AV-Comparatives.org Releases New Retrospective Test Results

[Dieselman]

Austrian security test lab AV-Comparatives.org puts antivirus products through various kinds of testing on a regular basis. Twice a year they run a on-demand static test that checks how well products recognize current malware samples, and twice a year they run what they call a "retrospective" test that evaluates how well products manage to detect malware for which no specific signature is available. In the just-released retrospective test, all tested products rated either ADVANCED+ (the highest rating) or ADVANCED.

http://securitywatch.pcmag.com/none/290841-av-comparatives-org-releases-new-retrospective-test-results

Notice the bottom paragraph. Now lets all do the Peter Panda Dance.

One standout among the results was Panda Cloud Anti-Virus . Despite its small size and cloud-based detection, this product managed an ADVANCED rating.

 

[NSG001]

Thanks for the link, surprising results.

 

[Jack]

One standout among the results was Panda Cloud Anti-Virus . Despite its small size and cloud-based detection, this product managed an ADVANCED rating.

Simple and pure advertising here...Wonder how the guys from PCMAG manage to get to this conclusion because Panda in fact performed really badly in the the latest AV-C Retrospective Test(November 2011).Not only they didn't manage to the A+ certification but they've also manage to finish last in the detection of new malware test with a really shameful 41,4% detection rate.

Here are the main spotlights from AV-C Retrospective Test (November 2011) :

ProActive detection of new malware :

1. Qihoo – 67.6% *,
2. G Data – 64%,
3. Avira – 62.4%
4. ESET - 61.6%
5. Trustport : 61.3% *
6. Kaspersky : 60.1%
7. F-Secure : 57,5%
8. Bitdefender : 57,2%
9. eScan : 56,9% *
10. Microsoft : 48,7%
11. Avast: 46,1%
12. Panda : 41,4%
    * = Product with 'many' false alarms (FP)

Certification levels reached in this test :
Advanced+ : G Data, Avira, ESET, Kaspersky. F-Secure, BitDefender
Advanced : Qihoo, TrustPort, eScan, Microsoft, Avast, Panda

Source : AV-C Retrospective Test (November 2011)

Note to self : Never trust or read PCMag again.

 

[NSG001]

Are Qihoo still using bitdefender's engine ?


Note to self : Don't believe all you read

 

[Jack]

Are Qihoo still using bitdefender's engine ?

Yes , as far as I know they still use the Bitdefender engine but they've had a really high False Positive ratio so that high detection rate comes with a price....That's why they didn't get the A+ rating.

 

[Dieselman]

Your forgetting something Jack. Panda is a cloud av. But if you look on other AVC tests Panada Cloud scored 99.3%. Also Jack read the chart they made. They include more then just AVC. Symantec was not included cause without Sonar its fairly useless.

 

[Jack]

Your forgetting something Jack. Panda is a cloud av. But if you look on other AVC tests Panada Cloud scored 99.3%. Also Jack read the chart they made.They include more then just AVC.

Don't know about which exact test are you talking but most likely they scored 99,3% in a traditional on-demand test that didn't involved only zero day malware and where all the other vendors got a detection rate above 99%.......From my point of view only Symantec with their Insight technology is really using the clouds while Panda is taking only a traditional approach which isn't something that can bring a high level of protection as proven in this test.

 

[Dieselman]

Lol. I love debates such as this. Take away Sonar and Norton is useless. Take away the internet from Panda or Norton and its useless. But explain to me when your ever in a situation where you are not connected. If you not connected then you do not need any protection. Avast is not a cloud and it scored low also. But this is only ONE test. One.

So I guess PCMag is also wrong when they give an Editors Choice to NIS?

 

[Jack]

Lol. I love debates such as this. Take away Sonar and Norton is useless. Take away the internet from Panda or Norton and its useless.

You didn't understood me , I didn't say that clouds were good or bad . My point was that Panda has a very simple idea on what the clouds can do and I gave you as example Symantec who is using their clouds to create reputations from each program and analyzing in real time in there clouds the programs.

But explain to me when your ever in a situation where you are not connected. If you not connected then you do not need any protection.

If lets say a new and unknown malware cuts down your internet connection....

Avast is not a cloud and it scored low also. But this is only ONE test. One.

Yes Avast didn't not performed really well , but you didn't see PCMAG praising them right? And yes this is just one test.........

 

[Dieselman]

Ok. So here is the situation. Malware would have to get passed Panda first. If that happened, which it can cause no av is perfect that malware would then have to disable my NIC. But what your missing is that Panda keeps a temp file for such situations. I do not need a video either Jack. I have seen infected pc's with malware running and no internet connection. I have seen Norton,Kaspersky,McAfee,AVG,Avast and Trend all fail. So does a traditional av have an advantage? No. In your scenario malware would first have to passed Panda just like it would get passed any other av. If you have no internet connection and your av cannot update then your just as vulnerable to malware as the cloud is. Languy99 proved that NIS is helpless when the internet connection is severed. So whats the difference? If malware gets passed your av then its already too late.No biggie. Rkill it and the MBAM it. Its that easy.

If your internet is down then what harm is there? Most malware just redirects you to there proxy and never completely disables your internet connection. I have seen rogue infections that have been there for months. Unless you pay them then there is nothing to steal. Panda Clouds real time tests prove that you you do not need a traditional av to stay safe. If a cloud av was not as good as a traditional av then explain why Panda Cloud has better overall protection then CIS. And CIS is a full suite. BTW......I wasn't comparing CIS vs.Panda. It was just for references. If a traditional av is disable by malware then it and does not have the latest definitions to combat new malware then how is it any different then a cloud av? Its not.

Here is another scenario. Suppose Norton,Kaspersky,Avast,Avira or AVG have not been been updated for about 2 hours or so. Some new malware gets released during that time. Now all 5 of those av's I mentioned are no different then Panda.

 

[Dieselman]

This is the first Retrospective test where Panda Cloud Antivirus participates. Unlike Panda Antivirus Pro 2012/2011 which participated in the previous Retrospective tests and which achieved #2 rankings in two of the last three tests, Panda Cloud Antivirus has parts of its heuristics from the cloud. The 2012/2011/etc Panda products have all its heuristics locally. So therefore we think the difference might be in the fact that part of the heuristics of Panda Cloud Antivirus were not allowed to run in this test (as it is performed offline). AV-Comparatives said they would clarify some points about PCAV and Retrospective test in the PDF but I haven't seen it. Maybe they forgot.

Also it is a little strange or suspicious but this is the only independent test I know of which doesn't provide missed samples, so there's no way to verify if what was missed was actually malware or not. In all other AVC tests before they are published vendors have a chance of reviewing the missed files and usually (more often than not) some are discarded because they are not malicious. The reason given is that those same samples will be reused in the next test, but this doesn't make a lot of sense to me as there are plenty other new malware coming out every day to make up the testbed for a test which is still 3 months from now.

Taking the above limitations into perspective we need to figure out if we will continue participating in AV-Comparatives Retrospective tests.

http://www.wilderssecurity.com/showpost.php?p=1975212&postcount=14

 

[Jack]

So by looking at this reply , I can assume that Panda admited the fact that they've failed pretty badly this test right?And PC Mag basically just give them free advertise for something that they didn't deserve..... :shy:
And a message for the Panda Team : Stop trying to explain why you fail and accuse AV-C of bad conduct, you knew the rules before you agreed to take part in this test....Just look at the results and try to improve the product....

 

[Dieselman]

Do you honestly believe that someone from the Panda Team is going to read your reply? Panda only scored low do to no internet connection. Thats how a cloud av works. So I guess Symantec also gets free advertising from PCMag cause every version of Norton gets an editors choice. A retrospective test is not real life.

 

[Jack]

Panda only scored low do to no internet connection. Thats how a cloud av works.

Then maybe the Panda developers should put a big warning on the site saying : We cannot properly protect you in case of 'internet connection' OR NO JOINING THIS TESTS!

A retrospective test is not real life.

Yes, in real life I can tell you that Avast Free or NAV would perform better in any test when compared with Panda Cloud AV Free.. I know that you use this product but believe it or not it's not the best...

 

[Dieselman]

No where did I say Panda was the best so PLEASE do not put words into my mouth. I love how people jump on the bandwagon of failure when they see a test like this. Not understanding how a cloud av works results in your replies. Norton is helpless without Sonar and Sonar is helpless without an internet connection. Please read my previous replies which clearly state different scenarios.

 

[Deleted member 178]

I love how people jump on the bandwagon of failure when they see a test like this.

Same with CIS, it has good results somewhere, they buy it; it failed, it is normal...

 

[Jack]

Ok lets not completely go off-topic in this thread.....Really don't want to start to talk again about CIS if it didn't make it to this test...


We have this results :

ProActive detection of new malware :

1. Qihoo – 67.6% *,
2. G Data – 64%,
3. Avira – 62.4%
4. ESET - 61.6%
5. Trustport : 61.3% *
6. Kaspersky : 60.1%
7. F-Secure : 57,5%
8. Bitdefender : 57,2%
9. eScan : 56,9% *
10. Microsoft : 48,7%
11. Avast: 46,1%
12. Panda : 41,4%
    * = Product with 'many' false alarms (FP)

If other members would like to comment on the subject , I'm waiting to read their opinions...
Avira seems to be doing just fine ..... They still got a very good heuristic engine and it's nice to see that the FP isn't a problem for them.

 

[Deleted member 178]

Qihoo is a surprise, Avira seems to take back the road to effectiveness. We are all surprised by Avast on this test.

 

[Jack]

We are all surprised by Avast on this test.

Yes, MSE manage to place above them with almost 3% detected malware...... That should be a strong sign for the Avast team that they really need to improve their behavior blocker...

 

[Dieselman]

Oh no..............Panda scored horribly. I better hurry up and jump on the CIS band wagon. Not. If a cloud av was so bad then how does Panda score extremely high in other tests? Explain Avast's score. One test and people are mocking Panda. When languy99 showed NIS failing did everyone stop using NIS?