July looks like the worst month so far for spam. So far – with about a week to go:
Spam per month (all years together - in development: display only for current year) Total in January: Ham: 924 Spam: 1666 Total in February: Ham: 1454 Spam: 1836 Total in March: Ham: 1843 Spam: 2474 Total in April: Ham: 1812 Spam: 2965 Total in May: Ham: 4549 Spam: 3269 Total in June: Ham: 2010 Spam: 5628 Total in July: Ham: 1559 Spam: 8595
That shows almost a 100% jump in spam in July. Admittedly – I did turn on bayesian filtering in May (lots of bad filtering as you can see) – but the number of spam is just climbing every month.
Spam filtering via SpamAssassin.
An increase in spam probably indicates that your address is making the rounds. In my own data I find a very consistent relationship between how old a published email address is and the amount of spam directed at it. Also the more references to an address can be found on the web, the high the spam level.
My main email address is about 8 years old and attracts about 400-500 spam emails per day 🙂
As you can see I couldn’t keep on using email without a good spam filter. All the spam that goes to my computers gets mined for “spamvertized” websites, a list of which feeds into SURBL, a realtime blacklist of spammer websites that SpamAssassin 3.x uses. In this way you probably benefit from the spam sent to my mailboxes.
Most of the mail is just broad – sent to multiple addresses.
I host my own mail server – so this is just sent to my mx.
SpamAssassin does a good job – I’ll dig more into the SURBL.
Final figures for July: