>...
>
>> This rule seems nearly as bad an idea as the one someone suggested a
>> while back that would penalize everyone who uses a middle initial in
>> their "From:" line.
>
>FWIW, I've been running that rule since before it was mentioned on the list,
>and it is still moderately useful. It does hit ham, but at one point or
>however I have it scored that isn't significant. On the other hand, that
>point has more than once pushed a spam over the limit.
>
>One has to remember that its ok for some tests to hit ham, as long as they
>don't result in an FP.
>
> Loren
>
Here Loren has the entire point I am trying to make - I'll even
repeat it to be clear:
"One has to remember that its ok for some tests to hit ham, as long as they
don't result in an FP."
Most of my "ham" hits at least two or three rules, but still ends
up with a negative score. What is more important than just the S/O ratios
is the "overlap" between rules - a lot of rules with small point values and
S/O rations as low as .7 can be *very* effective if the "overlap" is small.
Paul Shupak
[EMAIL PROTECTED]
