recent spam trends

[Bill McClain]

I keep a large "unsure" zone, just for peace of mind, so I can be more
certain that good mail is not being miscategorized. Here is a chart of
unsures plus false negatives as a percentage of total spam, summed into
10-day buckets for the last year.

With training, the unsure percentage had been declining steadily (it was as
high as 7.9% in the past) until this summer, when I started to see bursts of
more effective random word and image spam. It looks like training continues
to work on these (or the senders are only temporarily in business), but it
also seems the spammers are trying improved methods, probably directed
specifically against bayesian filters.

Does anyone know: is there a mailing list or web site where spammer's
strategies and techniques are discussed? I sometimes see "new stuff" in the
spam and would to find out "What is this for? What are they trying to
accomplish? Is it working for them"?

20051201 3.9 *******
20051212 2.3 ****
20051222 2.2 ****
20060103 1.6 ***
20060111 2.1 ****
20060121 1.8 ***
20060131 1.5 **
20060210 1.2 **
20060220 0.6 *
20060301 1.1 **
20060312 0.9 *
20060322 1.0 **
20060403 0.3
20060410 0.3
20060424 1.2 **
20060502 1.1 **
20060511 1.2 **
20060522 1.1 **
20060605 1.2 **
20060612 0.5
20060619 1.0 *
20060703 1.5 **
20060713 1.1 **
20060719 0.7 *
20060731 1.9 ***
20060814 2.2 ****
20060818 4.3 ********
20060828 1.4 **
20060910 0.7 *
20060918 0.5 *
20060927 2.4 ****
20061009 4.4 ********
20061023 2.8 *****
20061030 3.4 ******
20061106 3.8 *******
20061120 1.7 ***
20061129 1.0 **
20061208 1.6 ***

-Bill
-- 
Sattre Press                                    In the Quarter
http://sattre-press.com/                 by Robert W. Chambers
info at sattre-press.com         http://sattre-press.com/itq.html