AV-Comparatives: Real-World Protection Test - October 2017

[roger_m]

Is Tencent worth the spyware?

Well, I've never seen any actual proof of spyware, only claims made with no proof to back them up. It's also worth noting that while Softpedia removed the download link for PC Manager in April, they have restored the download link since then.

Softpedia has removed the download of Tencent PC Manager from their website with the following warning:

"WARNING: Download links for Tencent PC Manager have been removed temporarily due to suspicious activity (after installation, the application attempts to download some files which are detected and blocked by antivirus products, including Tencent itself)."

 

[Deleted member 178]

Emsisoft did quite well , zero FPs.

 

[HarborFront]

As a matter of fact, ESET has never shined because of its default settings (settings independent testers mostly use). ESET has never shined either by detecting threats in the moment, despite they update their databases pretty quick.
As an ESET user, i can affirm that anybody that has medium knowledge on the field, and understands what informatic keywords mean, can tweak ESET well enough to outplay almost any of the above software. ESET is no miracle nor is perfect. It's not the best out there but it's a solid product if handed carefully and wisely. For example, my ESET uses a lockdown configuration, warning me about every single inbound and otubound connection, and about every actions taking place on my system (HIPS interactive mode), blocking everything by default. Adding lots of network protection options and an amazing PUP blocker, i only need VoodooShield just in case and i have a nice fortress on my system.

Are you using the ESET IS? If yes, it may have a weakness based on an earlier version of Smart Security tested below. I'm not sure these have been rectified in the latest version

Analysis of ESET Smart Security 6 personal firewall’s thresholds and …

Unless someone carries out such tests again on the latest ESET IS version otherwise have to assume that the weakness remains

 

[amico81]

mmh the samples must be old.....Panda rocks in their test better than emsisoft or kaspersky. That cannot be normal.
Panda has mediocre signatures and bad behavior blocker against zero day and no glorious ransomware-protection, but seems to be the new star @av-c
And then the big difference from panda @ our malwarehub... I'm sure not a panda hater or fanboy of any av...but the test results should be more transparent from official av labs.

 

[Nightwalker]

The reason that Panda, Trend Micro and Bitdefender (and rebrands like VIPRE) excel in this particular test ("real world") is that they have aggressive URL filters, without this module the results would be much lower.

Personally I have much more trust in Avast, Kaspersky, ESET and Emsisoft even if they results arent stellar in a particular test; they have much more contributions, researchs and papers about malware than the rest.

I am almost sure that Kaspersky/Emsisoft/ESET can protect a machine much better than Panda for example, they have much better components/heuristics and signatures.

 

[Windows_Security]

Personally I have much more trust in Avast, Kaspersky, ESET and Emsisoft even if they results arent stellar in a particular test; they have much more contributions, researchs and papers about malware than the rest..

Bitdefender, ESET and Kaspersky are the AV-vendors (I know off) with two blogs. You can see/check which AV's are most active in blogging and research by looking at Malware in the media: Bad Rabbit ransomware and Wifi-Krack vulnerability - AV-Comparatives Weblog Most are mentioned there.

 

[ForgottenSeer 55474]

Hey, Bitdefender is performing well as I expected

 

[zzz00m]

Hey, Bitdefender is performing well as I expected

I use Bitdefender Free on one of my machines, and it has worked well without any issues. I would seriously consider Bitdefender Internet Security as an alternative to my current paid AV on my main PC.

 

[ForgottenSeer 67480]

I used Bitdefender just like Bullguard after him and both antiviruses made me dissapointed because they's detecting drivers from CD'S and installers of the games as a virus, for me: Emsisoft is much better choice because of no False Positives and much better knowledge about protection user, performance, stability and needs of user.

 

[razorfancy]

Remember this interactive chart is the sum up of various months in one chart.

October's chart is:

Personally I give zero credibility to these type of tests, a test where are multiple AVs in multiple times getting 100% clean systems seems so fake to me, for example Vipre, we can just check the malware tests done here(on the Malware Samples topic) and keeps getting infected...

 

[Nightwalker]

eset doing a regular work with unknow files
less smart than panda and symantec. but it keeps better privacy. that comparison has not been added.

About ESET results:

There were few CoinMiners in the test. These had been detected as potentially unsafe applications which are disabled by default and users enable detection at their discretion. AVC performs tests with default settings which is why these CoinMiners were "missed". After the test we reclassified them to trojans.

Dont worry, ESET protection is better than Panda and Symantec and with low false positives rate.

Symantec these days is relying too much on machine learning and file reputation, their products by default actually dont have legacy malware detection anymore and almost anything that isnt whitelisted is detected as malware.

Panda? I want to see those results without the URL filter module, change the infection vector and lets see how good Panda heuristics and behavior blocker actually are.

 

[RoboMan]

Personally I give zero credibility to these type of tests, a test where are multiple AVs in multiple times getting 100% clean systems seems so fake to me, for example Vipre, we can just check the malware tests done here(on the Malware Samples topic) and keeps getting infected...

Remember 100% does not mean no virus pass through certain antivirus, but all of the used files were detected. This only means X antivirus protected against the selected malware but can and will certainly fail sometimes. And you do good in not trusting these tests. Always take with a grain of salt.

 

[Windows_Security]

eset doing a regular work with unknow files
less smart than panda and symantec. but it keeps better privacy. that comparison has not been added.

I remember AV-Comparatives did one a few years ago. But that triggered fierce reactions of the AV-vendors participating in the test without asking. So they seem to have dropped that test. But that is logical when you have a good look at the AV-testing business model:

1. AV-testing model is for HOME users, but is not like the test models of CONSUMER organizations (who are paid by their members/readers). AV-testing is more simular to ISO business certification (companies getting tested, pay for the test). I guess truly independant testing is impossible on the internet where everything seems to be free (then we should pay the testlabs for the tests).
   
   
2. AV-Test is the biggest in terms of reader exposure, (with around 1 tot 1.5 million page views per month), with AV-Comparatives a close second (750K to 1.25 Million views per monh), then MRG and Virus Bulletin with (100-500K page views), the other smaller labs have less than 75.000 page views per month (like DennisLabs, NSS Labs, AV labs, etc).
   
   
3. When I randomly check some tests, it seems that AV-Test and AV-Comparatives are the most independant. The others often publish tests which are sponsored by one vendor and surprise. surprise that vendor also comes out best. Thruth needs to be told that AV-Test occasionally also applies simular tactics, have a look at their parental control test see picture https://www.av-test.org/fileadmin/_...erschutz_Software_Erkennung_EN_60e9e3445c.jpg That AV-test parental control test looks like a normal comparative test, but only two vendors paid (and passed the the test). So AV-Test publishes sponsored tests also, but sells them as independant tests (they should not do that IMO).

 

[Venustus]

That's it I'm switching to Panda!!

 

[Windows_Security]

Personally I give zero credibility to these type of tests, a test where are multiple AVs in multiple times getting 100% clean systems seems so fake to me, for example Vipre, we can just check the malware tests done here(on the Malware Samples topic) and keeps getting infected...

AV-Comparatives real world test are what they say they do: "whole product tests in real world environment". How real world is your scenario that a simple home users runs into ALL the malware collected by researchers in a month and EXECUTES them on his PC (bypassing 10 to 25% of the AV-defenses)? The average chance of running into malware for average home users is less then 0.0004% (I did the calculation in detail on Wilders Security once, search for it).

Remember the AV-industry is a 75 BILLION dollar industry. Their marketing is directed to giving us the impression that the sky falls on the earth when we don't use their products. They rely on people like you telling their friends and family they are in danger. Your friends don't have the knowledge to install a HIPS/FireWall/Sandbox. So when you scare them enought they will buy an antivirus. This is because most people think paid is better than free and they read that the paid version has the options you talk about.

Please don't take this as an insult or a personal attack: I am a Marketing & Sales guy who makes a living out of selling more to people who don't really need it (I am one of the bad guys :devil: ). Security is a hobby, I can be frank and open about stuff on this forum. That is why I helped Dan (of VoodooShield) in the past and will help Florian (of Excubits) in the future (to compensate and may be save my soul )

 

[dvdke]

VoodooShield with bitdefender free&comodo FW,works great for me.the latest infection i had was a virus in ccleaner.i dont use ccleaner no more.

 

[Mehdi Mohammadi]

Tencent & Vipre are better than Symantec. Really?

 

[Deleted member 65228]

Personally I give zero credibility to these type of tests, a test where are multiple AVs in multiple times getting 100% clean systems seems so fake to me, for example Vipre, we can just check the malware tests done here(on the Malware Samples topic) and keeps getting infected...

The tests are real and the results are truthful based on the tests carried out by the professionals who operate them. There's no harm in trusting the tests, and doing this doesn't mean you believe a product will have "100% detection" all the time. The vendors even note down on the result documents that a product scoring maximum does not mean it always will. All these tests should be taken with a grain of salt, it has always been this way and always will be... But that doesn't mean they aren't testing products correctly, effectively or demonstrating trustworthy results.

 

[roger_m]

Tencent & Vipre are better than Symantec. Really?

Tencent uses Bitdefender definitions, so it's understandable that it got a high score.

 

[Deleted member 65228]

Tencent uses Bitdefender definitions, so it's understandable that it got a high score.

I think that some people don't take a liking to some vendors doing better than their own preference, it has always been like this... Fact of the matter is that these testing companies pour so much work into testing professionally and conducting re-assessments prior to publishing the results. The results are based on fact... Product X detected X amount of samples, Product X blocked X amount of malicious links, Product X was light on resources or not under that testing environment, etc.

It doesn't mean a product is bad if it scores low and it doesn't mean a high scoring product always will score highly. It is just a test conducted by experienced people who are good with testing. Every vendor has good and bad days. Just because the results are factual doesn't mean they shouldn't be taken with a grain of salt and anyway these companies usually leave notes actually stating that 100% detection isn't always, just for their tests if it happens. We already know 100% of proper malware classification is impossible, the closest to that is blacklisted everything and allowing only the whitelisted (and then come file-less attacks from zero-day exploits from the web -> next big thing to overstepping the mark on that concept)

So also for what you said, I don't think it is weird that Tencent or VIPRE performed well in the test. Tomorrow? Maybe not so well. Its a hit and miss game with security software and malicious samples/website detection.