Well that doesn’t match their chart2.php page by a long way, the chart shows 98.4% blocked, add the User Dependant 0.8% gives 99.2%. Yet the Results section shows 98.0%, so 99.2% would have been a ranking of 4th; that would make a whole heap of difference.
Some sort of statistical ranking over and above the detection rate, presumably false positives and the cluster grouping fudge factor, drops the overall from 99.2% to 98.0%
There were only 20 products tested yet 21 shown in the chart only Microsoft not appearing on the chart.av-comparatives.org/chart2.php chart making that 22 tested ???
Colour me confused, not that I’m overly concerned.
They should target even higher but scores are very good as they are. Basically they are 4th best AV (the last 3-6 AV’s have nearly the same score). If there won’t be any drastic fluctuations in the upcoming months, i think its great.
@DavidR: please note that those are two COMPLETELY different types of tests. One is only a detection rate test, while the other one (with the March results), is the WHOLE PRODUCT REAL WORLD PROTECTION TEST.
OK, I’m still not convinced these real world tests reflect the real world of a user. As for me it doesn’t take into account the web/network shields (real world), which can block exploit scripts and or redirection to a malicious site, without ever knowing what the payload might be at the remote location. Whatever that payload might be detected by the conventional detections.
I guess I just don’t understand the methodology behind it. I have tried on a couple of time to read the methodoloug.pdf file, but I never seem to make it past page 4.
The WPDT did not exist in 2008, so it is not covered in the old methodology pdf. Please read its description in the PDF provided on http://www.av-comparatives.org/en/comparativesreviews/dynamic-tests
The Protection test takes into consideration all features of the product, including web shields, behaviorblockers, etc.
FileRep warning when downloading files is similar to AutoSandbox warning in v6, so I think it’ll be counted as user-dependent.
I thought it is showed by Web Sheild… ???