AV-Comparatives June (May 2007) Results (Retrospective / Proactive Tests)

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

e.g. f-secure, symantec, ~mcafee, bitdefender, avk, etc.
do not mix proactive DETECTION with proactive PROTECTION.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

give it a chance chris.
i think its a very good result since the heristics are still young.
just think what nod32 heristics where like back in the same position.
lodore

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

sure, but just 4% more than my doctor, but thats 4.33 which the heuristic engine is way over a year old.

i think this is the NEW kaspersky heuristics, it should be better than it is, thats all.

but at least they have the PDM to sort anything out, a safety net for sure

im not saying they are bad lodore, dont read it wrong..... im just saying i expected better

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

i do see your point.
but i wont be mean and say ye but kaspersky got 35percent with only 1fp
lodore

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Thanks IBK, I was misunderstanding.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

yes there FP count is/has always been impressive,

this has been drwebs problem, even though i dont think its too big amount of FP's, to score the ratings it apparently is.

if drweb had them lower, even the old 4.33 would be advanced all throughout the year.

-----------------
will Drweb be introducing a feature similar to kasperskys PDM?

Yes, we are planning a lot of improvements. Not similar to any proactive defense (PDM), but something else, better. I'm talking about version 5 ofcourse.
-----------------

hmm,

i get interested more and more, whenever i speak to them, and of course the new engine and heuristics are already in use in the 4.44 beta which should be released soon i would think, been using it for months now with no 'real-bugs' to report.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I'm not too surprised. The product is still new after all, and has to mature. Also the signature updates used were from February 2 with over 264,000 records. Had it been done with today's signature updates with over 336,000, I guess we might have seen a different picture.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

no tony, as that would totally make this test pointless.

all AV's were/should have been checked with the same signature-count for that day, to make it all equal and fair.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Just a out loud reflection...

I think KAV heuristics have been over exposed to user expectation (more than 1 year talking about it and a lot of secrecy around it)...

There is not outbreaking results here but I am sure plenty of room to improve it in the near future...

Fax

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Oh I agree. I'm just saying that with today's signatures, the results might have been different, I don't know. The emulator is still quite young, and there is room for improvement I'm sure, but as has been pointed out previously, we have the benefit of having the PDM if it is enabled.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

You're just mad (& drunk again perhaps?) .................................. because they made a better heuristic in a first try.

Dr Web is not an old heuristic btw, so no need to post excuses for them, they tweak & develop it all the time and frankly they were never really better................around 30% for the heuristics and 90% at best in total % with signatures...................it's been like that for years no matter what picture some fanatics like to paint, it an oldschool AV with a serious need for a makeover, so hopefully they will soon bring out the much talked about version 5.0.

P.s. There could be a simlie or two missing.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

who is making excuses, all heuristics get tweaked as they go along, but the new drweb engine and heuristics in the beta have been in the process for 18 months, this shows how old the engine is for 4.33

i think it is YOU who is making the excuses for such a lame heuristic engine, lame is a strong word for sure, but it should be better, and you know it should, or maybe my expectations for kaspersky on this were too high, i dont know.

if it really does matter that they are tweaked as you say, ALL THE TIME....... why have kaspersky such rubbish results all the time for its heuristics,

laughable, so you can point your finger somewhere else if you dont mind



maybe enough smileys



also, yeah im gonna go get drunk, its friday... ive had an hard weeks work, and i deserve a night out with some drinking and dancing, thank you vry much

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

well I like Dr Web and have to give Kaspersky kudos for their first shot at it. We all know what they are capable of. But my number one is Eset. No one can touch them.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I, for one, am very happy with the results for KAV's new heuristics. Of course it is not as good as the likes of NOD, but given time im sure it will be just as good as any of its competitors. Remember that this is still the first release of KAV 7. The heuristics will constantly get better with time... MP1 will probably work wonders!

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I think we shouldn't put too much credence on heuristics whichever AV is used. These tests are bearing this out. As has been said before, they do have to be constantly updated and tweaked, especially if false positives want to be avoided.

Signatures are easier to create and can be distributed within minutes if necessary. Heuristics take a bit longer to fine tune. Obviously, if an AV isn't delivering as many signature updates, such as NOD32, they work more on the heuristics or some other form of proactive protection.

It's a cat and mouse situation even with heuristics.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

i never said in my above posts they were bad,

i said 'i expected better', sure this may be my high expectations of kaspersky,

but for DON to come in, .. o0(once again) .. because he has some clout with kaspersky and say all that, un-called for.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Nice results. Avira is fixing the FPs really fast and as they mentioned they are not present anymore. I think in the future they'll polish their heuristic engine even better to have fewer FPs. Regarding my own experience after couple of months of using it I had only about 4 FPs, which is not that much I think.

Congratulations for NOD32.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Changing topic, I notice that none of the AVs tested picked up on the 13 'Other OS malware' samples. I assume all vendors have this now, and are able to detect them.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

all FP's listed are fixed because we sent them to the vendors. but apparently the HEUR/Exploit.HTML false positives still occur in other files, like can be seen in this post http://forum.avira.com/thread.php?threadid=22990
so its not that now the heuristic is flawless and causes no false positives.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

None did very well with scripts either.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Completely agree, Bill

I have had this discussion with Andreas before . In the real world, the majority of users would simply stay with the default settings and never see this degree of FP's.

Maybe if he has time, in future tests, he can post the results for detection and FP's at the different heuristic settings for the relevant AV's such as KAV/AntiVir/(VBA32).

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

Thanks Andreas

Besides their reliability, NOD's heuristics have good detection of almost all malware categories (excepting other OS malware and scripts). The others have good detection of trojans and/or backdoors, but poor detection of file viruses, worms, other malware, etc. However, trojans and backdoors are the most important threats nowadays.

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

I didn't read all posts in details (sorry) but what about False Positives which Kaspersky 7 produced ... IBK , I didn't see false-positives in KAV7's test of yours . FP is important . It will be one situation with 35% detection rate and few FPs and 35% but with a lot of FPs

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

read again the report please

 

Re: AV-Comparatives June Results (Retrospective / Proactive Tests)

KAV 7 had only one FP (same as KAV 6)