This should be a FAQ... scores between 0.4999 and 0.5001 do not match any Bayes rule in 2.63. (If they did, it would have a zero score anyhow...)
So you either hit no Bayes tokens, or they added up to within a whisker of 50%. The latter actually happens much more often than you would expect; it seems to be a known issue. Pierre Thomson BIC -----Original Message----- From: Roedel, Mark [mailto:[EMAIL PROTECTED] Sent: Monday, August 02, 2004 4:35 PM To: Jim Maul; [EMAIL PROTECTED] Subject: RE: Why is there no BAYES_ score? Looks like its final result from the bayes scoring was an 0.5 -- is it possible you've got the BAYES_50 rule zeroed out? -- Mark Roedel Web Programmer / Analyst LeTourneau University -----Original Message----- From: Jim Maul [mailto:[EMAIL PROTECTED] Sent: Monday, August 02, 2004 3:21 PM To: [EMAIL PROTECTED] Subject: Why is there no BAYES_ score? I have a couple addresses that i have aliased to one spam mailbox. Anything that gets sent to these addresses is most definitely spam because the addresses were only used on an old whois record ([EMAIL PROTECTED]). Occasionally i get messages to info@ which is fine, but i noticed that spamassassin didnt tag a couple of them as spam. Researching a little, i found that there was no BAYES scoring at all for the message. I cant figure out why. I have posted the output of a "spamassassin -D < message" at https://external.elih.org/debug.txt Can someone look through this incredibly long debug and possible tell me why there is no bayes scoring at all? Bayes has been on and autolearning for many months now and i have been manually learning it as well. All other messages i have checked have atleast some BAYES scores.
