What he may have meant is that he thinks it is more "normal" to use
something like spamassassin to assign scores to BL hits (and other
metrics), and to block only when the resulting score exceeds some
threshold. I have no idea if he really believes that, or if he's just
trying to make debating points.

For some blocklists (those with a high built-in chance of an FP), it
isn't obviously a good idea to block based only on a list hit. But for a
list with a low chance of an FP, the other heuristics may have less
predictive accuracy than the list itself; "mixing" the list result with
results from these other less-accurate tests is just going to generate
an overall result that's less accurate than the result from the list alone.

That is: I disagree with his POV (in fact there's very little that Rob
has ever uttered here that I agree with).

--
MrD.
http://ipquery.org

--
        Comments posted to news.admin.net-abuse.blocklisting
        are solely the responsibility of their author.  Please
        read the news.admin.net-abuse.blocklisting FAQ at
        http://www.blocklisting.com/faq.html before posting.