Latest AV-Comparatives test; Antivirus proactive on-demand results.

In their latest testing round, AV-Comparatives have focused on proactive on-demand testing.
This means that antivirus products have been tested solely on their generic heuristic scanning; how well does an AV program do if it isn't updated for a month?

The final result is a combined score for detection rate and false positive rate; an AV program that has the highest detection rate but a high number of false positives also, gets a lower final score.
The top-10 detection scores over a 4 week period are;
1 Avira 67%
2 Kaspersky 60%
3 GDATA 59%
4 NOD32 51%
5 Sophos 50%
6 TrustPort 49%
7 BitDefender 46%
8 Microsoft, Symantec 44%
9 AVG 40%
10 Avast 39%

In the final score, where the number of false positives is taken into account, the winner is NOD32 with the highest rating Advanced+.

Final AV-Comparatives scores;
Advanced+; NOD32
Advanced; Avira, Kaspersky, Microsoft, Symantec, McAfee, GDATA
Standard; Trustport, BitDefender, AVG, Avast, Norman, VBA32

The scoring table can be found in the comparatives section on this page here, the full test results here; PDF link.
Cheers.

Edit;
Keep in mind that these scores are not representative for 'real-life' situations, they only show the heuristic engine capability of an AV program.
The virus signatures/daily updates and the heuristic engines combined are what makes an AV program effective.
Please also take into account that AV-Comparatives judges quite arbitrary about the number of false positives produced.
Just a few more FP's than a competing AV program can result in a lower final score.

The scores from the AV-Comparatives test from last August has 'real-life' scores; PDF link

 

Thnx.Would be useful.

 

No wonder they give Avast away for free. Maybe I need to get something else.

 

Avast is actually very good.

It updates almost daily so when it doesn't update for a month it is bound to cause problems.

 

But the scores don't agree with you.

 

Try Avira. Also free.

 

So whos the best here? NOD32?

 

Well, with all tests like these it's pretty important to read the results and testing methodology closely.
For instance, the final result has been influenced a lot by the number of false positives produced.
AV-Comparatives decided rather arbitrary about the (amount of) FP's classification.
For instance 14 FP's was categorized as ' few' while 17 FP's were considered ' many'.
So Avira, which scored better in detection than NOD32, got a lower classification (Advanced) than NOD32 (Advanced+) because of more FP's.

I would rather have superior detection and a bit more FP's (e.g. Avira) than a lesser detection rate with fewer FP's (e.g NOD32).
So I will still say; Avira (free or paid) for the win...

 

avira+threatfire is pretty good

 

Thanks for posting this.

 

baserk i do not believe he allows use of or linking to that site's results or pdf's without permission..you can link to the main page though then they can click the links from there.

 

Mmmm, you're right.
I've mailed AV-Comparatives to see if they mind, if so, I'll remove the links in post one.
Cheers.

 

I got an answer from Andreas Clementi.
He points out that he does mind links to the 'online results'.
While those end scores point out the winner, they don't offer any information on testing methodology, number of samples used, i.e. all the data on which the final scores are based upon.
The PDF's have all this information and linking to them is allowed so I will only link and refer to them and to the home page.
Cheers.