Groups | Search | Server Info | Keyboard shortcuts | Login | Register

Groups > comp.os.linux.advocacy > #116568

Re: cc is proved wrong about "outliers"... but he will never admit to it.

Show all headers | View raw

On 6/28/12 5:41 AM, in article
1c75fb3d-764b-4649-a173-17637b42af8d@googlegroups.com, "cc"
<scatnubbs@hotmail.com> wrote:

...
>>>>> Once again you don't understand outliers. Just because you can fit a
>>>>> trendline to them, doesn't mean they aren't outliers.
>> 
>> You now think outliers forn trends. You can't help but make an ass out
>> of yourself with your every post and just prove how fucked in the head
>> you are.
>> 
>> http://tmp.gallopinginsanity.com/outliers.pdf
> 
> I don't think outliers form trends.

Not in all cases, perhaps, but certainly in this case!

<http://tmp.gallopinginsanity.com/LinuxTrend2011-2ndhalf.png>

Remember, it was *you* who claimed four of those six points, which
unambiguously for a trend, are "outliers".

Hey, speaking of how lost you are on this topic... what method did you use?
What you describe might be the Grubb's test.  Is that what you used to come
up with these four points you mistakenly see as outliers?

> Outliers can come in any shape or form.

Actually, no, they cannot.  As you have been told,

<http://en.wikipedia.org/wiki/Outlier>
    ----- 
    An outlying observation, or outlier, is one that appears to
    deviate markedly from other members of the sample in which it
    occurs. 
    ----- 

You see, if 17% of the data are forming a very clear pattern then it is
absurd to say they "deviate markedly from other members of the sample"!

> You're an absolute moron.

Ah, poor cc: he calls people names when he is proved wrong.  You are
completely ignorant on the topic of statistics - hence the reason you have
been wrong on almost every claim you have made:

1) You were wrong to claim it was a "fact" that some data points
   were outliers: the reality is that there is no clear
   definition nor single standard way to determine such, and thus
   such determinations are subjective and *not* "facts".
2) You made a rather silly claim to say that 17% of the data are
   "outliers" - especially when those data are in a direct series
   tied to a *very* clear trend that includes 25% of the data.
3) Outliers are atypical - you clearly are stretching to call 17%
   of the data, which occurs in a clear pattern, "atypical" to the
   data.  Tied to the top 3: <http://tinyurl.com/7hymgfg>.  Makes
   it very clear just how clueless cc is.
4) You have clearly placed a great deal of focus on the
   correlation coefficient even though it is unwise to do so at
   the expense of a focus on the visual graphs / visible trends.
5) You were wrong to deny that the sigma lines can be correctly
   drawn based on the distance from the mean to the inflection
   points. 
6) You were wrong to deny that the poorly done depictions I showed
   you were, in fact, poorly done.  They were unambiguously wrong.
7) You were wrong to say I missed steps in the creation of a
   linear trend line in Excel.  I did no such thing (and you
   never were able to list any steps I missed in creating a
   linear trend line, nor explain why the MS site and others
   would also miss these "steps" you spoke about)
8) You were wrong to say I was pushing the correlations I noted
   as being proof of the causation I had spoken of earlier.  I
   did no such thing.
9) You repeatedly snip and ignore comments which are contrary to
   your claimed views.
 
>> Look at that and learn! Snit goes out of his way to make lessons for
>> you and you are too fucked in the head to thank him.
>> 
>>>>>> 3) Outliers are atypical - and it is a stretch to call 17% of the data
>>>>>>    "atypical".
>>> 
>>>>> By all measures of outliers (quartiles, avg+stddev, etc), they are
>>>>> outliers.
>> 
>> It is your opinion that they are outliers and that opinion is based on
>> you being retarded. There is nothing more to this. You are a fucking
>> retard.
> 
> It is not opinion. I gave two STANDARD operations for finding outliers. If you
> have a method that does not make them outliers, them I'm all ears. But you
> don't.

See, you keep denying it is your opinion... but that is just silly.  The
determination of what is an outlier is almost always based on opinion
(though there are times when there would be pretty good consensus on such
opinions).

But you have been told this already.  Here, linked:

Details on cc's "outliers" BS: <http://tinyurl.com/84r3ypq>

A "fact"?  Based on what.  Also from the same Wikipedia page:
    ----- 
    There is no rigid mathematical definition of what constitutes
    an outlier; determining whether or not an observation is an
    outlier is ultimately a subjective exercise.
    ----- 

And I went into more detail here: <http://tinyurl.com/7vyhttc>

    ----- 
    These are just a few of the ways you can identify outliers in
    your data set. Frequently, classifying a value as an outlier
    is a judgment call...
    ----- 

You made a judgment call... and it is a silly one.  Completely absurd.  I go
into that a lot more in that link, above.  More on cc's ignorance about
outliers: <http://tinyurl.com/7vyhttc>.

But, hey, what methods did you use?  Sounds like the Grubb's test, perhaps,
though if you did so with what significance level?  I would *love* to hear
the details of your determination!

>>>>> If you can find a method that doesn't make them outliers, I would like to
>>>>>  see it.
>> 
>> Looking at the pattern of the graph shows you they are not you fucking
>> retard.
> 
> You can't look at a graph to find outliers, you fucking retard.

Again you have been shown to be wrong!  And again from:

More on cc's ignorance about outliers: <http://tinyurl.com/7vyhttc>.

----------------------------------------------------------------------
<http://wp.me/ps1hD-bb>
    ----- 
    Diagnosing Outliers
    There are a couple of ways to check data for outliers. These
    include: 

    Visually Inspect Data

    Plot your data on a chart or graph. Do some points stand out
    from the "crowd?" If so, what is the record? Can you verify
    that it was entered correctly?
    ----- 

As you have been shown, the data set is rather "smooth" - not much stands
out: <http://tmp.gallopinginsanity.com/outliers.pdf>

In the first images, there is one point that clearly stands out... it is an
outlier.  In the second two images there are no such points - no obvious
outliers.  With the second image there is a pretty clear curve - with the
third image (the one that charts the full set of data we are discussing),
the high point is actually smoother and has fewer points that stand out from
the data which you are *not* claiming are "outliers".  When you claim the
high points from end of 2011 are "outliers" you are making a very, very poor
judgment call based on this method.

Based on this method cc screwed up - but he will never admit to it.
----------------------------------------------------------------------

Not only are you denying you screwed up based on this method, you are now
denying this method even exists! Not even I predicted you would be *that*
extreme in your denial of reality!

>>  If they were not atypical, they wouldn't be outliers. I would like to
>>>>> see some sort of cite where it is discussed what is the proper percentage
>>>>> for
>>>>> outliers. Oh, you can't find one? Imagine that.
>> 
>> Proper percentage? What the fuck? Now you think there should be a
>> specific percentage of data points that should be outliers. That is
>> more fucked up than your normal crack-induced babbling.
> 
> Actually, I said the opposite you fetal-alcohol-syndromed fuck.

You are the one who is speaking of looking for the "proper percentage for
outliers".

I merely noted that your looking at 17% of the data (a subset which forms a
part of a very clear trend) and calling then "outliers" is absurd.  Idiotic,
really.

> There is no proper percentage, so why you're harping on the fact that there is
> 17% outliers or some nonsense is bullshit. It doesn't matter how many outliers
> there are! I did standard tests for outliers, using Excel (AVERAGE, STDDEV,
> QUARTILE functions), so unless you have some proof those functions are
> incorrect, or unless you would like to argue some other method that doesn't
> show those points as outliers, then you need to just shut the fuck up.

Oh, you have refused to give the details of your "method" but now you want
to know the specifics of how you screwed up (though above and elsewhere you
have implied the quartile and Grubbs methods). Well, given any of a number
of methods I have already told you:

More on cc's ignorance about outliers: <http://tinyurl.com/7vyhttc>

Completely rips your BS about outliers apart.

> You do realize that from the same site you can go all the way back to 2007
> with stats, don't you? You do realize how much more credence that lends to
> those points being outliers?

LOL!  So now you want to change your methodology and use more data.  Enjoy!
But show it!  Give your set of data and your method!  Explain why you think
your method is correct... and be specific as to what method you are using.

Funny how every time you are proved wrong you want to use a different set of
data and a different method.  Gotta love it!
 
... 
>> You give fucked up lying answers to numbers 2 and 3 and ignore all the
>> rest of the points because you are a yellow bellied fucktard.
> 
> I believe there's a statistical term for when a single state (out of the
> entire WORLD) causes such a fluctuation in the usage numbers. What does that
> make those data points again... oh yeah, it makes them outliers!

More on cc's ignorance about outliers: <http://tinyurl.com/7vyhttc>

Come on, now, stop pretending to be knowledgeable about a topic you are
completely ignorant about, esp. once you have been proved to be completely
ignorant!

> Since you can't seem to understand even the simplest concepts, or even show
> basic reading comprehension, I suggest you shoot yourself in the head and
> spare the world your brain-damaged idiocy. I can give you a link to a video
> showing how it's done since you apparently need your hand held for absolutely
> everything.
> 
> --
> <http://tmp.gallopinginsanity.com/LinearTrendLineCreation.mov>

Thank you for the link - but that is to my own video and it is showing how
to make a linear trend line - *not* how to find outliers!

LOL!

You completely screwed up.  Again!  Oh, wait... maybe you meant you had some
other video?  If so then show it... and I will happily show you why you are
wrong to think whatever method it is showing gives the results you think it
does.  Remember: you have been proved wrong on this - absolutely proved.

By now even you must know how wrong you are - it is *the* reason you run
from all the points I keep noting where you are wrong and why you refuse to
say what methods and with what parameters you used to come up with your
idiotic claims.

> - Snit's 
> ignorance of Excel and his hilarious attempt at statistical analysis

LOL!  Poor cc... always lashing out when he is proved wrong.

-- 
Summary of cc's statistical BS: <http://tinyurl.com/7rwazxw>
Details on cc's "outliers" BS: <http://tinyurl.com/84r3ypq>
More on cc's ignorance about outliers: <http://tinyurl.com/7vyhttc>
Details on cc's sigma and R^2 BS: <http://tinyurl.com/7vambev>

Back to comp.os.linux.advocacy | Previous | Next — Previous in thread | Next in thread | Find similar

Thread

Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-25 05:31 -0700
  Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-25 06:45 -0700
    Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-25 21:47 -0700
      Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-26 05:19 -0700
        Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-26 09:46 -0700
          Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-26 10:07 -0700
            Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-26 16:59 -0700
              Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-27 05:28 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-27 08:19 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-27 10:02 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-27 11:49 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-27 11:59 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-27 12:10 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Onion Knight <onionknightgot@gmail.com> - 2012-06-27 21:49 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Onion Knight <onionknightgot@gmail.com> - 2012-06-27 22:05 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. cc <scatnubbs@hotmail.com> - 2012-06-28 05:41 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-28 09:01 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Onion Knight <onionknightgot@gmail.com> - 2012-06-29 22:04 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-30 11:03 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Onion Knight <onionknightgot@gmail.com> - 2012-06-30 21:23 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Snit <usenet@gallopinginsanity.com> - 2012-06-30 23:11 -0700
                Re: cc is proved wrong about "outliers"... but he will never admit to it. Onion Knight <onionknightgot@gmail.com> - 2012-06-27 22:08 -0700
  Re: cc's new, weaker claim about "outliers is proved wrong ... cc still in denial Snit <usenet@gallopinginsanity.com> - 2012-06-25 22:12 -0700

csiph-web