5 ms·

And on the flip side, I get very little spam to any of my inboxes. This includes gmail, an exchange server using Frontbridge, and some personal domains running

And on the flip side, I get very little spam to any of my inboxes. This includes gmail, an exchange server using Frontbridge, and some personal domains running on dreamhost. I'd be more interested to hear about what portion of this spam gets through modern filters.

tokenadult 18y ago
I'd be more interested to hear about what portion of this spam gets through modern filters. It's much higher to me on any email service but Gmail. In a typical week on Gmail, I get no spams to my inbox. A bad week is getting two or three, among hundreds of good messages. But I also get one or two false positives (good emails sent to spam folder) in a typical week on Gmail. Those are not high-priority messages, typically, but of course as a user I expect Gmail to be perfect in winnowing good messages from spam. I've given up using any other email service, mostly because even one that I subscribed to years ago because of its boasted spam-filtering capability shows me much more spam than good messages in my inbox. (That account was my primary email account for a long time, but now I have everything from that account forwarded to Gmail, to add one more layer of filtering before I read the messages.)
jgrahamc 18y ago
Virus Bulletin has started testing the efficacity of spam filters and will be publishing the results. In the initial tests the best product caught 96.24% with a false positive rate of 0.06% (although there are some caveats---see the article below). Using the ratios 97% is spam/3% is ham the ratio of spam to ham in your inbox using the best solution would be about 1.2. So you should see more spam in your inbox than ham, but not by much. So it's probably manageable to manually delete the spam in the inbox (if annoying). But the flip side is interesting. What's the ratio between false positives in the spam folder and captured spam messages? Using those figures above it's 1 in 50,000. i.e. almost impossible to manually look for false positives. http://www.virusbtn.com/vbspam/trialresults.xml http://www.virusbtn.com/vbspam/trialresults.xml
- Zak 18y ago
  I see significantly better results than that using gmail combined with the filter built in to Evolution. Spam almost never gets through. I would estimate that no more than 1 in 200 actual spams get through. A somewhat higher number of newsletter-type messages from businesses I've bought from in the past show up. Many times, these messages are sent to everyone who orders, with no ability to not opt in when placing an order. I would assume these messages are not counted in that 97% figure.

tokenadult 18y ago

I'd be more interested to hear about what portion of this spam gets through modern filters. It's much higher to me on any email service but Gmail. In a typical week on Gmail, I get no spams to my inbox. A bad week is getting two or three, among hundreds of good messages. But I also get one or two false positives (good emails sent to spam folder) in a typical week on Gmail. Those are not high-priority messages, typically, but of course as a user I expect Gmail to be perfect in winnowing good messages from spam. I've given up using any other email service, mostly because even one that I subscribed to years ago because of its boasted spam-filtering capability shows me much more spam than good messages in my inbox. (That account was my primary email account for a long time, but now I have everything from that account forwarded to Gmail, to add one more layer of filtering before I read the messages.)

jgrahamc 18y ago

Virus Bulletin has started testing the efficacity of spam filters and will be publishing the results. In the initial tests the best product caught 96.24% with a false positive rate of 0.06% (although there are some caveats---see the article below). Using the ratios 97% is spam/3% is ham the ratio of spam to ham in your inbox using the best solution would be about 1.2. So you should see more spam in your inbox than ham, but not by much. So it's probably manageable to manually delete the spam in the inbox (if annoying). But the flip side is interesting. What's the ratio between false positives in the spam folder and captured spam messages? Using those figures above it's 1 in 50,000. i.e. almost impossible to manually look for false positives. http://www.virusbtn.com/vbspam/trialresults.xml http://www.virusbtn.com/vbspam/trialresults.xml

Zak 18y ago
I see significantly better results than that using gmail combined with the filter built in to Evolution. Spam almost never gets through. I would estimate that no more than 1 in 200 actual spams get through. A somewhat higher number of newsletter-type messages from businesses I've bought from in the past show up. Many times, these messages are sent to everyone who orders, with no ability to not opt in when placing an order. I would assume these messages are not counted in that 97% figure.