AV-Comparatives (February 2009)

kaspersky 97.1 !! man its a good result but their detection rate is getting lower in every new test !! i dont know where is the problem ?? maybe they must use some pulse update like norton

 

no, they have something beter than pulse, they have a HIPS and HIPS isnt used in on-demand tests, and im not sure whether or not they make use of KIS HIPS in the proactive tests, we will have to see.

 

While I am an avid reader of Wilders I do not post very often. I must say all this talk about Avira and false positives has me laughing. A quick simple look at the numbers between Avira and Symantec seems obvious which app I want installed. There is a detection difference of 12,236 samples. That is a lot of malware in my eyes. While false positives are an unbelievable difference of 17. Come on now.

 

Quick Reply seems to be disabled ...

Anyways, first thing I said was nice site! Like the new layout/color scheme, orange and white.

Noticed something of questionable nature; on page 6 the author(s) state Symantec asked for the heuristics to be set at max, however "...it made practically no difference". I remember in the last 2008 test Norton with the highest settings reached 99% while at default only 97.9.

I see that you use KIS09. If you are happy with it, there is no reason to be disappointed by the lab tests.

 

You are right but the nature of the test has led the discussion to the false positives. As you can see are the false positives that mainly determine the rank. And even if we discuss it and why not...laugh a bit...believe me there are users that buy a product or use its free version because they see it in the top 5.

So get listed in the top 5 it is important (specially when the test comes from AV-Comparatives)...and if for example Avira wants to be in the top 5...has to reduce its false positives, even if it's about a small number.

 

avira f-secure bitdefender etc is very excellent~~~(these antivirus always very goode in any tests )
norman which i used and i loved is very poor ~in particular its script malware detection~~i don't know which updated on 19TH march.and his scan speed is always very slow not any improvement~~oh i think norman should make effort to change the sitation~~

 

Maybe I missed it but shouldn't "Hit to system performance" be part of the testing? If the AV detects 100% but jams my system then it isn't much use to me... Just a thought.

 

That would be hard to test, since everyone computer is different (different hardware, software, OS's, etc). Many AV's that people have said are too heavy seem to run just fine on my computer.

 

That's why the whole advanced/ advanced + is a bit of a joke a la "Virus bulletin". People should do some simple maths, the differences between the results of Avira and GData versus the rest can't be even compared: we are talking about thousands in terms of detection (25,000- 30,000 range of samples). Have people tried to divide say in Avira's case 24/1,274,928 and Symantec 7/1,274,928 and compare the percentages in terms of FPs?

 

So they dumped Kaspersky 6 engine The old engine was really making their detection mediocre.

Kudos for Eset, IMO They have been doing well nearly every time, no matter what the sample set or the FP rules. That consistency is hard to match.

 

They are a consideration if you have any of the packages installed prior to the false detections being fixed. As I've said before, I hardly know a lot of those packages, let alone have them installed.

Users who have not had many or no FPs is likely due to (a) the heuristic being set at medium by default and not being changed to high and/or (b) not having any of the packages installed even when heuristic is set to high.

I note that among the AVIRA FPs are packages relating to DrWeb and Trend Micro; my thought is if you're using AVIRA in the first instance, why would you have either of those two on your system anyway? It's good to know the FPs for these has been fixed, but in the real world scenario I doubt many novice users would have these or indeed several of those other packages installed at the same time.

 

FPs notwithstanding, Avira did swell.

 

As I have stated many times here- my only concern is detection. If some are willing to trade hundreds/thousands of non-detected malware as a trade-off for fewer FPs- so be it. But not for me.

 

Agree 100%

 

Something strange: F-Secure has significantly lower detection rates compared to Kaspersky and they were using the same engine....

 

i know and its great and giving such a good result on my test ، but you know its also something great to have such a good detection !! anyway the days will tell us

 

did you know the effect of FP for ordinary joe?....or let us say the country like china?

 

and do u also know the effect of a serious infection that could have been missed because the AV strived to have 0 FP's?

 

sorry i didn't get serious infection yet....let AV-Comparative result do the talking.

 

I agree, especially G Data has too many FP! Sometimes a FP (winlogon.exe, svchost.exe) is more dangerous than malware

 

Everybody has FPs, but the results of this test don't say much, just that the product X caught Y fps on the AVC set. But you can't make any assumptions of out it (better said: you should not). That's why I think the penalization is wrong here, because attention span of most 'readers' doesn't go much beyond 'bright green'

 

F-Secure using the old Kaspersky 6 engine also has its own huerstic thats why..

 

I've been satisfied with Avira.. false positives have been fewer lately on my computer at least. Does anyone have any explanation for so many AVs surpassing Kaspersky? Even Kaspersky's trojan detection, which was #1 for a long time, is now pretty average in this test.

 

Hi

F-Secure is currently using the V.7. Take a look at this:
http://forum.f-secure.com/topic.asp?TOPIC_ID=8397

Best regards,
Nightwatch

 

even if its using the version 7 engine doesnt mean its using the kaspersky heristics. its probaly just engine plus sigs that are supplied by KL.

btw doint base f-secure's proactive detection based on this.
If the files are excuted deepguard 2 will catch pretty much any infection.