Hmmm, that would depend though on how the score was computed. If it hit a couple of URIBLs which pushed it over the kill score, but then another round of emails came in which were not in the URIBL, then the words from the first SPAM email would not be added into the bayes.
__________________ |