New results of Virusinfo.info for September-October 2008

Here is the link to the results of Virusinfo.info for September-October 2008. Everything looks more or less logical, decent and fair.

http://virusinfo.info/images/vttesting/sep-okt.PNG

 

The results are pretty much realistic, still they're not as reliable as tests done with the full antivirii. That is because the VirusTotal engines really miss features that the real scanners have.

BTW: Very good results by Avast.

 

Not a single heuristic detection for either McAfee or DrWeb, what happend here?

 

Wow, F-Secure is doing great

 

It would make it more valid if they showed what version of each software they were using. to. Oooooppps, the are, I just saw it, Prevx1, yes they are real up to date.

 

Pretty colors but as far as being a definitive test, "not happening"

 

do not take it as critics, but I think it was often posted why misusing virustotal to compare detection rates is wrong. The commandline scanners used at VirusTotal are run with different settings: some vendors want that their product runs at virustotal with lowest detection settings while other vendors want their product being run with the most paranoid settings. this is one reason why using virustotal to compare the detection rates of products based on a ~40 files test-set is not very meaningful.

 

IBK, txs for clarification

Regards,

Smokey

 

Actually they have labeled it Prevx1, but it seems the version of engine is v2.
Regardless, many other products are not latest, like Symantec,Kaspersky.

 

Yep, very good results by avast. It might be better than kaspersky by version 5 since its self defence and packer support is already just as good.
Despite AVG's three giant fp's in 1 month it still is up there with avast and avira.

 

Finally some good news for AVG at this time of great disappointment .

 

I'm surprised McAfee did so well. But of course, one could question the method of testing, and the 'real-world' effectiveness of the antiviruses.

 

Avira & Ikarus superb, as always. Twister not tested (sigh).

 

poor test, poor samples, poor methods, poor presentation.

why are people really bothering with such results?

im sure the guy is a great guy, but this test has many flaws that i can see, and im sure more flaws the experts can.

this weekend, ive sent over 200 individual samples to Prevx, if i ran these through VirusTotal and made a graph, does this make me a professional tester?

...highly doubtfull.

infact, that would make around 5 of these tests alone.

 

Yep
i agree

 

I think they make quite clear that you shouldn't swear by their results.... it is just samples they have collected through their "live" infection cleaning, and obviously should be used in conjunction with many other tests etc to come to a reasonable conclusion.


edit: should=shouldn't

 

Kaspersky & Symantec have ZERO heuristic malware detection....GREAT.....keep up the good work.

 

Data are data are data are data. You might say the burden is on the USER of data as to whether to use, mis-use, or ignore.

To regard the data as gospel is foolish. To regard them as totally meaningless is equally foolish (IMO) -- they are what they are, & what they are is clearly stated. I view the data with curiosity & say "Hmmm... how about them apples?" -- I see nothing wrong with that.

 

sure, your allowed to post your opinion.

.. as ive done

 

Question,if I may?How does Dr.Web clean well yet detect only fair?How can it clean what it can't detect?

 

Using KAV 7.0.0.125.... stay tuned for a very nice big juicy update before the end of the year to the heuristics of the current v2009. It will blow your socks off

 

^^^^^ nice cant wait to see what they are doing. i had issues with the current beta so i uninstalled it and went back to the released version for my office pc's...

 

Please Please ... Make it available for Vista 32/64 too

 

Antivir, and Avg.

 

This test is BOGUS! It doesn't run the malware, and hence is not test behavioral heuristics.