Unsolved

This post is more than 5 years old

5 Journeyman

 • 

15.6K Posts

 • 

45K Points

3174

June 8th, 2010 06:00

AV Comparatives Retrospective/Proactive Tests, May 2010

http://www.av-comparatives.org/index.php?option=com_content&view=article&id=144&Itemid=152

The testing this time focused on the ability to detect UNknown threats (i.e., before they were placed in a signature database) using generic/heuristic methods.

As a result, ALL the scores were low (relative to what these products normally receive when considering their detection rates of KNOWN, signature-based malware):   the TOP products only scored 63%.

of the popular FREE products, Microsoft Security Essentials and Avira received an "Advanced+" ranking, while Avast receieved only an "Advanced" ranking.

4 Apprentice

 • 

20.5K Posts

June 8th, 2010 19:00

Thank you for posting this, ky331. :emotion-1:

5 Journeyman

 • 

15.6K Posts

 • 

45K Points

June 8th, 2010 20:00

BB,

you'll be happy to know that Microsoft's Security Essentials did especially well in this particular test:   not only was it toward the top of the chart in proactive detection rate, it was also near the top in fewest false positives.   An excellent "1-2 punch".

in fairness, I think one other critical aspect of this test must be noted/emphasized here:   it considered only the generic/heuristic capabilities of a passive on-demand scan.  in other words, it did NOT consider on-execution / behavioral analysis.   This can (possibly) explains avast's significantly poorer showing, in that this test did not factor-in the proactive nature of avast's "behavior shield", which "monitors activity on your computer using a number of sensors (file system, registry and network based) and reports/blocks any suspicious behavior"; nor avast's [optional] ability to perform "code emulation: If avast! detects some suspicious code in a file, it will attempt to run the code in a virtual environment to determine how it behaves. If potential malicious behavior is detected, it will be reported as a virus. Running the code in this virtual environment means that if the code is malicious it will not be able to cause damage to your computer".


5 Journeyman

 • 

5.8K Posts

 • 

17.3K Points

June 9th, 2010 19:00

Good points, ky.

On-demand detection is perhaps a less important indicator of an AV's overall protection. I'm more interested in prevention, than in detection. AV-C indicated in its latest newsletter that new dynamic tests are forthcoming.

And in fairness it should be noted these tests were actually conducted in early February/2010, and the analysis only now just published. As I recall, avast! 5.0 had just been released in its first version back then.

Nonetheless, it's a bit disappointing to see that of  the 15 AVs undergoing this test who were all last tested  6 months previously, fully 13 had lower proactive detection rates this time around (the average drop was about 10%  in absolute score).

I'm glad to see that MSE did not fall in this group, and actually improved a bit.

But the moral here seems to be: don't rely on your on-demand scans to protect you, and don't choose an AV solely on tests like these.

No Events found!

Top