On 20 Jun 2006 01:47:24 -0700, "Larry Lard" <email@example.com> wrote:
>Larry Hammick wrote: >> "As the number of responses approaches infinity, the content of the >> thread converges to a limit which is independant of the original >> topic." > >No, I don't think this is true (or what you mean). As threads get >longer the content doesn't *converge* to anything - it's not the case >that by the end of a 10,000-post thread, all posts are on exactly the >same topic. I think what you want is: > >As the number of responses increases, the distance of the content from >the original topic is non-decreasing
I think this is too broad, as it would imply that once the thread gets long enough it will cover topics arbitrarily distant from the original topic. That doesn't really happen in sci.math. Either the thread dies, becomes an exchange between two people only, or mutates into one of the crank topics. It would be interesting to find out what are the basins of attraction.