the distance from the mean to the inflection
>> points on a bell curve relate to a standard deviation...
>> and how does this mesh with your claim that the distance
>> from the mean is irrelevant to the standard deviation?
>>
>> 3) You repeatedly claim I used both Excel and Numbers
>> incorrectly to come up with the linear trend lines I
>> showed you. What do you think is the right procedure and
>> what do you think I did wrong? Be roughly as specific in
>> your response as I was in showing you what I did to get
>> the trend line in Excel.
>>
>> <http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
>> <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>
>>
>> Keep in mind: I have already given you the answer to each of those
>> questions.
>>
> This discussion, a very boring one, is leading nowhere, please try to accept
> that cc has his arguments to disagree with you.

Of course: cc argues with me just to disagree - it is not as though he
really believes what he claims. He is a liar. And while I know it bores
others, I respond by shoving his nose in his own messes. And he will run...
he has no answer where he can save face because even he knows he made a
complete ass out of himself. Watch him respond with denials of that
*without* answering the above questions. Quite predictable.



--
🙈🙉🙊

[cc]

On Thursday, May 31, 2012 2:50:45 PM UTC-4, Clogwog wrote:
> "Snit" <use...@gallopinginsanity.com> schreef in bericht
> news:CBEBCEFF.2221%use...@gallopinginsanity.com...
> > You have not answered these questions:
>
> IMHO, he doesn't owe you an answer.

I already answered them.

> >
> > 1) What is an inflection point on a curve?

Where the second derivative changes sign. I already told you that.

> > 2) How does the distance from the mean to the inflection
> > points on a bell curve relate to a standard deviation...

The first standard deviation is always at the inflection point and vice versa.

> > and how does this mesh with your claim that the distance
> > from the mean is irrelevant to the standard deviation?

I pointed out that the distance to the first standard deviation/inflection points is not the same for every curve, which is true. You seemed to think that standard devation measured distance from the mean, not area under the curve which is why you were confused when different curves had the first standard deviation/inflection points at different distances, but always covering ~68% of the area.

> >
> > 3) You repeatedly claim I used both Excel and Numbers
> > incorrectly to come up with the linear trend lines I
> > showed you. What do you think is the right procedure and
> > what do you think I did wrong? Be roughly as specific in
> > your response as I was in showing you what I did to get
> > the trend line in Excel.

I offered many times before to show you the right procedure, you just shrugged it off. Since you're going to continue to lie I'd prefer to just laugh at your ignorance. Here's a hint, your line was a poor predictor of the months it covered, not to mention the very next month! Now compare that to mine. I (and I believe others) have pointed out the term for how good of a fit a trend line is in other posts, which you conveniently ignored at the time. Go back, find that term, and then you'll be on the right track. Happy to help.


> > <http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
> > <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

I would love it if you kept posting those links though. I know I will.

> > Keep in mind: I have already given you the answer to each of those
> > questions.
> >
> This discussion, a very boring one, is leading nowhere, please try to accept
> that cc has his arguments to disagree with you.

I have more than arguments, I have mathematical fact.

--
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov> - Snit's ignorance of Excel and his hilarious attempt at statistical analysis

[cc]

On Thursday, May 31, 2012 3:16:46 PM UTC-4, cc wrote:
> On Thursday, May 31, 2012 2:50:45 PM UTC-4, Clogwog wrote:
> > "Snit" <use...@gallopinginsanity.com> schreef in bericht
> > news:CBEBCEFF.2221%use...@gallopinginsanity.com...
> > > You have not answered these questions:
> >
>
> > >

> > > 3) You repeatedly claim I used both Excel and Numbers
> > > incorrectly to come up with the linear trend lines I
> > > showed you. What do you think is the right procedure and
> > > what do you think I did wrong? Be roughly as specific in
> > > your response as I was in showing you what I did to get
> > > the trend line in Excel.
>
> I offered many times before to show you the right procedure, you just shrugged it off. Since you're going to continue to lie I'd prefer to just laugh at your ignorance. Here's a hint, your line was a poor predictor of the months it covered, not to mention the very next month! Now compare that to mine. I (and I believe others) have pointed out the term for how good of a fit a trend line is in other posts, which you conveniently ignored at the time. Go back, find that term, and then you'll be on the right track. Happy to help.

Sorry to reply again, but this question is laughable. I've already told you what you need to know, you've just done no leg work to connect the dots. Here are some questions you never answered:

1) Why did I take an average of the data?

2) Why did I get the standard deviation of the data?

3) Why did I set control limits?

4) How do I know my line is a better fit than yours?


If can answer these questions, then congratulations, you know some very basic statistical analysis (it can all be done in Excel for shit's sake).

[Snit]

On 5/31/12 12:16 PM, in article
b73e54ea-62ba-4529...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 2:50:45 PM UTC-4, Clogwog wrote:
>> "Snit" <use...@gallopinginsanity.com> schreef in bericht
>> news:CBEBCEFF.2221%use...@gallopinginsanity.com...
>>> You have not answered these questions:
>>
>> IMHO, he doesn't owe you an answer.
>
> I already answered them.

Well, but you showed no understanding.

>>> 1) What is an inflection point on a curve?
>
> Where the second derivative changes sign. I already told you that.

Looking back I do see where you said:
-----
I know exactly what an inflection point is. It's where the
second derivative changes sign, and it's exactly where the
sigma lines are in your supposed incorrect examples.
-----

But then you also denied the top example here:
<http://tmp.gallopinginsanity.com/sd.png> has the inflection points (the
standard deviation lines) drawn incorrectly. It clearly does. You had (and
maybe still have) no idea that the inflection point is where the curve's
concavity changes. *Clearly* in the first example this is wrong.

A good explanation of this can be seen here: <http://goo.gl/7iOHF> - and on
a bell curve here: <http://goo.gl/BA6WP>.

There is no doubt you at least were completely clueless about this, or you
would not have claimed:

cc:
-----
Yes, the first sigma lines are always at an inflection
points. And ALL THE EXAMPLES YOU LINKED TO HAD THE LINES
DRAWN AT INFLECTION POINTS.
-----

The emphasis, by the way, is yours. You were so amazingly sure of yourself.
And so amazingly and obviously wrong. But you will not admit to it, even
though the examples I showed you *were* clearly wrong... with the most
obvious example being the one shown at the top of this image:
<http://tmp.gallopinginsanity.com/sd.png>.

I mean, really, how hard is it to see that is very, very wrong. It is.
Your denials just prove your ignorance. But have fun denying that - it is
what you do!

>>> 2) How does the distance from the mean to the inflection
>>> points on a bell curve relate to a standard deviation...
>
> The first standard deviation is always at the inflection point and vice versa.

It is the distance from the mean to the inflection point... which is
completely contrary to your claims before:

cc:
-----
It's area under the curve, not distance from center you
idiot. Jesus Christ, give up now.
-----
It is not distance from the mean. Repeat: it is not
distance from the mean.
-----

Good to see you now know your previous claims were flat our wrong. Or will
you deny that, too... even though you have now admitted I was right to tell
you it is the distance from the mean to the inflection points (on a bell
curve, of course).

>>> and how does this mesh with your claim that the distance
>>> from the mean is irrelevant to the standard deviation?
>
> I pointed out that the distance to the first standard deviation/inflection
> points is not the same for every curve, which is true.

It is *always* the distance from the mean to the inflection point on a bell
curve. Always. Exactly the same. And remember, this is easy to see: it is
where the concavity changes (*always*!). And thus it is easy to see when a
depiction is significantly wrong (but, as I note above, you did not know
this... but you will deny it... as is your habit).

Of course, bell curves can be tall or short or wide or think or big or
small... depending on the data and how big the image is. This, apparently,
was some big epiphany for you that you felt the need to share, even though
it is very, very basic knowledge and there as no reason to think anyone else
would not have known that (I mean, really, for an adult to find this to be
even slightly noteworthy he or she would have to be quite daft).

> You seemed to think that standard devation measured distance from the mean,
> not area under the curve which is why you were confused when different curves
> had the first standard deviation/inflection points at different distances, but
> always covering ~68% of the area.

It is the distance from the mean to the inflection points on a bell curve.
The distance. Measured from the mean. To the inflection points on a bell
curve.

You claimed I was wrong to say it was the distance. You were the one who
was flat out wrong.

>>> 3) You repeatedly claim I used both Excel and Numbers
>>> incorrectly to come up with the linear trend lines I
>>> showed you. What do you think is the right procedure and
>>> what do you think I did wrong? Be roughly as specific in
>>> your response as I was in showing you what I did to get
>>> the trend line in Excel.
>
> I offered many times before to show you the right procedure, you just shrugged
> it off.

I already showed you the right procedure:

<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

It is not like there is a question to that being correct, in either Excel or
Numbers (and I show both).

You like to pretend otherwise because you know I am right. There is no "if"
here... what I show is exactly fine for drawing the linear trend lines.

And you will *always* run from this question... not even you believe your BS
claims about me somehow getting it wrong. You simply do not. If you did
you would be rubbing it in my face and having a party - just imagine, cc
*finally* getting to show where Snit was wrong! You would be jumping for
joy and posting it in your signature every day.

You can pretend otherwise, but we both know you would be lying. And, guess
what, you will lie... because you are a liar. You pretend to be
knowledgeable about things you are clearly very, very ignorant about.

> Since you're going to continue to lie I'd prefer to just laugh at your
> ignorance. Here's a hint, your line was a poor predictor of the months it
> covered, not to mention the very next month! Now compare that to mine. I (and
> I believe others) have pointed out the term for how good of a fit a trend line
> is in other posts, which you conveniently ignored at the time. Go back, find
> that term, and then you'll be on the right track. Happy to
> help.

You have *nothing* to add when it comes to what I might have done wrong when
drawing those linear trend lines. Nothing. Because, well, I was (and am)
right.

It really is that simple. Stop trying to make it seem more complex.

>>> <http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
>>> <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>
>
> I would love it if you kept posting those links though. I know I will.

Are you suggesting you *really* believe your own BS? LOL! That would be
funny if you did.

Keep in mind: the only claim of mine is that those are correctly drawn and
calculated linear trend lines. I know you have tried to push other topics
such as what type of trend line was selected and why, etc., but that is not
relevant to your claim that I did something wrong in creating those linear
trend lines.

Remember that... if you can!

>>> Keep in mind: I have already given you the answer to each of those
>>> questions.
>>>
>> This discussion, a very boring one, is leading nowhere, please try to accept
>> that cc has his arguments to disagree with you.
>
> I have more than arguments, I have mathematical fact.

LOL! You have your ignorance and your refusal to admit when you are wrong.
I mean, really, you have yet to even admit you were wrong with the quotes
above (and repeated here):

cc:
-----
Yes, the first sigma lines are always at an inflection
points. And ALL THE EXAMPLES YOU LINKED TO HAD THE LINES
DRAWN AT INFLECTION POINTS.
-----

cc:
-----
It's area under the curve, not distance from center you
idiot. Jesus Christ, give up now.
-----
It is not distance from the mean. Repeat: it is not
distance from the mean.
-----

Flat out wrong on your part. Not if. Not maybe. Not could be. You simply
were wrong. Completely and utterly wrong.

But you will *never* admit to it. Never. And if you do happen to prove me
wrong I will admit to it readily... it is what I do. You simply are not as
moral of a person as I am. But you are amusing!

--
🙈🙉🙊

[Snit]

On 5/31/12 12:35 PM, in article
86c78db3-6cc5-4cb5...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 3:16:46 PM UTC-4, cc wrote:
>> On Thursday, May 31, 2012 2:50:45 PM UTC-4, Clogwog wrote:
>>> "Snit" <use...@gallopinginsanity.com> schreef in bericht
>>> news:CBEBCEFF.2221%use...@gallopinginsanity.com...
>>>> You have not answered these questions:
>>>
>>
>>>>
>>>> 3) You repeatedly claim I used both Excel and Numbers
>>>> incorrectly to come up with the linear trend lines I
>>>> showed you. What do you think is the right procedure and
>>>> what do you think I did wrong? Be roughly as specific in
>>>> your response as I was in showing you what I did to get
>>>> the trend line in Excel.
>>
>> I offered many times before to show you the right procedure, you just
>> shrugged it off. Since you're going to continue to lie I'd prefer to just
>> laugh at your ignorance. Here's a hint, your line was a poor predictor of the
>> months it covered, not to mention the very next month! Now compare that to
>> mine. I (and I believe others) have pointed out the term for how good of a
>> fit a trend line is in other posts, which you conveniently ignored at the
>> time. Go back, find that term, and then you'll be on the right track. Happy
>> to help.
>
>
> Sorry to reply again, but this question is laughable. I've already told you
> what you need to know, you've just done no leg work to connect the dots. Here
> are some questions you never answered:
>
> 1) Why did I take an average of the data?

You did not know Excel calculated a trend line as easily as it does:
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Based on your ignorance you tried to calculate it yourself - and if you got
a difference answer than what Excel and Numbers did then you did something
wrong in your calculations (I do not care where you goofed and will not be
double-checking your work for you... assuming you even post it).

> 2) Why did I get the standard deviation of the data?

You did not know Excel calculated a trend line as easily as it does:
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Based on your ignorance you tried to calculate it yourself - and if you got
a difference answer than what Excel and Numbers did then you did something
wrong in your calculations (I do not care where you goofed and will not be
double-checking your work for you... assuming you even post it).

> 3) Why did I set control limits?

You did not know Excel calculated a trend line as easily as it does:
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Based on your ignorance you tried to calculate it yourself - and if you got
a difference answer than what Excel and Numbers did then you did something
wrong in your calculations (I do not care where you goofed and will not be
double-checking your work for you... assuming you even post it).

> 4) How do I know my line is a better fit than yours?

You do not know your line is "better" than mine, because mine is 100%
correct.

> If can answer these questions, then congratulations, you know some very basic
> statistical analysis (it can all be done in Excel for shit's sake).

Nobody is saying you cannot calculate an average or a standard deviation in
Excel. Wow. You really are lost.

But, hey, if you want to know how to calculate a linear trend line:
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

But you claim Excel and Numbers both get it wrong... and only you have been
able to figure it out! Oh, wait... you then claimed I made some mistake in
how I used the tools in those programs - but you could never figure out what
that mistake was!

LOL!

Seriously, cc, you are a complete idiot. I like posting to you because you
cannot help but be completely arrogant in your ignorance. It amuses me to
no end.



--
🙈🙉🙊

[cc]

On Thursday, May 31, 2012 3:47:49 PM UTC-4, Snit wrote:
> On 5/31/12 12:16 PM, in article
> b73e54ea-62ba-4529...@googlegroups.com, "cc"
> <scat...@hotmail.com> wrote:
>
> > I pointed out that the distance to the first standard deviation/inflection
> > points is not the same for every curve, which is true.
>
> It is *always* the distance from the mean to the inflection point on a bell
> curve. Always. Exactly the same.

You're getting confused. The first standard deviation point is always at the inflection point, like I said. But the distance from the mean to the first inflection point is not the same for every curve, like you seem to believe. Standard deviation covers area under the curve, not percentage distance from the mean.

[cc]

On Thursday, May 31, 2012 3:52:36 PM UTC-4, Snit wrote:
> On 5/31/12 12:35 PM, in article
> 86c78db3-6cc5-4cb5...@googlegroups.com, "cc"
> <scat...@hotmail.com> wrote:
>
> >
> > 4) How do I know my line is a better fit than yours?
>
> You do not know your line is "better" than mine

I do. It's called an R^2 value or coefficient of determination. Excel gives you the value.

This is getting quite funny.

[Snit]

On 5/31/12 1:27 PM, in article
0d1fed23-5252-41ae...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 3:47:49 PM UTC-4, Snit wrote:
>> On 5/31/12 12:16 PM, in article
>> b73e54ea-62ba-4529...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>> I pointed out that the distance to the first standard deviation/inflection
>>> points is not the same for every curve, which is true.
>>
>> It is *always* the distance from the mean to the inflection point on a bell
>> curve. Always. Exactly the same.
>
> You're getting confused.

Nope. I am *exactly* correct. And you were *exactly* incorrect when you
made the following claims:

cc:
-----
It's area under the curve, not distance from center you
idiot. Jesus Christ, give up now.
-----
It is not distance from the mean. Repeat: it is not
distance from the mean.
-----

But it *is* the distance from the center (the mean) to the inflection point.
There is no "confusion" about this... it *is* the way it *is*. No argument.
No maybe.

This is not a place where there is any chance of my being wrong or any
chance of you being right. This is a case of accepted *math*. You will not
prove math wrong (just as you will not prove Excel and Numbers wrong, nor
*ever* find a place where my linear trend line is wrong in the following
examples):

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Remember: there is no debate here - you, simply, are wrong. Period. The
game now is to see how extreme you will get in your denial - it is not as if
you are ever going to show a flaw with my work or with Excel or with Numbers
or with *math* itself. You will *never* do this.

But goodness, you will be amusing as you squirm and insist you are
oh-so-right when you are just blowing smoke.

> The first standard deviation point is always at the inflection point, like I
> said.

Well, as you said after I told you about it, sure. Here is the message
where I told you:
<http://groups.google.com/group/comp.os.linux.advocacy/msg/148c20587f1a5898>

------------------------------------------------------------
Snit:
-----
Ok, time to give you a hint: it *very much is* the "distance
from the center"... which, by the way, is generally referred
to as the "mean" (which is not the same way you are a mean
person, but the average). Again: There is a very specific
distance from the mean - one which is easy to see on a graph
- which you clearly are ignorant about. And that is a very
clear hint... maybe even you can figure out your error from
there (happy Googling!).
-----

cc:
-----
Holy Shit! You are completely wrong! Standard deviation for a
normal distribution is the percentage of the area UNDER THE
CURVE!
-----

------------------------------------------------------------

I even told you what type distance it was:

cc:

-----
It is not distance from the mean.

------

Snit:
-----
It is *also* that... a very specific distance. Again, not
in inches or millimeters or something like that, but based
on a property of the curve. Wow... that is a *huge* hint
for you. I mean it is almost screaming the "secret" to
you. Can you figure it out from that hint? I am being
kind and giving you more and more, even though you are
refusing to admit you do not know and you are even
fabricating stories about me.
-----

------------------------------------------------------------

And then I gave you the big hint... and apparently you took my advice and
looked up the concept of an inflection point:

Snit:
-----
Ok, one more *huge* hint for you. Huge. I mean, really,
this is giving it away... but I am taking pity on your
ignorance and your self-humiliation and feeling a bit bad
for you. Consider the concept of an "inflection point".
Look it up if you have to (as you almost surely will).
Yeah, then see if you can find how that concept relates to
the distance from the mean in relation to standard
deviations. Yes: I said it - the distance from the mean.
And, yes, in relation to standard deviations.
-----

------------------------------------------------------------

And, remember, you repeatedly insisted that the standard deviations were
depicted correctly even on images where the line was *clearly* drawn
someplace other than the inflection point - the most extreme of these being
the top image here: <http://tmp.gallopinginsanity.com/sd.png>

Again, there is no "if" here - there is math. You were simply wrong to say
that the graph was a correct depiction. It is not and it was not. If you
had *really* known what an inflection point was before I suggested you look
it up up you would not have made the mistake you did.

> But the distance from the mean to the first inflection point is not the
> same for every curve, like you seem to believe. Standard deviation covers area
> under the curve, not percentage distance from the mean.

What do you mean by "percentage distance"? Remember, as you were told: it
is the distance from the mean to the inflection points on a bell curve...
or, before I was being so specific for you and merely giving you hints, it
is "based on a property of the curve". And this is true of *every* bell
curve. Every. Single. One. It is always the same. When you claim it is
"not the same for every curve" you are wrong. Again: this is not a debate -
it is my simply telling you that you are wrong. That *is* where the first
standard deviation is. Period. Your denials and your squirming *will not
change this*. Got it yet?

On the other hand: good for you for finally showing *some* understanding of
this. Based on your new understanding can you see why the top image on my
link has the standard deviation *clearly* incorrect:
<http://tmp.gallopinginsanity.com/sd.png>

If you *really* get the concept of an inflection point (visually: where the
concavity changes - do you need help understanding what that means?) you
will now see why you were very, very wrong to deny my observation that the
image shown there is quite wrong.

But you will *never* admit to that. You *hate* to admit you are wrong - no
matter how massively wrong you are.





--
🙈🙉🙊

[Snit]

On 5/31/12 1:30 PM, in article
96e4e3dd-f963-4573...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 3:52:36 PM UTC-4, Snit wrote:
>> On 5/31/12 12:35 PM, in article
>> 86c78db3-6cc5-4cb5...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>>
>>> 4) How do I know my line is a better fit than yours?
>>

>> You do not know your line is "better" than mine, because mine is 100%
>> correct.
>>
>>> If can answer these questions, then congratulations, you know some very
>>> basic statistical analysis (it can all be done in Excel for shit's sake).
>>>
>> Nobody is saying you cannot calculate an average or a standard deviation in
>> Excel. Wow. You really are lost.
>>
>> But, hey, if you want to know how to calculate a linear trend line:
>> <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>
>>
>> But you claim Excel and Numbers both get it wrong... and only you have been
>> able to figure it out! Oh, wait... you then claimed I made some mistake in
>> how I used the tools in those programs - but you could never figure out what
>> that mistake was!
>>
>> LOL!
>>
>> Seriously, cc, you are a complete idiot. I like posting to you because you
>> cannot help but be completely arrogant in your ignorance. It amuses me to no
>> end.
>

> I do. It's called an R^2 value or coefficient of determination. Excel gives
> you the value.
>
> This is getting quite funny.

This is hilarious.... the topic is how to create a linear trend line... and
you claimed I made some error in both Excel and Numbers.

And will go to your grave never admitting you are wrong but also never
actually describing any error I made - because I made *no* error in working
with Excel and Numbers and having them calculate the linear trend line.

Oh, and so you know, the "coefficient of determination" is not synonymous
with the linear trend line in discussion. You are, again, showing your
ignorance.



--
🙈🙉🙊

[cc]

On Thursday, May 31, 2012 5:12:12 PM UTC-4, Snit wrote:
> On 5/31/12 1:30 PM, in article
> 96e4e3dd-f963-4573...@googlegroups.com, "cc"
> <scat...@hotmail.com> wrote:
>
> >
> >
> > I do. It's called an R^2 value or coefficient of determination. Excel gives
> > you the value.
> >
> > This is getting quite funny.
>
> This is hilarious.... the topic is how to create a linear trend line... and
> you claimed I made some error in both Excel and Numbers.

Yes. You can't even answer why standard deviation, average, and control limits are used.

> And will go to your grave never admitting you are wrong but also never
> actually describing any error I made - because I made *no* error in working
> with Excel and Numbers and having them calculate the linear trend line.
>
> Oh, and so you know, the "coefficient of determination" is not synonymous
> with the linear trend line in discussion. You are, again, showing your
> ignorance.
>
>

Synonymous is not the word you're looking for, but R^2 values determine the best fit for trend lines, just like the one in this discussion. My R^2 value is closer to 1 than yours, meaning my trendline is a better fit.

http://en.wikipedia.org/wiki/Trend_estimation#Goodness_of_fit_.28R-squared.29_and_trend

If you'll remember before I asked why you chose a linear trend line. You didn't know why. Well you unwittingly chose the best fit based on it's R^2 value (unless you wanted to get needlessly complicated).

http://office.microsoft.com/en-us/help/choosing-the-best-trendline-for-your-data-HP005262321.aspx


Here's how you can get an R^2 value for your trendline in Excel:

http://pages.towson.edu/racasey/tools/genchemexcel.htm#Best

Care to try again?

[cc]

On Thursday, May 31, 2012 5:04:49 PM UTC-4, Snit wrote:
> On 5/31/12 1:27 PM, in article
> 0d1fed23-5252-41ae...@googlegroups.com, "cc"
> <scat...@hotmail.com> wrote:
>
> > On Thursday, May 31, 2012 3:47:49 PM UTC-4, Snit wrote:
> >> On 5/31/12 12:16 PM, in article
> >> b73e54ea-62ba-4529...@googlegroups.com, "cc"
> >> <scat...@hotmail.com> wrote:
> >>
> >>> I pointed out that the distance to the first standard deviation/inflection
> >>> points is not the same for every curve, which is true.
> >>
> >> It is *always* the distance from the mean to the inflection point on a bell
> >> curve. Always. Exactly the same.
> >
> > You're getting confused.
>
> Nope.

Unfortunately you still are.

Let's say x = distance from mean to inflection point on curve1. y = distance from mean to inflection point on curve2. x != y in all cases (only if they are the same curve).

Let's say i = area under curve1 to first sigma line. j = area under curve2 to first sigma line. i = j in all cases, even for different curves.

Get it now?

[Snit]

On 5/31/12 2:21 PM, in article
0605d161-cf57-404e...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 5:12:12 PM UTC-4, Snit wrote:
>> On 5/31/12 1:30 PM, in article
>> 96e4e3dd-f963-4573...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>>
>>>
>>> I do. It's called an R^2 value or coefficient of determination. Excel gives
>>> you the value.
>>>
>>> This is getting quite funny.
>>
>> This is hilarious.... the topic is how to create a linear trend line... and
>> you claimed I made some error in both Excel and Numbers.
>
> Yes. You can't even answer why standard deviation, average, and control limits
> are used.

I did answer why *you* used them: you had no idea how to use the build in
tools in Excel (or Numbers). This is true even though I showed you... and
then you *incorrectly* claimed I made a mistake. You are, of course, wrong,
as you are so often. But you will not admit it.

My favorite example of your utter incompetence (and subsequent denial) is
shown here: <http://tmp.gallopinginsanity.com/sd.png>

You insisted the depiction of a standard deviation was correct in that top
image. Now that you know what an inflection point is and that the standard
deviation is the distance from the mean to the inflection points on those
curves, surely you *must* know you were wrong to deny I correct in noting
how poorly the image was drawn. You must... not even you can be so lost as
to *still* not see that.

But you will never admit to it. Never. You simply are going to pretend you
were not ignorant... but you were. Hey, here are some quotes from you about
that image and others which were clearly wrong:

cc:
-----
There'se nothing wrong with the image, other than some weird
axis labeling.
-----
Snit's so fucking stupid he thinks the sigma lines are drawn
based on distance from the mean, not area under the curve.
-----
| The sigma lines are drawn based on the area of the curve -
| which is easy to see when the images screw it up, esp. when
| they do so really badly, like in some of the ones I showed
| you.
They are not wrong.
------
LOL!!!! All of those links are fine. The first sigma lines
cover 68% of the area UNDER THE CURVE.
-----
If you would like to prove, on any single one of the links
you call incorrect, that the first sigma lines do not bound
an area that is 68.2% of the area UNDER THE CURVE, then I
would like to see it.
-----
Hahahaha your "approximate inflection points" are hilarious.
Please, post more on this subject.
------

I know exactly what an inflection point is. It's where the
second derivative changes sign, and it's exactly where the

sigma lines are in your supposed incorrect examples. Funny
how you're now questioning the applications used to generate
those graphs! Face it, you're wrong.
------

On and on and on and on and on. You just made a complete and total fool of
yourself.

Seriously, now that you have been educated on the subject, can you not see
why the top depiction at my link is very wrong?

<http://tmp.gallopinginsanity.com/sd.png>

Simple yes or no answer: can you tell it is wrong. It is wrong, of course,
that is not in debate - the question is if you can or cannot tell.
Remember, before you repeatedly denied it and the other poorly done examples
I showed you were wrong. Can you see, now, where they were wrong?

...



--
🙈🙉🙊

[Snit]

On 5/31/12 2:21 PM, in article
0605d161-cf57-404e...@googlegroups.com, "cc"
<scat...@hotmail.com> wrote:

...

>> And will go to your grave never admitting you are wrong but also never
>> actually describing any error I made - because I made *no* error in working
>> with Excel and Numbers and having them calculate the linear trend line.
>>
>> Oh, and so you know, the "coefficient of determination" is not synonymous
>> with the linear trend line in discussion. You are, again, showing your
>> ignorance.
>
> Synonymous is not the word you're looking for

It is the exact word I was "looking for", for it is the correct word.

Remember, before you claimed you had made a linear trend line:

-----
I made a simple linear trend line that is the CORRECT way of
doing things (and is incredibly easy in Excel).
-----

But you also claimed to get a different result than what Excel and Numbers
get when they calculate their trend lines:

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

> , but R^2 values determine the best fit for trend lines, just like the one in
> this discussion. My R^2 value is closer to 1 than yours, meaning my trendline
> is a better fit.
>
> http://en.wikipedia.org/wiki/Trend_estimation#Goodness_of_fit_.28R-squared.29_
> and_trend
>
> If you'll remember before I asked why you chose a linear trend line. You
> didn't know why.

You made that up.

> Well you unwittingly chose the best fit based on it's R^2
> value (unless you wanted to get needlessly complicated).
>
> http://office.microsoft.com/en-us/help/choosing-the-best-trendline-for-your-da
> ta-HP005262321.aspx
>
>
> Here's how you can get an R^2 value for your trendline in Excel:
>
> http://pages.towson.edu/racasey/tools/genchemexcel.htm#Best
>
> Care to try again?

No need to "try" to show you are wrong... you simply are. Remember: this is
*not* a debate about who is wrong or right... it is simply me being amused
by how far you will go in denying your error.

Oh, and if you want to see the R² value for the chart, it is trivial to
produce... in the case of the trend line I showed you, it is: R² = 0.20083

Do you need me to show you a video of how to get that? LOL!



--
🙈🙉🙊

[Snit]

On 5/31/12 2:58 PM, in article
3c5824b1-f203-4b27...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 5:04:49 PM UTC-4, Snit wrote:
>> On 5/31/12 1:27 PM, in article
>> 0d1fed23-5252-41ae...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>> On Thursday, May 31, 2012 3:47:49 PM UTC-4, Snit wrote:
>>>> On 5/31/12 12:16 PM, in article
>>>> b73e54ea-62ba-4529...@googlegroups.com, "cc"
>>>> <scat...@hotmail.com> wrote:
>>>>
>>>>> I pointed out that the distance to the first standard deviation/inflection
>>>>> points is not the same for every curve, which is true.
>>>>
>>>> It is *always* the distance from the mean to the inflection point on a bell
>>>> curve. Always. Exactly the same.
>>>
>>> You're getting confused.
>>

>> Nope. I am *exactly* correct. And you were *exactly* incorrect when you
>> made the following claims:
>>
>> cc:
>> -----
>> It's area under the curve, not distance from center you
>> idiot. Jesus Christ, give up now.
>> -----
>> It is not distance from the mean. Repeat: it is not
>> distance from the mean.
>> -----

And you never have admitted you were wrong about this. Never. And you will
not. You simply are going to lie until the end of time. 100% predictable.

>> But it *is* the distance from the center (the mean) to the inflection point.
>> There is no "confusion" about this... it *is* the way it *is*. No argument.
>> No maybe.

Again: no comment from cc. You just snipped and ran.

>> This is not a place where there is any chance of my being wrong or any chance
>> of you being right. This is a case of accepted *math*. You will not prove
>> math wrong (just as you will not prove Excel and Numbers wrong, nor *ever*
>> find a place where my linear trend line is wrong in the following examples):
>>
>> <http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
>> <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>
>>
>> Remember: there is no debate here - you, simply, are wrong. Period. The
>> game now is to see how extreme you will get in your denial - it is not as if
>> you are ever going to show a flaw with my work or with Excel or with Numbers
>> or with *math* itself. You will *never* do this.
>>
>> But goodness, you will be amusing as you squirm and insist you are
>> oh-so-right when you are just blowing smoke.

And you are squirming and snipping and praying someone - anyone - will
believe your BS.

But there is no debate: just the amusement of seeing how long you will
insist you are right when you are *clearly* wrong.

>>> The first standard deviation point is always at the inflection point, like I
>>> said.
>>
>> Well, as you said after I told you about it, sure.

And no comment about how you only said this *after* I educated you on the
topic... and how I gave you hint after hint and you showed *no* sign of
understanding.

But, sure, you did eventually say it. LOL!

>> Here is the message where I told you:
>> <http://groups.google.com/group/comp.os.linux.advocacy/msg/148c20587f1a5898>
>>
>> ------------------------------------------------------------
>> Snit:
>> -----
>> Ok, time to give you a hint: it *very much is* the "distance
>> from the center"... which, by the way, is generally referred
>> to as the "mean" (which is not the same way you are a mean
>> person, but the average). Again: There is a very specific
>> distance from the mean - one which is easy to see on a graph
>> - which you clearly are ignorant about. And that is a very
>> clear hint... maybe even you can figure out your error from
>> there (happy Googling!).
>> -----
>>
>> cc:
>> -----
>> Holy Shit! You are completely wrong! Standard deviation for a
>> normal distribution is the percentage of the area UNDER THE
>> CURVE!
>> -----
>>
>> ------------------------------------------------------------

Gee, cc, you snipped that - no admission of your error. Man, who could have
guessed you would run away from your own mistakes? LOL!

>> I even told you what type distance it was:
>>
>> cc:
>> -----
>> It is not distance from the mean.
>> ------
>>
>> Snit:
>> -----
>> It is *also* that... a very specific distance. Again, not
>> in inches or millimeters or something like that, but based
>> on a property of the curve. Wow... that is a *huge* hint
>> for you. I mean it is almost screaming the "secret" to
>> you. Can you figure it out from that hint? I am being
>> kind and giving you more and more, even though you are
>> refusing to admit you do not know and you are even
>> fabricating stories about me.
>> -----
>>
>> ------------------------------------------------------------

Look below where you pretend that you thought I did mean in inches or
millimeters or whatever. LOL!

Come on, cc, why not just be honest and admit you had no clue?

Because to do so you would have to be honest - and you are cc, and you are
unable to be honest when you are proved wrong.

>> And then I gave you the big hint... and apparently you took my advice and
>> looked up the concept of an inflection point:
>>
>> Snit:
>> -----
>> Ok, one more *huge* hint for you. Huge. I mean, really,
>> this is giving it away... but I am taking pity on your
>> ignorance and your self-humiliation and feeling a bit bad
>> for you. Consider the concept of an "inflection point".
>> Look it up if you have to (as you almost surely will).
>> Yeah, then see if you can find how that concept relates to
>> the distance from the mean in relation to standard
>> deviations. Yes: I said it - the distance from the mean.
>> And, yes, in relation to standard deviations.
>> -----
>>
>> ------------------------------------------------------------
>>
>> And, remember, you repeatedly insisted that the standard deviations were
>> depicted correctly even on images where the line was *clearly* drawn
>> someplace other than the inflection point - the most extreme of these being
>> the top image here: <http://tmp.gallopinginsanity.com/sd.png>

No comment from you. Come on, cc, can you *now* see where that depiction is
wrong?

But you *will* run from that question. LOL! Always. You *fear* honesty.
The fact is I have educated you enough where you can tell it is wrong... but

you will never admit to it.

>> Again, there is no "if" here - there is math. You were simply wrong to say
>> that the graph was a correct depiction. It is not and it was not. If you
>> had *really* known what an inflection point was before I suggested you look
>> it up up you would not have made the mistake you did.
>>
>>> But the distance from the mean to the first inflection point is not the same
>>> for every curve, like you seem to believe. Standard deviation covers area
>>> under the curve, not percentage distance from the mean.
>>>
>> What do you mean by "percentage distance"?

No answer from you. Really: what the heck did you mean by that? Given that
the curve goes off into infinity, how do you get a percentage of *that*? A
percentage of infinity? You just keep showing off how ignorant you are.

>> Remember, as you were told: it is the distance from the mean to the
>> inflection points on a bell curve... or, before I was being so specific for
>> you and merely giving you hints, it is "based on a property of the curve".
>> And this is true of *every* bell curve. Every. Single. One. It is always
>> the same. When you claim it is "not the same for every curve" you are wrong.
>> Again: this is not a debate - it is my simply telling you that you are wrong.
>> That *is* where the first standard deviation is. Period. Your denials and
>> your squirming *will not change this*. Got it yet?
>>
>> On the other hand: good for you for finally showing *some* understanding of
>> this. Based on your new understanding can you see why the top image on my
>> link has the standard deviation *clearly* incorrect:
>> <http://tmp.gallopinginsanity.com/sd.png>

How many times have I asked you this question? How many times have you run?
I do not know... but if you answer the first question you will have the
answer to the second... they are the same. Yeah, you run *every* time.

>> If you *really* get the concept of an inflection point (visually: where the
>> concavity changes - do you need help understanding what that means?) you will
>> now see why you were very, very wrong to deny my observation that the image
>> shown there is quite wrong.
>>
>> But you will *never* admit to that. You *hate* to admit you are wrong - no
>> matter how massively wrong you are.

Other than the *first word*, you snipped all of the above. My goodness you
are becoming quite the coward. Yeah, you know you are wrong... your actions
are proving it. You know I am right. Oh well... how far will you go to
pretend you do not know you are wrong?

> Unfortunately you still are.
>
> Let's say x = distance from mean to inflection point on curve1. y = distance
> from mean to inflection point on curve2. x != y in all cases (only if they are
> the same curve).
>
> Let's say i = area under curve1 to first sigma line. j = area under curve2 to
> first sigma line. i = j in all cases, even for different curves.
>
> Get it now?

In other words you are looking to measure it in inches or centimeters or the
like - exactly what I told you *not* to do:

cc:
-----
It is not distance from the mean.
-----

Snit:
-----
It is *also* that... a very specific distance. Again, not in
inches or millimeters or something like that, but based on a
property of the curve. Wow... that is a *huge* hint for you.
I mean it is almost screaming the "secret" to you. Can you
figure it out from that hint? I am being kind and giving you
more and more, even though you are refusing to admit you do
not know and you are even fabricating stories about me.
------

Again: if you understand the concept, you would not have argued with me when
I noted how incorrect some of the depictions of the standard deviation
were... the worse of the examples being shown as the first example here:

<http://tmp.gallopinginsanity.com/sd.png>

But you have yet to admit that the top depiction is wrong. But it is. Come
on, cc, can't you at least try to admit to that now? LOL! No... that would
mean you would have to admit you were wrong to deny it before.

And you will *never* admit you were wrong.

You are a very, very funny troll.


--
🙈🙉🙊

[cc]

I just gave you directions on how to do it, retard. Do you know how bad an R^2 value of .2 is?

[cc]

On Thursday, May 31, 2012 6:40:43 PM UTC-4, Snit wrote:
> On 5/31/12 2:58 PM, in article
> 3c5824b1-f203-4b27...@googlegroups.com, "cc"
> <scat...@hotmail.com> wrote:
>
> > On Thursday, May 31, 2012 5:04:49 PM UTC-4, Snit wrote:
> >> On 5/31/12 1:27 PM, in article
> >>

> >> cc:
> >> -----
> >> Holy Shit! You are completely wrong! Standard deviation for a
> >> normal distribution is the percentage of the area UNDER THE
> >> CURVE!
> >> -----
> >>
> >> ------------------------------------------------------------
>
> Gee, cc, you snipped that - no admission of your error. Man, who could have
> guessed you would run away from your own mistakes? LOL!
>

So you're maintaining that the first sigma lines do not mark the boundaries of 68.2% of the area?

"To be more precise, the area under the bell curve between μ − nσ and μ + nσ"
http://en.wikipedia.org/wiki/Normal_distribution#Standard_deviation_and_confidence_intervals

"The check buttons below will help you realize the appropriate percentages of the area under the curve."
http://www-stat.stanford.edu/~naras/jsm/NormalDensity/NormalDensity.html

I suppose the Stanford statistics department fucked up?

"The table below can be used to find the area under the curve from the central line to any "Z-score" value up to 3, in steps of 0.01 This will then tell you what portion of the population are within "Z" standard deviations of the mean."
http://www.mathsisfun.com/data/standard-normal-distribution-table.html

Did't I bring up the Z-table before?

[Snit]

On 6/4/12 6:07 AM, in article
050e80c4-683d-48d2...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 6:40:43 PM UTC-4, Snit wrote:
>> On 5/31/12 2:58 PM, in article
>> 3c5824b1-f203-4b27...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>> On Thursday, May 31, 2012 5:04:49 PM UTC-4, Snit wrote:
>>>> On 5/31/12 1:27 PM, in article
>>>>
>>>> cc:
>>>> -----
>>>> Holy Shit! You are completely wrong! Standard deviation for a
>>>> normal distribution is the percentage of the area UNDER THE
>>>> CURVE!
>>>> -----
>>>>
>>>> ------------------------------------------------------------
>>
>> Gee, cc, you snipped that - no admission of your error. Man, who could have
>> guessed you would run away from your own mistakes? LOL!
>
> So you're maintaining that the first sigma lines do not mark the boundaries of
> 68.2% of the area?

You made that up.

Really: you do that *lot*. A whole lot. You just make things up. And then
you run from questions. Here, one of the questions you are running from
(along with some of my commentary on how absurd your BS is):

-----
Do you see now why I was (and am) correct: it is easy to see
when the image is screwed up (at least when the screw up is
significant, as in this example: <http://goo.gl/HtsTx>).

I mean, really, now that you know to look for where the
concavity changes, can't you admit you can see where they
image is wrong? Here is more info on that if you need it:
<http://tmp.gallopinginsanity.com/sd.png>.

Face it, cc, it is not as if you were wrong just once or
twice - you kept making the same *clearly* incorrect claim
and saying it was hysterical how wrong I was... when I was
*clearly* right.

Don't you find that at least a little bit funny? I know I do!
------

Come on, don't you find your ignorance, in the face of your arrogance, to be
funny? I know I do!

> "To be more precise, the area under the bell curve between μ − nσ and μ + nσ"
> http://en.wikipedia.org/wiki/Normal_distribution#Standard_deviation_and_confid
> ence_intervals
>
> "The check buttons below will help you realize the appropriate percentages of
> the area under the curve."
> http://www-stat.stanford.edu/~naras/jsm/NormalDensity/NormalDensity.html
>
> I suppose the Stanford statistics department fucked up?

How so? Keep in mind, that is *your* comment, not mine.

cc:
-----

I suppose the Stanford statistics department fucked up?

-----

That is not something I said or implied. If you look at the link you show,
just where do you think they are making mistakes? Where do you think anyone
other than you has suggested they are?

> "The table below can be used to find the area under the curve from the central
> line to any "Z-score" value up to 3, in steps of 0.01 This will then tell you
> what portion of the population are within "Z" standard deviations of the
> mean."
> http://www.mathsisfun.com/data/standard-normal-distribution-table.html
>
> Did't I bring up the Z-table before?

Yes. So what? Really, do you think that makes the fact you proved your
ignorance go away? I mean, are you trying to imply someone was denying the
existence of such a table? Do you even have a point, other than to try to
obfuscate what you are running from:

<http://www.udel.edu/htr/Statistics/Images/Class12/normal2.gif>

You repeatedly denied that such images, which are clearly depicted
incorrectly, were in fact depicted incorrectly.

You had no clue what you were talking about.

Why not just admit to the fact you were ignorant and arrogant? You have no
idea what an inflection point was or how it was significant to the concept
of a standard deviation. None. You were lost. And when this was proved
you freaked out.





--
🙈🙉🙊

[Steve Carroll]

Save for the controls I've asked these of him already (trying to give
the poor fool hints), he apparently didn't understand any of it. I
even suggested that he question using a line at all. Again, no
effect... trust me, he's WAY outta his league here. That he keeps
harping on you is the proof. Just stick to the above, don't engage him
on his red herring crap. He doesn't have a CLUE why the topic ever
included standard deviation, he now has to go find one and we both
know what'll happen when he does ;)

(cue up Snit claiming he never saw any hints, by either of us).


My prediction: Snit will focus on his statements about standard
deviation and veer away from his trend line. He will remove his movie
(you shoulda taken it to You Tube) and he'll replace it with one that
utilizes the above. How do I know this? He's done the exact same thing
in csma already, several times.

> <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov> - Snit's ignorance of Excel and his hilarious attempt at statistical analysis

(for the record, I have a copy of his goofy movie... but he'll just
claim you or I made it... same kind of thing he did in csma).

[cc]

On Monday, June 4, 2012 11:38:15 AM UTC-4, Snit wrote:
> On 6/4/12 6:07 AM, in article
> 050e80c4-683d-48d2...@googlegroups.com, "cc"
> <scat...@hotmail.com> wrote:
>
> > On Thursday, May 31, 2012 6:40:43 PM UTC-4, Snit wrote:
> >> On 5/31/12 2:58 PM, in article
> >> 3c5824b1-f203-4b27...@googlegroups.com, "cc"
> >> <scat...@hotmail.com> wrote:
> >>
> >>> On Thursday, May 31, 2012 5:04:49 PM UTC-4, Snit wrote:
> >>>> On 5/31/12 1:27 PM, in article
> >>>>
> >>>> cc:
> >>>> -----
> >>>> Holy Shit! You are completely wrong! Standard deviation for a
> >>>> normal distribution is the percentage of the area UNDER THE
> >>>> CURVE!
> >>>> -----
> >>>>
> >>>> ------------------------------------------------------------
> >>
> >> Gee, cc, you snipped that - no admission of your error. Man, who could have
> >> guessed you would run away from your own mistakes? LOL!
> >
> > So you're maintaining that the first sigma lines do not mark the boundaries of
> > 68.2% of the area?
>
> You made that up.

You just said my quote above was an error. Standard deviation is area under the curve, which is what I said in the quote. So either what you quote above is not a mistake or error, or you believe that the first sigma lines do not mark the boundaries of 68.2% of the area. Your call.

[Snit]

On 6/4/12 9:33 AM, in article
a25e7037-0e2a-4739...@googlegroups.com, "cc"

Your error was to deny that the images were shown incorrectly. And you were
wrong.

But you will never admit to that.

> Standard deviation is area under the curve, which is what I said in the quote.

You did say that, but you *also* claimed I was wrong to note that the images
I showed you were depicted incorrectly. It is in your denial of my correct
statement that you were wrong.

> So either what you quote above is not a mistake or error, or you believe that
> the first sigma lines do not mark the boundaries of 68.2% of the area. Your
> call.

You made that up. And in context it is clear:

I showed you some clearly incorrectly depicted sigma lines. You responded
with all sorts of bizarre and ignorant denials:

cc:
-----
There'se nothing wrong with the image, other than some weird
axis labeling.
-----
Snit's so fucking stupid he thinks the sigma lines are drawn
based on distance from the mean, not area under the curve.
-----
| The sigma lines are drawn based on the area of the curve -
| which is easy to see when the images screw it up, esp. when
| they do so really badly, like in some of the ones I showed
| you.
They are not wrong.
------
LOL!!!! All of those links are fine. The first sigma lines
cover 68% of the area UNDER THE CURVE.
-----
If you would like to prove, on any single one of the links
you call incorrect, that the first sigma lines do not bound
an area that is 68.2% of the area UNDER THE CURVE, then I

would like to see it.
-----

I know exactly what an inflection point is. It's where the
second derivative changes sign, and it's exactly where the

sigma lines are in your supposed incorrect examples. Funny
how you're now questioning the applications used to generate
those graphs! Face it, you're wrong.
------

Hahahaha your "approximate inflection points" are hilarious.
Please, post more on this subject.
------

Of note: in the last quote you asked me to post more about the inflection
points. And I did (doing you a favor... one you never thanked me for).
Even made a video to help you (and others): <http://youtu.be/MoW3hMq-eIc>.
Given the information you now have, you should be able to answer a simple
question. Looking at the following:

<http://tmp.gallopinginsanity.com/sd.png>

With the top image do you agree the depiction is wrong based on the vertical
lines not being drawn at the inflection points?

Easy question, but you will not answer - for by now even you must know you
are wrong... but to answer you would have to admit this... and *that* is
something you will never do.

You see, cc, you repeatedly pretend to have knowledge on topics you are
clearly ignorant about. And I repeatedly call you on it. Then you run...
and as you do so you lie about me and my views. Some of your lies:

1) That I denied the area under the sigma lines was incorrect.
I never did so.

2) That I claimed the linear trend line you asked me to generate
would be predictive of future data (I noted it fit with past
predictions, not that it would offer the predictive value you
keep pushing)

3) That I claimed the linear trend line was the best tool to use
to make predictions for the data in question. I never did so.

4) That your percentages of Linux usage are representative of
the percentage of *desktop* usage. They are not. You are
including in the total all web usage, including mobile
devices and the like.



--
🙈🙉🙊

[Snit]

On 5/31/12 1:27 PM, in article

0d1fed23-5252-41ae...@googlegroups.com, "cc"

<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 3:47:49 PM UTC-4, Snit wrote:
>> On 5/31/12 12:16 PM, in article
>> b73e54ea-62ba-4529...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>> I pointed out that the distance to the first standard deviation/inflection
>>> points is not the same for every curve, which is true.
>>
>> It is *always* the distance from the mean to the inflection point on a bell
>> curve. Always. Exactly the same.
>
> You're getting confused.

Nope. I am *exactly* correct. And you were *exactly* incorrect when you
made the following claims:

cc:
-----

It's area under the curve, not distance from center you
idiot. Jesus Christ, give up now.
-----
It is not distance from the mean. Repeat: it is not
distance from the mean.
-----

But it *is* the distance from the center (the mean) to the inflection point.
There is no "confusion" about this... it *is* the way it *is*. No argument.
No maybe.

This is not a place where there is any chance of my being wrong or any
chance of you being right. This is a case of accepted *math*. You will not
prove math wrong (just as you will not prove Excel and Numbers wrong, nor
*ever* find a place where my linear trend line is wrong in the following
examples):

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Remember: there is no debate here - you, simply, are wrong. Period. The
game now is to see how extreme you will get in your denial - it is not as if
you are ever going to show a flaw with my work or with Excel or with Numbers
or with *math* itself. You will *never* do this.

But goodness, you will be amusing as you squirm and insist you are
oh-so-right when you are just blowing smoke.

> The first standard deviation point is always at the inflection point, like I
> said.

Well, as you said after I told you about it, sure. Here is the message

where I told you:
<http://groups.google.com/group/comp.os.linux.advocacy/msg/148c20587f1a5898>

------------------------------------------------------------
Snit:
-----
Ok, time to give you a hint: it *very much is* the "distance
from the center"... which, by the way, is generally referred
to as the "mean" (which is not the same way you are a mean
person, but the average). Again: There is a very specific
distance from the mean - one which is easy to see on a graph
- which you clearly are ignorant about. And that is a very
clear hint... maybe even you can figure out your error from
there (happy Googling!).
-----

cc:
-----
Holy Shit! You are completely wrong! Standard deviation for a
normal distribution is the percentage of the area UNDER THE
CURVE!
-----

------------------------------------------------------------

I even told you what type distance it was:

cc:

-----
It is not distance from the mean.

------

Snit:
-----
It is *also* that... a very specific distance. Again, not
in inches or millimeters or something like that, but based
on a property of the curve. Wow... that is a *huge* hint
for you. I mean it is almost screaming the "secret" to
you. Can you figure it out from that hint? I am being
kind and giving you more and more, even though you are
refusing to admit you do not know and you are even
fabricating stories about me.
-----

------------------------------------------------------------

And then I gave you the big hint... and apparently you took my advice and
looked up the concept of an inflection point:

Snit:
-----
Ok, one more *huge* hint for you. Huge. I mean, really,
this is giving it away... but I am taking pity on your
ignorance and your self-humiliation and feeling a bit bad
for you. Consider the concept of an "inflection point".
Look it up if you have to (as you almost surely will).
Yeah, then see if you can find how that concept relates to
the distance from the mean in relation to standard
deviations. Yes: I said it - the distance from the mean.
And, yes, in relation to standard deviations.
-----

------------------------------------------------------------

And, remember, you repeatedly insisted that the standard deviations were
depicted correctly even on images where the line was *clearly* drawn
someplace other than the inflection point - the most extreme of these being
the top image here: <http://tmp.gallopinginsanity.com/sd.png>

Again, there is no "if" here - there is math. You were simply wrong to say
that the graph was a correct depiction. It is not and it was not. If you
had *really* known what an inflection point was before I suggested you look
it up up you would not have made the mistake you did.

> But the distance from the mean to the first inflection point is not the
> same for every curve, like you seem to believe. Standard deviation covers area
> under the curve, not percentage distance from the mean.

What do you mean by "percentage distance"? Remember, as you were told: it

is the distance from the mean to the inflection points on a bell curve...
or, before I was being so specific for you and merely giving you hints, it
is "based on a property of the curve". And this is true of *every* bell
curve. Every. Single. One. It is always the same. When you claim it is
"not the same for every curve" you are wrong. Again: this is not a debate -
it is my simply telling you that you are wrong. That *is* where the first
standard deviation is. Period. Your denials and your squirming *will not
change this*. Got it yet?

On the other hand: good for you for finally showing *some* understanding of
this. Based on your new understanding can you see why the top image on my
link has the standard deviation *clearly* incorrect:
<http://tmp.gallopinginsanity.com/sd.png>

If you *really* get the concept of an inflection point (visually: where the
concavity changes - do you need help understanding what that means?) you
will now see why you were very, very wrong to deny my observation that the
image shown there is quite wrong.

But you will *never* admit to that. You *hate* to admit you are wrong - no
matter how massively wrong you are.

--
🙈🙉🙊

[Steve Carroll]

Aren't you Mr. Move On.

Move on.

Idiot.

[Snit]

On 5/31/12 1:27 PM, in article
0d1fed23-5252-41ae...@googlegroups.com, "cc"
<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 3:47:49 PM UTC-4, Snit wrote:
>> On 5/31/12 12:16 PM, in article
>> b73e54ea-62ba-4529...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>> I pointed out that the distance to the first standard deviation/inflection
>>> points is not the same for every curve, which is true.
>>
>> It is *always* the distance from the mean to the inflection point on a bell
>> curve. Always. Exactly the same.
>
> You're getting confused.

Nope. I am *exactly* correct. And you were *exactly* incorrect when you
made the following claims:

cc:
-----

It's area under the curve, not distance from center you
idiot. Jesus Christ, give up now.
-----
It is not distance from the mean. Repeat: it is not
distance from the mean.
-----

But it *is* the distance from the center (the mean) to the inflection point.
There is no "confusion" about this... it *is* the way it *is*. No argument.
No maybe.

This is not a place where there is any chance of my being wrong or any
chance of you being right. This is a case of accepted *math*. You will not
prove math wrong (just as you will not prove Excel and Numbers wrong, nor
*ever* find a place where my linear trend line is wrong in the following
examples):

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Remember: there is no debate here - you, simply, are wrong. Period. The
game now is to see how extreme you will get in your denial - it is not as if
you are ever going to show a flaw with my work or with Excel or with Numbers
or with *math* itself. You will *never* do this.

But goodness, you will be amusing as you squirm and insist you are
oh-so-right when you are just blowing smoke.

> The first standard deviation point is always at the inflection point, like I
> said.

Well, as you said after I told you about it, sure. Here is the message
where I told you:
<http://groups.google.com/group/comp.os.linux.advocacy/msg/148c20587f1a5898>

------------------------------------------------------------
Snit:
-----
Ok, time to give you a hint: it *very much is* the "distance
from the center"... which, by the way, is generally referred
to as the "mean" (which is not the same way you are a mean
person, but the average). Again: There is a very specific
distance from the mean - one which is easy to see on a graph
- which you clearly are ignorant about. And that is a very
clear hint... maybe even you can figure out your error from
there (happy Googling!).
-----

cc:
-----
Holy Shit! You are completely wrong! Standard deviation for a
normal distribution is the percentage of the area UNDER THE
CURVE!
-----

------------------------------------------------------------

I even told you what type distance it was:

cc:

-----
It is not distance from the mean.

> But the distance from the mean to the first inflection point is not the
> same for every curve, like you seem to believe. Standard deviation covers area
> under the curve, not percentage distance from the mean.

[Snit]

On 5/31/12 2:21 PM, in article
0605d161-cf57-404e...@googlegroups.com, "cc"
<scat...@hotmail.com> wrote:

> On Thursday, May 31, 2012 5:12:12 PM UTC-4, Snit wrote:
>> On 5/31/12 1:30 PM, in article
>> 96e4e3dd-f963-4573...@googlegroups.com, "cc"
>> <scat...@hotmail.com> wrote:
>>
>>>
>>>
>>> I do. It's called an R^2 value or coefficient of determination. Excel gives
>>> you the value.
>>>
>>> This is getting quite funny.
>>
>> This is hilarious.... the topic is how to create a linear trend line... and
>> you claimed I made some error in both Excel and Numbers.
>
> Yes. You can't even answer why standard deviation, average, and control limits
> are used.

I did answer why *you* used them: you had no idea how to use the build in
tools in Excel (or Numbers). This is true even though I showed you... and
then you *incorrectly* claimed I made a mistake. You are, of course, wrong,
as you are so often. But you will not admit it.

My favorite example of your utter incompetence (and subsequent denial) is
shown here: <http://tmp.gallopinginsanity.com/sd.png>

You insisted the depiction of a standard deviation was correct in that top
image. Now that you know what an inflection point is and that the standard
deviation is the distance from the mean to the inflection points on those
curves, surely you *must* know you were wrong to deny I correct in noting
how poorly the image was drawn. You must... not even you can be so lost as
to *still* not see that.

But you will never admit to it. Never. You simply are going to pretend you
were not ignorant... but you were. Hey, here are some quotes from you about
that image and others which were clearly wrong:

cc:
-----
There'se nothing wrong with the image, other than some weird
axis labeling.
-----
Snit's so fucking stupid he thinks the sigma lines are drawn
based on distance from the mean, not area under the curve.
-----
| The sigma lines are drawn based on the area of the curve -
| which is easy to see when the images screw it up, esp. when
| they do so really badly, like in some of the ones I showed
| you.
They are not wrong.
------
LOL!!!! All of those links are fine. The first sigma lines
cover 68% of the area UNDER THE CURVE.
-----
If you would like to prove, on any single one of the links
you call incorrect, that the first sigma lines do not bound
an area that is 68.2% of the area UNDER THE CURVE, then I
would like to see it.
-----

Hahahaha your "approximate inflection points" are hilarious.
Please, post more on this subject.
------
I know exactly what an inflection point is. It's where the
second derivative changes sign, and it's exactly where the
sigma lines are in your supposed incorrect examples. Funny
how you're now questioning the applications used to generate
those graphs! Face it, you're wrong.
------

[Snit]

On 5/31/12 2:21 PM, in article
0605d161-cf57-404e...@googlegroups.com, "cc"
<scat...@hotmail.com> wrote:

...

>> And will go to your grave never admitting you are wrong but also never
>> actually describing any error I made - because I made *no* error in working
>> with Excel and Numbers and having them calculate the linear trend line.
>>
>> Oh, and so you know, the "coefficient of determination" is not synonymous
>> with the linear trend line in discussion. You are, again, showing your
>> ignorance.
>

> Synonymous is not the word you're looking for

It is the exact word I was "looking for", for it is the correct word.

Remember, before you claimed you had made a linear trend line:

-----
I made a simple linear trend line that is the CORRECT way of
doing things (and is incredibly easy in Excel).
-----

But you also claimed to get a different result than what Excel and Numbers
get when they calculate their trend lines:

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

> , but R^2 values determine the best fit for trend lines, just like the one in
> this discussion. My R^2 value is closer to 1 than yours, meaning my trendline
> is a better fit.
>
> http://en.wikipedia.org/wiki/Trend_estimation#Goodness_of_fit_.28R-squared.29_
> and_trend
>
> If you'll remember before I asked why you chose a linear trend line. You
> didn't know why.

You made that up.

> Well you unwittingly chose the best fit based on it's R^2
> value (unless you wanted to get needlessly complicated).
>
> http://office.microsoft.com/en-us/help/choosing-the-best-trendline-for-your-da
> ta-HP005262321.aspx
>
>
> Here's how you can get an R^2 value for your trendline in Excel:
>
> http://pages.towson.edu/racasey/tools/genchemexcel.htm#Best
>
> Care to try again?

No need to "try" to show you are wrong... you simply are. Remember: this is
*not* a debate about who is wrong or right... it is simply me being amused
by how far you will go in denying your error.

Oh, and if you want to see the R² value for the chart, it is trivial to
produce... in the case of the trend line I showed you, it is: R² = 0.20083

Do you need me to show you a video of how to get that? LOL!

--
🙈🙉🙊

[Snit]

On 5/31/12 2:21 PM, in article
0605d161-cf57-404e...@googlegroups.com, "cc"
<scat...@hotmail.com> wrote:

...

>> And will go to your grave never admitting you are wrong but also never
>> actually describing any error I made - because I made *no* error in working
>> with Excel and Numbers and having them calculate the linear trend line.
>>
>> Oh, and so you know, the "coefficient of determination" is not synonymous
>> with the linear trend line in discussion. You are, again, showing your
>> ignorance.
>

> Synonymous is not the word you're looking for

It is the exact word I was "looking for", for it is the correct word.

Remember, before you claimed you had made a linear trend line:

-----
I made a simple linear trend line that is the CORRECT way of
doing things (and is incredibly easy in Excel).
-----

But you also claimed to get a different result than what Excel and Numbers
get when they calculate their trend lines:

<http://tmp.gallopinginsanity.com/LinuxTrendMar2012Snit-vs-cc.png>
<http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

> , but R^2 values determine the best fit for trend lines, just like the one in
> this discussion. My R^2 value is closer to 1 than yours, meaning my trendline
> is a better fit.
>
> http://en.wikipedia.org/wiki/Trend_estimation#Goodness_of_fit_.28R-squared.29_
> and_trend
>
> If you'll remember before I asked why you chose a linear trend line. You
> didn't know why.

You made that up.

> Well you unwittingly chose the best fit based on it's R^2
> value (unless you wanted to get needlessly complicated).
>
> http://office.microsoft.com/en-us/help/choosing-the-best-trendline-for-your-da
> ta-HP005262321.aspx
>
>
> Here's how you can get an R^2 value for your trendline in Excel:
>
> http://pages.towson.edu/racasey/tools/genchemexcel.htm#Best
>
> Care to try again?