On Fri, Feb 26, 2010 at 05:40:23PM -0500, dar...@chaosreigns.com wrote: > On 02/26, Ned Slider wrote: > > dar...@chaosreigns.com wrote: > >> The majority of spams making it through my filters in the last couple weeks > >> have been (confirmed) delivered by hotmail, *while* I was giving all those > >> emails a +1. > >> > >> I just went over the numbers again and set it to +3. > > > > Although it's easy to see the temptation in doing this, it's likely to > > lead to false positives for ham from hotmail (yes, occasionally one can > > receive ham from hotmail!) > > I'm aware. I'm just curious how the numbers would come out if a re-score > were run with such tests. And hoping that it might result in motivating > hotmail to work on their spam problem.
Are you aware of the ruleqa app? There are two subrules that give some impression. http://ruleqa.spamassassin.org/?rule=%2F%28HOST%7CFROM%29_HOTMAIL (other checks *.hotmail.com relay, other From: @hotmail.com) When the ratio is about fifty-fifty, it's pretty obvious what the score would be like. Ofcourse if you are controlling mail only for yourself, you are free to bump it up to +100 if you want.
