Bogofilter accuracy plummets starting around March 10, 2010

[Jonathan Kamens]

On 4/1/2010 8:59 AM, David Relson wrote:
> These attacks have been around for years.
Perhaps, but at least for me, it's never been anywhere near this bad before.

I have daily statistics going back to July 2006.  Before March 10, 2010, 
the lowest weekly success rate for bogofilter I saw was 91.1%.  My 
current weekly success rate is 79.1%.  It's more than nine standard 
deviations below my average weekly success rate for the entire period 
for which I have statistics.  Something about this attack is different.

   jik