[Spambayes] Odd behavior with messages

[Tony Meyer]

> This is POP not IMAP.
> 
> The 0% I was referring to is in the score column.
> 
> I attached a clues list for one of these messages.

[from the clues list]
> # ham trained on: 10255
> # spam trained on: 291

This is *not* good.  SpamBayes works best with a reasonably balanced
database - ratios of 2::1 or even 5::1 can work ok, but you have over 35::1!
Retraining would be a good idea.  <http://entrian.com/sbwiki/TrainingIdeas>
has more.

=Tony.Meyer

-- 
Please always include the list (spambayes at python.org) in your replies
(reply-all), and please don't send me personal mail about SpamBayes.
http://www.massey.ac.nz/~tameyer/writing/reply_all.html explains this.