AV-Test Product Review and Certification Report - 2010/Q2

Something's very fishy with this test,Avast in all of these tests is among the lightest av's and only this one claim the opposite? and based on what? did they mention the tools they used to test the impact on systems performance?

~ Removed Direct PDF Link as per AV-Comparatives Request - See Performance Tests page for the actual PDF ~

http://www.maximumpc.com/article/features/kill?page=0,2

http://www.raymond.cc/antivirus/

 

How can you even compare AV-Test.org with Raymond.cc ?!

And they don't need to mention the tools in each and every report , this is AV-Test.org - AFAIK the biggest testing organisation in the AV industry.

 

Agreed. A sample set of 16 seems very low.
Does anyone know for how much the Dynamic Detection Testing-part counts in the total score?

Not so fishy.
The Avast security suite was tested, not the AV.

 

I too am surprised that only 16 samples were used in the “blocking of malware on or post execution” portion of the test. I wonder: what was the rationale for AV-Test to use so few?

The “real-world testing” component is based upon 59 samples. However, over about a 90-day duration (April-June, 2010), I would have expected the test to encounter a greater quantity of zero-day malware attacks.

 

Okay i take back what i said about this test being fishy

This test is the laughingstock of the tests EVER

 

I totally agree with you. For me the surprise was the result of bitdefender
Bitdefender is no way lighter than avast or avira

 

Is that your opinion or do you have something to back that up? Sorry if I'm being hard, but again typical Wilders behavior has shown itself. One minute a test is looked at with interest and congratulations are spread around. Then, inevitably, someones favorite vendor doesn't do so hot, the test starts to have "something wrong with it", and from there on, reason after reason comes along from others to dismiss the entire thing. And no, I'm not saying that because I have a favorite that did bad or good, I say that because this always, never without fail happens. Is 16 samples enough? Maybe not, but that doesn't automatically mean the testers are complete idiots.

 

Overall, I think the test is pretty good. Symantec and Kaspersky scored at the top, which shouldn't be a surprise. I don't know enough about Panda to comment about it. Historically, Eset has never done that well with AV-test.
A big surprise to me is the high performance of AVG. The only part that seems unreliable is the average system slowdown. This is only one test, and should be treated as such.

 

Perhaps there is some confusion here and there as some seem to think that only AV's have been tested where AV-Test.org has tested complete security suites (besides MSE and SpywareDoctor+AV) afaict.

I don't understand why they would test the usability/slowdown of different products.
Suites, AV's and combinations of AM/AV products are likely to have a different impact on a system.

A test like this seems to imply that AV-Test.org assumes that those different products should, in principle, have the same resource usage.

I don't think I could agree with such an assumption.

 

Just to add for those of y'all wondering why AVG scored so high remember the suite and pro version has Identity Protection = Behavior Protection = The old Norton Antibot and from reading about tests AVG has really kept it a strong tool.

 

Don't jump to conclusions so easily,on each of the three pc's (2 desktops 1 laptop) i have at home i have installed 3 different av's so the part you said about favorite vendors is totally wrong,at least for me.

Now on the other you said,nobody said the testers are idiots but this doesn't mean we have to ignore some real facts just because they happen to be the "biggest testing organisation in the AV industry"

 

There is no AVG pro. And only the IS suite has IDP.
http://www.avg.com/gb-en/free-antivirus
Free and AV can add IDP for $20. Which if you think about it Free and IDP looks like a pretty good deal for really good protection.

 

Spain - Holland : 1 - 0

I posted that suggestion long time ago at PrevX, wait and see

PrevX4 - Panda2 : 1 - 0

 

Oh my mistake I thought AVG had a PRO Av that had the IDP module.

 

I think when v10 comes out in a few months, the packages are supposed to change.

 

Bullguard use the Bitdefender scanner, and should have the same detection

 

Spain - Holland : 1 - 0

http://doc.noticias24.com/1007/patada630rg11.jpg

 

I ranked them based on AV-Test's protection score, with 0 equal to worst and 6 equal to best.

Product - (Protection Score)
AVG: Internet Security 9.0 - (5.5)
G Data: Internet Security 2010/2011 - (5.5)
Panda: Internet Security 2010 - (5.5)
Symantec: Norton Internet Security 2010 - (5.5)
F-Secure: Internet Security 2010 - (5)
Kaspersky: Internet Security 2010 - (5)
McAfee: Internet Security 2010 - (5)
PC Tools: Spyware Doctor with AntiVirus 7.0 - (5)
BitDefender: Internet Security Suite 2010 - (4.5)
Avira: Premium Security Suite 10.0 - (4)
Microsoft: Security Essentials 1.0 - (4)
Avast: Internet Security 5.0 - (3.5)
BullGuard: Internet Security 9.0 - (3.5)
Eset: Smart Security 4.0 - (3.5)
Webroot: Internet Security Essentials 6.1 - (3.5)
Trend Micro: Internet Security Pro 2010 - (2.5)
Norman: Security Suite 8.0 - (2)

Why ranked on protection? As others have stated, I'm more interested that it prevent infection and there are good, free tools for removing malware.

 

Yep, I have seen them loose of Germany, Argentina and now Spain. I corrected the post.

Have you seen how the Germans were beaten like sheep. Spain has the most capital (hence most talent) in their team. You have to show spirit: see http://www.youtube.com/watch?v=WO4tIrjBDkk

 

Very inspirational speech but a bit aggressive for football maybe jeje

 

I haven't formed my opinion on this new testing, but so far and from the consumer point-of-view, this one seems to be the more important certification the indrusty is currently issuing, and i'm thinking of competitors like VB100, WCL and ICSA. Probably more re$ource$ are needed to increase the statistical relevancy of the "Blocking of Malware on or post Execution" test (the same can be said of the latest Dynamic Testing from AV-C), and i also wish they were more transparent regarding methodology used. Looking forward for follow-up.

 

last night I took 16 samples and tried them on product A. It got 15 of them, and would have recieved a A rating from me.

Then I tried another 16 with product A and it only got 4 of them. In this test a D rating.

How quickly one can rise or fall depending on how narrow the scope is.

 

The extremely low number of samples does seem to fly in the face of reason,with such an amount there's a huge possibility of wide variance in results.

 

For the 0 day testing they used 59 samples. I think that is reasonable considering how new the samples have to be, and the time frame to test in.

I think this is one of the better tests I have seen done. I like the way the results are laid out, and the fact there is a system performance impact measurement.

I just wish more products were included.

 

About the 0-day protection testing,
I think that it would be better to test 4-5 sets of 0day samples and then take the average mean of each product score.
Most of the products, would detect some more samples after the second signature update, the same update, so it's a matter of some hours for one product to score 10/16 or 15/16.