Why do these two tests have so many differences ?

I found a site of antivirus software testing recently, it is called “Malware-Test”. You can see it here.

The test team was founded by some experts once worked for Trendmicro.

I found that the result of this test had many differences with the test done by AV-Comparatives.

For instance, NOD32 v2.5 , in AV-Comparatives , the detection percentage is 98,61% . At Malware-Test , the result is 49.51%.

Another example, Kaspersky and Dr.web both performs good in these two tests. Kaspersky's detection percentage is higher at AV-Comparatives, Dr.web's detection percentage is higher than Kaspersky at Malware-Test.

After read thest two tests ,you can find the statistics had so many differences. Some differences are very obvious. Can someone tell me why ?

PS:Malware-Test provide a kind of test that shows the treatment rate of anti-virus software, I think this is very useful. I hope AV-Comparatives could also do this kind of tests. Thank you.

 

Unfortunately malware-test.com is not a trustworthy source and their results are extremely skewed to say the least.

 

"some experts" ... "once worked for Trendmicro" ...

As what? The test is flawed to maximum.

 

I found this info on the website:

About us

Anyway, the tests are done using samples got from a honeypot. Which is why the test is inaccurate, or so I think.

 

Oops, I forgot to post a link to the previous discussion.

https://www.wilderssecurity.com/showthread.php?t=157575&highlight=malware-test

 

if you look at http://malware-test.com/smf/index.php?action=dlattach;topic=3125.0;attach=47 you see that
a) some products (if not all) were just trial versions, which means they had not the full potential as the full versions (= it is not true that they had the best possible settings enabled, cause in some trial versions that is not possible) and
b) not all products were updated the same day.

 

Samples from honeypots? I hope they carefully sort out the garbage/damaged samples - which make up most of the files you get from this source.

Obviously they don't do that - the NOD32 detection rate is a good indicator for this.

 

I don't know who's behind the Malware-Test, i don't know who's behind AV-Comparatives, I'm not working for anyone of them etc...but since i'm a member of Wilders forums i observed something interesting: Any test that doesn't rate NOD32 detection rate 98% is inaccurate and not a trustworthy source. How does it sounds ?

 

You want an honest answer? That sounds stupid.

Why would someone working for another av company than ESET (Example: Stefan and myself) write their doubt's? We're not working there. Or in my case no longer anymore. Don't you think your statement is quite "bold"? I mean you should have noticed that somebody else confirmed that too, not only the usually NOD users. That should ring somehow a bell.

 

if any of the top tier AVs scored such a low result then it would be cause to doubt the test, it isn't just a NOD32 thing.

 

The virus trends in Asia could be diffrent from in Australia, I think.

I thought IBK works for them.

 

I am the "chief" of av-comparatives

 

And if you will look at all the reports at AV Comparitives done from last year, you will see certain ones where Nod wasnt the winner. So that pretty much blows your theory out of the water.

 

Hmm Nod didnt win this one. IBK does a great job at ensuring things are as fair and equal as can be. In his summazation of all the products for the year, yes Nod won, but he also stated that Avira will stand a great chance of taking that honor this coming year if they keep up the good work.

 

well done last year

cheers

 

So, you own av-comparatives. Cool.
Sorry for my ignorance.
Keep up your great work!