On 3/30/2012 11:51 AM, Kevin A. McGrail wrote: >> Some human being decided on the -3.0 score. > Yes, and I could argue based on my corpus that -3.0 is not harsh enough > is my basic point. > > I agree I'm not looking below the surface of the rules very much, > though. I am simply saying that XYZ rule on my corpus has an extremely > good S/O. > > Can anyone add information on this rule from their personal corpora?
I checked my logs for the last 30 days. RCVD_IN_RP_CERTIFIED hit 505 times, but did not hit a single spam message. There is also not a single case where it would have changed the ham/spam designation of the message if it had not hit. -- Bowie
