Hadron <hadronquark@xxxxxxxxx> writes:
> Mart van de Wege <mvdwege_public@xxxxxxxxxxxxx> writes:
>
>> Hadron <hadronquark@xxxxxxxxx> writes:
>>
>>> Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> writes:
>>>
>>>> ____/ Mart van de Wege on Sunday 04 January 2009 13:28 : \____
>>>>
>>>>> OK <otto@xxxxxxxxx> writes:
>>>>>
>>>>>> On Sat, 03 Jan 2009 16:04:18 -0600, Terry Porter
>>>>>> <linux-2@xxxxxxxxxxxxxxx> wrote:
>>>>>>
>>>>>>>On Sat, 03 Jan 2009 16:32:21 -0500, DFS wrote:
>>>>>>>
>>>>>>>> Dave wrote:
>>>>>>>>> http://www.howstuffworks.com/question246.htm
>>>>>>>
>>>>>>>> Popularity graph:
>>>>>>>
>>>>> <snip stats>
>>>>>>
>>>>>> Wow, stats from a wopping 3000 hits... is that a year's worth of
>>>>>> visitors on your site, or two years perhaps?
>>>>>>
>>>>>> Go, Linux, Go!!!
>>>>>>
>>>>>> http://marketshare.hitslink.com/report.aspx?qprid=9&qpcustom=Linux
>>>>>
>>>>> But Terry gives his source and raw data, so we can verify *his* stats.
>>>>>
>>>>> That link you give that keeps getting bandied about does no such
>>>>> thing, so its numbers *are* suspect.
>>>>>
>>>>> Yes, even if they show Microsoft share dropping. Any statistics quoted
>>>>> should be assumed to be lies unless the source gives you access to the
>>>>> raw data and methodology used.
>>>>
>> <snip>
>>>>
>>>> The 'study' ('survey') cited by the trolls is flawed by design. Always check
>>>> the methods; they don't want to conceal that embarrassment.
>>>
>>> Yet your trojan infected site is? LOL!
>>>
>>> Their study was a non biased hit count on OS agnostic sites. I think we
>>> know who is to be trusted.
>>
>> So, tell me, Quack, where do the good folks from hitlink tell us what
>> sites exactly are tested and how?
>>
>> Mart
>
>
> Wow. Talk about missing the point.
>
> Go read here:
>
> http://marketshare.hitslink.com/
>
Yes, I saw that site.
Now, how about some details? What sites are tested? How do they
correct their numbers for confounding factors? What is their exact
methodology? Do they offer access to the raw data?
The answers are in order: we don't tell, we don't tell, we don't tell, no.
In other words, their numbers may as well be made up. In terms of
serious statistics, they are *worthless*.
Mart
--
"We will need a longer wall when the revolution comes."
--- AJS, quoting an uncertain source.
|
|
