AV-Comparatives February 2007

There's no problem here, Sjoeii. All systems are
I'm using the latest official 6.0.2.614 build.

 

Several ot posts split out of this thread and placed into a separate thread for further discussion\target practice by the Wilders Security Forums Staff

Please carry on with the AV-Comparatives February 2007 discussion

 

3 non surprises to me.

Microsoft - Need not say more

Norton - In real world use, doesn't approach it's detection rates achieved in these tests

NOD32 - Relies too much on it's Heuristics (slow to add malware samples)

Again, Thank You IBK for your efforts!

 

I agree with you. I also believe Dr. Web may not score well on these "standardized" test but does provide good real world protection. Dr. Web updates 4-6 times a day and a new engine should be coming out of beta soon. The new engine has improved heuristics. The latest beta key is due to expire this week and hopefully the new engine will be pushed out via auto update soon afterward to the masses.

 

Interesting results, I suppose we'll have to see the full report/results before we can deduce if the surprises are caused by the removal of DOS malware from the test set.

Londonbeat

 

Out of curiosity, why do you think that an AV might not score well on the test, but provide good real world protection?
It would seem to me that if one could not detect a virus/malware on the test machines it would not be any better in the real world.

I admit to not much knowledge regarding computers. However, I have been very involved at various times in the tests of some very complicated missile systems. I can assure you that if one did not perform well on the test range, it would do much worse in the real world. The test range is an easier set of circumstances than the real world.
Maybe you might conclude that the lab tests of AVs are more difficult than in the real world.

Regards,
Jerry

 

I have to disagree somewhat, tobacco.

Advanced+ status reached 7 consecutive testing periods......

and the first time since February of 2005 (two years ago) that it did not reach Advanced+ (simply "Advanced").....

and this falls into your "non-surprise" category

Personally, I'm of the belief that perhaps Eset's attention and concentration has been more spread out more recently than at any time in the past. When you consider the new Anti-Stealth/Rootkit detection that was implemented in version 2.7, along with work on the "Auto-Component Upgrade" installer....AND the firewall, spam filter, etc. that will be included in the Eset Security Suite....and that probably has as much (or more) to do with NOD failing to reach Advanced+ for the first time in 2 years as anything else. But I've been hearing the "slow to add definitions" talk for quite some time....even when NOD was consistently reaching Advanced+ testing status......

 

So do I! I've always said it is the long-time standard. It has been solid for a long time now, and not just for one year...like some overrated ones I won't mention, ahem...NO...no, no...I said I wouldn't mention it. Great job Kaspersky, you've succeeded to continue to make KAV solid and now light as a feather! Congrats also to Avira! IMHO, Kaspersky's only true competition!
But, hey!...competition is what keeps vendors on their toes...better for us all!

 

i feel there will be a few surprises in this on-demand test, but never count the scores before they are released, as you just... never know.

 

Hello

Please don't misunderstand me as i do feel NOD is one of the top tier AV's. And this is just my observation as i follow many different categories here. Yes, they have been busy but its just my opinion that they should accept their Heuristics as a 'bonus' and update sigs more frequently. Again, just my opinion.

 

as long as it gets the threat, it dont matter......

but i do agree, that nod32 is mainly heuristic-based and maybe should add more sigs, if their heurstics were not excellent, i dread to think what their detection would be.

 

Aren't the results of the August 2006 tests indicative of the detection rate without heuristics? NOD scored 98,61% on that test.

Best,
Jerry

 

heuristics are used in the on-demand tests. if no heuristics were used the only AV that would do any good would be kaspersky.

 

That definitely sounds like a fanboy statement that could not possibly be backed up by any facts.

 

There's the fact Kaspersky has awesome detection rates.

 

Not denying that.

But is it a "fact" that...

"if no heuristics were used the only AV that would do 'any good' would be kaspersky"?

Nope. Definite "fanboy" stuff....

 

Seeing as though his signature links to antivir's website, I'll let you think about that one , but to tell you the truth, I think SAV would stay pretty much the exact same, and same goes for a few other avs.

Anyways, I have been pretty impressed with antivir's rise and continual commitment to keeping amoung the elite, so the crew there definetly deserves a round of applause

Also, kaspersky's consistence is pretty darn good, too bad it doesn't like my CPU

However, the results just go to show that no av is perfect, and that an av isn't the perfect solution. It can't been relied on as a cure all, whether its kav, dr. web, clamav, antivir, nod, etc.

Cheers,

Alphalutra1

 

I guess it was extremely exaggerated, but I didnt know what to say to prove that heuristics are used in the on-demand test. And i am not a KAV fanboy. I prefer Avira though I do think with no heuristics were used kaspersky would be number 1 in all on-demand tests and that was how i pathetically tried to prove heuristics must be used in the on-demand tests.

 

Wouldnt it be neat to change the ranking system? Standard 85%-89.99%
Standard+ 90%-93.99% advanced 94%-96.99% Advanced+ 97%-100% or if no AV gets 99% on the february on-demand test Advanced+ 97%-98.99% Stellar 99%-100%

 

IBK, can you tell the exact percentage detected by AVG Pro?

 

Does anyone know if the AVG Internet Security Suite would score the same as AVG Anti-Malware? It has the Antispyware Component in it, but I'm a little confused from some of the posts here about the differences between AVG Anti-Malware, AVG Anti-Spyware, and AVG Pro. Also, since the Antivirus uses Heuristics for RealTime Detection of Viruses, does the Antispyware also use Heuristics for RealTime Detection of Spyware? Thanks for any help with this.

 

I admit I cannot argue the point of how heuristics impacts the on-demand tests, however, the bottom line is the results. How much is heuristics, and how much signatures I don't really care. If it runs well, and has a high detection rate it seems unreasonable to be critical of the reasons.

I also think the KAV vs NOD argument is relatively useless in the overall scheme of things. Both have a proven record in tests, and in the experience of users. Both and some others have 1st class records, and in my case it boils down to which one runs best on my system.
I like KAV best of any AV, except it has not run well on my system lately, while NOD does. That makes the choice obvious at this present time. Who knows what will happen in a month or so?

Considering IBK's tests, I would not want to go to the Standard level, but I really believe that I would be as safe with Avast as KAV. Yet, some swear by certain AVs at the Standard level. Evidently their experiences have been satisfactory overall.

Regards,
Jerry

 

This is somehow true, KAV heuristics are really poor, it looks that they focused on fast-updates for new threats more than any other AV, now with PDM it is better, also we all are still waiting for the new heuristics.

 

What AVG IS has over AVG AM are the Firewall and Anti-Spam. The rest are the same.

 

I thought so, but I guess the Ewido (AVG Anti-Spyware) addition gives AVG Anti-Malware, and the ISS more protection against trojans and the like, right? I still wonder though if the Heuristics used for Virus detection is also used for RealTime Spyware detection?