On 6/18/12 9:21 PM, in article firstname.lastname@example.org, "Onion Knight" <email@example.com> wrote:
>> Agreed, which is why I previously toldSnithe should be questioning >> the use of a line at all... but the argument was really a dick >> measuring contest based on which line was a better fit andSnit'sline >> lost. >> >> Onion Kinight is aSnitsock puppet or shill who has never posted to >> usenet prior to this debate...Snithas done things like this for >> years. He is one of the biggest liars you'll probably ever see on >> usenet, there are scores and scores of posters who have all labeled >> him a liar, troll or worse because of his online antics: >> >> https://groups.google.com/group/comp.os.linux.advocacy/msg/6e8fd0b61d... >> >> My suggestion: Don't get sucked into his BS. > > Snit predicted an increase in Linux usage and he showed where the data > supported this for 2011 but not 2012. You the confused him showing a > correlation with a claim of causation and have since proved you can't > understand the difference.
Carroll claims my method did not have as good of a fit... bit it showed better results:
* The increase in usage in the second half of 2011 * The decrease in 2012
cc admitted his method missed these trends... he screwed up and assumed the data was wrong because it did not fit his model... even though it was data he asked us to use!