On 30/03/07, Richard Lockwood <[EMAIL PROTECTED]> wrote:
The next "round number" above 0.4% is 0.5%.
Yes, but I was stating what I would have expected the value to be, not stating the value presented with some rounding.
On a sample of "visitors to BBC home page" - an inflation of over 1000% (as you are suggesting now - 0.4% to 5%) is, frankly, unlikely.
Odd then that an official BBC news article claims that the value is closer to 6% isn't it? Quote:
However, analysts believe that approximately 6% of computers users run Linux
Link: <http://news.bbc.co.uk/1/hi/technology/6506027.stm?ls> (Incidentally Thursday's most popular story)
If you'd argued that Linux use was more likely to be 0.5% than 0.4% you'd: a) Possibly have a point b) Have been wasting everyone's time.
I am (or was) arguing that it was likely to be closer to the 5% mark than the 0.4% mark (approximately one order of magnitude out).
However, in your previous posts, you state that as the BBC stats "suggest" that Linux use is only 0.4%
Indeed, what do you think they suggest?
, they are obviously wrong due to a conspiracy, and that Linux use is, in 'fact' (with no evidence), over 1000% higher than that.
I don't _remember_ using the word conspiracy. The chances of them being completely accurate is extremely remote. Hitting such a small target (the true value) without taking account of inaccuracies would be more to do with luck than actual statistics. a 1000% difference isn't unrealistic considering we are dealing with the low end of the percentages. And the fact one of the things that was not accounted for is the traffic generated by spam robots. In email traffic spammers contribute to 90% (rounded down figure from the Guardian) of the traffic. I.e. Only 1 in 10 messages are genuine. Why are you assuming that they would not be generating similar traffic over HTTP? In fact it would make sense for it to be higher. (Harder to filter out hacked home boxes over HTTP than email, no dynamic IP should ever be passing on mail, in HTTP you would expect connections from dynamic addresses).
"Possible inaccuracies" do not cover that kind of imagined error margin.
Then maybe there is something to your conspiracy theory. Seem as the BBC's stats disagree with the BBC news articles. Something is not quite right wouldn't you agree? Either: 1. Browser stats are inaccurate 2. BBC news article is wrong 3. The BBC is attracting less of the Linux users to it's site (something that should be looked at seriously as this could be an indication the BBC is interfering with commercial markets). Pick one. (or add another). Andy -- First they ignore you then they laugh at you then they fight you then you win. - Mohandas Gandhi - Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please visit http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html. Unofficial list archive: http://www.mail-archive.com/backstage@lists.bbc.co.uk/