EmailDiscussions.com - View Single Post - spam in inbox

JeremyNicoll · 4 Dec 2019, 07:18 PM

According to the chart at https://www.futurequest.net/docs/SA/

your positive SA scores, ordered from biggest to smallest are for

ME_VADESPAM_MED 2.5,

FREEMAIL_FORGED_REPLYTO 2.095, Freemail in Reply-To, but not From

BAYES_80 2,

DCC_CHECK 1.1, bulk mail

FSL_BULK_SIG 0.264, Bulk signature with no Unsubscribe

PP_MIME_FAKE_ASCII_TEXT 0.195, MIME text/plain claims to be ASCII but isn't

There was a discussion here a while ago about VADESPAM scores but I don't understand enough about the x-vs- headers in your mails to see why this mail is thought medium-likely to be spam whereas your earlier example was thought to be clean. See: http://www.emaildiscussions.com/showthread.php?t=74033

That aside that BAYES_80 presumably means you've seen fairly similar mail bodies before and classed them as spam, but for some reason the earlier example wasn't? Maybe those lists of random words are not randomly picked from many thousands but from much smaller pools and the spammer is using a different pool?

4 Dec 2019, 07:18 PM	#7
JeremyNicoll Essential Contributor Join Date: Dec 2017 Location: Scotland Posts: 490	According to the chart at https://www.futurequest.net/docs/SA/ your positive SA scores, ordered from biggest to smallest are for ME_VADESPAM_MED 2.5, FREEMAIL_FORGED_REPLYTO 2.095, Freemail in Reply-To, but not From BAYES_80 2, DCC_CHECK 1.1, bulk mail FSL_BULK_SIG 0.264, Bulk signature with no Unsubscribe PP_MIME_FAKE_ASCII_TEXT 0.195, MIME text/plain claims to be ASCII but isn't There was a discussion here a while ago about VADESPAM scores but I don't understand enough about the x-vs- headers in your mails to see why this mail is thought medium-likely to be spam whereas your earlier example was thought to be clean. See: http://www.emaildiscussions.com/showthread.php?t=74033 That aside that BAYES_80 presumably means you've seen fairly similar mail bodies before and classed them as spam, but for some reason the earlier example wasn't? Maybe those lists of random words are not randomly picked from many thousands but from much smaller pools and the spammer is using a different pool?