AV-Comparatives released Sept file detection tests

[YeOldeStonecat]

http://www.av-comparatives.org/detection-test/

Top 4 spots taken by brands we don't usually see up towards the top.

 

[Moltuae]

Interesting result.

Here's the real world test chart: http://chart.av-comparatives.org/chart1.php

Seems Trend and Qihoo are the top two this month. Who on earth uses Qihoo!?

 

[nlinecomputers]

I've never heard of Kingsoft AV (Don't they do a Office suite?) or Qihoo (gesundheit!).

 

[YeOldeStonecat]

Qihoo is getting popular, believe more overseas on the Pacific part of the world. I've tried out their free, "360" security. Recall it was triple engine. I have mentioned it a few times in the "free AV" thread I created.

However, "in the real world"....meaning outside of tech forums, doubt we'll ever run across it in use on the Atlantic side of the world.

 

[NYJimbo]

I might have to start reconsidering Avira for old XP boxes. I switched to Avast on XP's a while ago but if its really getting better and I can get their AV license so cheap it might be a good time to switch.

 

[TechPJC]

Who on earth uses Qihoo!?

http://www.techsupportalert.com/content/antivirus-number-one-contender-block.htm

 

[Metanis]

Surprised

Somewhat surprised they didn't test a Norton/Symantec product and also the one from Webroot. Sorta like a Consumers Report shootout of compact cars while neglecting to include a Honda or a Toyota.

-Mike

 

[mrapoc]

Is Threattrack Vipre what GFI use? If so...not good!

 

[Eagle21]

Somewhat surprised they didn't test a Norton/Symantec product and also the one from Webroot. Sorta like a Consumers Report shootout of compact cars while neglecting to include a Honda or a Toyota.

-Mike

This misunderstanding regularly occurs on this forum.
It's funny how people are surprised by this when they wrongly assume that AV-Comparatives is some kind of independent consumer testing organization.

Just take time to read their funding page
http://www.av-comparatives.org/funding/
and your 'surprise' may disappear.

If you 'follow the money' you will see that they test only the AV products that AV vendors pay them to test.

Perhaps we should ask YeOldeStonecat to add a disclaimer to this effect for when he starts one of his regular " AV-Comparatives" threads. Something to the effect "Here is a list of AV tests by companies who have elected to pay for the AV tests".

 

[Larry Sabo]

These companies all pay the same amount to be tested, according to the article to which you linked. Are you suggesting that the results are biased by their payment of these fees? I seem to recall that Symantec withdrew because they didn't like the abysmal test results for their product, although I don't have a citation handy.

EDIT: According to this 2012-04-16 post by "IBK" of AV-Comparatives in a Wilders Security Forum. . .

"Symantec only wanted to take part in our public tests if they could choose which of the tests from our yearly public test-series they participated in; specifically, they did not want to take the File Detection Test (formerly called On-Demand Test). As an independent testing organization, we require all vendors to take part in all the basic tests in the series, and do not allow them to cherry-pick tests. We feel that the File Detection Test is essential to showing the overall capabilities of an anti-virus product, especially with regard to threats that do not come directly from web pages, but are spread via email, LAN or flash drive. Other independent testing organizations (such as AV-Test, VirusBulletin, ICSALabs, WestCoastLabs) include, or rely exclusively on, file detection tests. We know that such tests are not easy to pass, especially considering the need to minimise false positives. As we cannot allow any vendor to opt out of any of the core public tests, Symantec has decided not to take part into our public tests this year."

 

[TechLady]

I'd still rather look at test results where no money was involved, personally

 

[YeOldeStonecat]

Somewhat surprised they didn't test a Norton/Symantec product and also the one from Webroot. Sorta like a Consumers Report shootout of compact cars while neglecting to include a Honda or a Toyota.

-Mike

All the players get invited....but some companies choose not to participate.

AV-Comparatives is the most neutral, unbiased AV testing site out there, it's the only one I really care to read. Many of their testing methods, such as the real world/dynamics tests, as the most "realistic" AV test out there...simulating actual scenarios that end users do to get infected.

Costs money...yes...but it's the same price, and...large staff, lots of time, it's expensive. These guys at AV-C have to put food on the table at home.

 

[YeOldeStonecat]

Is Threattrack Vipre what GFI use? If so...not good!

Yes...and for years, it's usually in the bottom 50% of the list.

 

[mrapoc]

I'm gonna have to rethink using that one then. Hmm

 

[MikeLierman]

Also check out the real-world detection test in September as well. It's a little different and focuses on in the wild samples and the antivirus programs ability to prevent the computer from being compromised and infected.
http://www.av-comparatives.org/wp-content/uploads/2014/10/avc_factsheet2014_09.pdf

Again, seen in this test Qihoo 360 shows nearly 100%. Never even heard of this product.

 

[YeOldeStonecat]

Yeah their Dynamics tests rock...I usually link those. Busy September 'n October so I failed to post it. But it's my favorite tests...very "real world" in how the replicate typical end users and infection methods.

 

[PcTek9]

Considering that there are millions of computer viruses, trojans, malware, spyware, rats, and etc., we need to remember that the "best" antivirus is really a subjective experience.
As an example, suppose one antivirus detects only 500,000 viruses, while another antivirus detects say 1.5 million viruses. It's possible that if the test maker chose, he could choose his samples in such a manner that the antivirus that only detects 500,000 viruses would come out shining like a star.
In fact, a researcher could make the one that detected 500,000 viruses show as 100% effective and the one that has 1.5 million signatures at 0% effective.
So they try to set up computers, and find viruses "in the wild", that are typically infecting normal people surfing the web.
The thing about statistics is that even though you can make a "science" out of choosing the perfect random sampling from a population, there is ALWAYS the chance that it could paint a picture which is completely wrong.