> -----Original Message----- > From: jdow [mailto:[EMAIL PROTECTED] > Sent: Sunday, July 31, 2005 12:14 PM > To: users@spamassassin.apache.org > Subject: Re: unwanted breakthrough > > From: "Herb Martin" <[EMAIL PROTECTED]> > > > * -3.5 HM_URIBL_SC_DBL Prevent SC-SC2 double score > > * -2.5 HM_URIBL_SC_XS Prevent SC-XS double score > > Just set these scores to 0.1 or something tiny like that. > Setting them to minus scores them minus, which you do not want.
No, I do want the MINUS value as large or nearly as large as the lesser of the two scores. I am avoiding scoring BOTH SC and SC2 (test version) at full combined value on the same email. I am also avoiding scoring both SC and XS fully. The idea is that either of those probably represent a duplicate finding so score X or Y, but not X+Y. Since X AND Y are both already scored when this test succeeds we subtract the smaller of the two scores or a proportion of the smaller scores if we feel there is some value in "double hits." -- Herb
