[SURBL-Discuss] Question about scoring in SA3
jeffc at surbl.org
Wed May 18 16:31:09 CEST 2005
On Wednesday, May 18, 2005, 6:54:02 AM, Kevin McGrail wrote:
> Truthfully, I was looking for something a bit more conservative that still
> reflected the possibility of a false-positive here and there ;-)
One of the reasons the scores are below the default spam
threshold of 5 is to take into account the possibility of false
positives. The Bayesean math lets SpamAssassin score many
different features of spam to make the determination. It
certainly is possible to score some features higher than 5,
which I do myself, but it should probably be evaluated in the
context of the local installation.
Don't harm innocent bystanders.
More information about the Discuss