AV-Comparatives Forum

why is the link never posted !!

Oo !?!?!
I have a good Adress for you: http://www.av-comparatives.org

On the right side you will find all new Comparatives. I dont think that this "Latest Reports" Modul is too small to find!

No offense meant!

First time in years (if ever) that MS aced the test. What changed? Only AVIRA beat it out, yet MS came out far better than AVIRA in regards to FPs. New engine?

Hi,

I'm a frequent reader of these comparative reports and find them very informative. Both the testing methodology and the detailed reports provided here are above anything similar available on the Net.
I was wondering about the only piece of information that apparently is not given in the reports: the size of the clean set used to test False Positives. The absolute number of false positives detected for every product tested is given, but not the total size of the clean set, so that we have a better idea of what it means. It's clearly not the same to have 50 FPs in a clean set of one thousand files than to have it in a set of one million clean files, right? Could you provide us with that number for the last few reports and maybe include it in future reports?

Thanks

Quoted from "admin"

no.

OK, I understand, but could you elaborate a bit more about the reasons why you can't and won't?

Quoted from "admin"

if i remember fine, this is explained already in the faq section of the methodology document.

You're right, I had missed that bit.
But let me say that I partly disagree with you about that. The FAQ section says: "Giving a percentage for false alarms is in our opinion senseless and highly misleading". I think I know why you say that. I'm assuming that the clean set size is over one hundred thousand files, and therefore if you gave the number of false positives as a percentage rate, the numbers would be minuscule (e.g.: 0.003%) and most people would disregard them as practically nonexistent and also would disregard such apparently subtle differences between the percentages of different AVs (e.g., between 0.003 % and 0.004%). If that's the reason, I agree with you on that. But it wouldn't hurt to just give the total number of clean files used in the test, without expressing the numbers for each AV as a percentage but keeping them as absolute numbers just like they are reported now.
Just my humble opinion.