AV-comparatives Retrospective/ProActive Test - November 2008

i don't care about many FP as long as my anti virus protect me from trojan/viruses.

even average user will choose protection compare to few FP.

If you don't wan't FP, then uninstall your AV.....no FP at all, that's Advance ++

 

I agree, sometimes malware is less dangerous than a FP

 

The other thing to remember about the FPs in relation to these tests is that many of the packages listed in the report won't be on everyone's machine. I've said previously I hadn't even heard of most of the packages let alone have them installed on my system.

These packages would have flagged an alert if you had them all on your system, but by the time the test came out, the FPs were fixed anyway so if you tested a package after test publication, there should have been no alert.

I guess this is why you see a lot of people here say they've not had a FP in all the time they've used AV x, y or z. It does depend what you have on your system I think. This could be why some of these FPs don't get fixed sooner because most people don't have these programs that flag the alerts otherwise they'd be reported and corrected by time testers like av-comparatives come to test the whole lot.

 

What I don't like about this rating system is that having more false positives has more impact than having a good detection. You can very easily loose your advanced++ because of a handful of false positives, but having 25% more detection don't give you any advantage in the rating.

So if Avira would have been tested on heuristic level 2 instead of 3, it would have scored advanced++. Which is the default setting. Oh well...

 

nothing makes sense until you know how many files were used to scan for FPs. Why is it so hard to find that out.

 

And I think less than 1% of the average users ever heard of a FP.
They buy (if they do) what is in the shops (boxed) and mostly what is advised by the shop seller
There choice is limited to about 5 AV's or so, or more likely Secuity Suites.
They will return to the computer shop when their system is crippled and pay additional money for fixing their system.

Gerard

 

Hi,

may someone answer this riddle...

On-demand comparatives August 2008
Very high detection rate + many FPs = ADVANCED+
Retrospective/ProActive Test November 2008
Highest detection rate + many FPs = ADVANCED

For a Certification level you need three shells and a pea?

Cheers

 

AntiVir 67% & 71% is what counts to me

 

The number of clean files scanned for testing false positives is not so important. It would not make much sense in turning that number to % on the total of scanned clean files.
However, there is no good way to explain why 15 fps are "few" and 17 are "many". Those ranges he selected are pretty arbitrary in my opinion.

 

Same here, nothing else matters.

 

Stefan, any ranges will be arbitrary (not that I'm arguing the current ones make sense). But that's exactly why a % rate is lot more informative for readers/users - not only for detection, but also for FPs.

 

I think that's the wrong way. I could create a tool which flags 100 % and then it's the best SuperAntiVirus?

Just my 2 cents ...

 

well the problem is that you wont see much on percentages. the clean set has surely serveral 10000 files so the percentage will be for all below 0.1 %
and there you wont see anything.

 

Hmm? Now that's a whole lot more useful information, isn't it? Means the amount of FPs is pretty much negligible for most of the tested products. OTOH saying 17 FPs gives you no idea how "many" or "few" those actually are. It's like providing the number of detections as an absolute number without mentioning the number of samples. How's information like "Foo AV detected 12345 malware samples" useful without knowing how many it missed?

 

If the tool detects and blocks, thats the right way.

 

In VB you'll loose your Vb100% even with a single FP .

Time to have a talk with IBK and tell him test AVIRA with the default heur. settings

 

That ranking is ok!

 

that's what av-comparatives has proposed to do in future, but avira prefers that also in future the highest settings are used.

 

Then Stefan has no reason to complain that they didn't get Advanced +

 

True HiTech, he really does have nothing to complain about. Especially with that 99.6% detection rating.

 

On a side note, seems like their site got the "do not link to the results" madness to an entirely new level. Going to that site via http://www.av-comparatives.org/ link works, you can go to Comparatives section, click on the "Online results" links and get to the results. Now go to the site via http://av-comparatives.org/, try to do the same and enjoy the nifty 404 error. Folks, if you really intend to nag people with nonsensical policies, at least get your referrer checks right.

 

no 404 error for me here

 

Sure, 404 for me.

 

404 for me as well

 

Sure you can.....