MGR test #19

This is not good for comodo, all of this work and it only moved up 1% from the last test. :frowning:

Program Detection Rate (%)
a-squared 99.7%
Avira 99.5%
G DATA 99.4%
Kaspersky 98.8%
Avast 98.7%
BitDefender 98.6%
eScan 98.5%
Norton 98.2%
Nod32 97.4%
COMODO 97.1%
McAfee 96.8%
F-Secure 96.4%
AVG 96.2%
Norman 95.4%
Twister 94.6%
Sophos 93.5%
Spy Emergency 82.4%
ClamAV 82.3%
Dr.Web 79.5%

Well, in the zone that it’s now, it will get harder and harder to get those %'s. If you are pretty new, and there are so many virii in the world. You grab a few of those and then test… it’s hard to compete with Avira etc.


Only a 2.6% gap with the first scorer… on 639 424 samples

Changes form the previous Malware Research Group Project #18 (395 844 Samples):

a-squared : + 0.1%
Avira : - 0.1%
G DATA : 0.0%
Kaspersky : 0.0%
Avast : -0.5%
BitDefender : -0.1%
eScan : 0.0%
Norton : -0.8%
Nod32 : -0.2 %
Comodo : + 0.9 %

Good to go Comodo. :-TU

Well Cavs is definitely going in the right direction. Most of the “big avs” are roundabouts where they always are in these tests. Comodo is creeping up bit by bit and as i always said, once the family signatures come in and some fixes are done to the av, i am confident the gap % wise will be miniscule.

Well done Comodo, keep on improving and get ready for the kicking of ■■■■, its just around the corner.


I’m giving g-data a try if I like it I might move my whole family over there. Who can argue with two engines, heuristics, behavior analysis and firewall for $40 for 3 computers. I am sorely disappointed at comodo. :cry:

Choice is always available. That’s what life is all about.

Unfortunatly they didn’t test for false positives this time around.

I think this is pretty good for Comodo. The first steps are always rather easy. It is more difficult to grow at the last steps

I can see CIS catching the 6,585 Malware that it missed the facts iis that 1 well placed Family Signature could do the trick :wink: and i realy get the felling that there will be adding well over 500 Family Signatures!

And then add CIMA on top for v4, boy if i ever see a AV with 100% or 99.99% detection ratio i would try it!
And languy dont buy G-Data yet wait until MRG tests 3.10 and matt tests 3.10

Im not sure but is the memory scanner tested (BOclean) (does the manual scan use BOclean sigs)?

Thanks for this guys.

As I have stated previously, give us 12 months from our launch. I believe we have made some huge progress in less than 8 months and still 4 months to go… it is thanks to your support we have come so far, pls keep it going, it will continually improve! We are determined to continue to improve if you are determined to help us! Only thru working together can we achieve our goal of a top notch AV thats free to all! Thank you all and keep up the support pls.


Totally agree Omelet guy. Languy is being a wee bit impatient on this one i feel. Lets cast our minds back to january this year when Cavs was tested by the same company and it did 90%.

In 6 months that has increased by 7%.This i laregely due to the size of the data base, and if Languy can hang on a week or two then i think once 3.10 is tested by Matt at remove-malware .com and others then i feel we will see the detection rate gap drop dramatically.

However remember its not just about detection. Prevention will always take first place in my book and i have yet to see anything get through D+ in its present form.

The planned improvements and family sigs should make a big difference, so please dont jump ship yet, give Cavs the time it deserves.

Just my few pennies worth chaps.


I’ll wait, I have to really give g data a work out (especially if I have to pay for it) and I have a month to try it. But so far it has impressed me. :stuck_out_tongue:

Apparently they never did ( but as for FP there are some handpicked ones in the FP boards…

Some results of this review are really surprising. For example, I don´t belive that Dr.Web is so bad AV. I haven’t had problems with malware yet. It is possible to trust this review? :-\

Also we have to take into account that MRG has malware from 1 year ago, CAVS was just a kid back (was CAVS around 1 year ago?) then and i think that’s the malware that is missing detection by CAVS, as each month goes by i would say detection will keep going up until vs 4.

CAVS 2 was in BETA for 2 years. Plenty of time to collect signatures :slight_smile:

Choosing CIS for its antivirus is crazy. ??? ??? You act like the AV is “the” layer.
CIS has a much better history against ANY samples. The D+ and Firewall has always scored top and been able to prevent even the newest infections(such as conflicker), before it even was out CIS prevented it. Thats more than you can say about a AV. :wink:

However you pick what you like, and I wish you the best… Hopefully we won’t have to see you turn into a new “anti comodo” dude posting at Wilders to sell Gdata, or whatever… 88)

CIS will have a top notch AV as well but it will take time, 3.10 will be a shot in that direction probably. CIS badly will need some heuristic to match the bigboys…

We are very thankful for all samples you provided however… Have a nice day, and hopefully you find a security product you like! =)

Hi Monkey boy. You are a man after my heart in that you also know that the av is only part of the “layers”.

We know that prevention is always better than detection.Yes, we would all like to see Cavs have the best detection rate as well as the top notch D+ and firewall. However while Cavs antivirus matures and gets better and better we can sit back and say " if a nasty does bypass Cavs, D+ will kill it stone dead!".

So its just a matter of time before Cavs catches up, and those who are impatient will see Melihs promise come true.


Hey no problem but the deal with 3 licenses is too good. So far G-data has fairly good behavior analysis also, I found samples that were not detected by avast or bitdefender and ran them sandboxed, and I actually got pop ups from G-data. I have yet to see how I can test heuristics on it but nothing has popped up. The firewall seems pretty good if you set it right. I also want to set up my brother and father with g-data becasue basically there is no user interaction and both of them are computer illiterate. But before I do that I really want to give it a work out.

The main reason I am disappointed at comodo is because between MGR test 18 and 19 they almost doubled the DB but only gained 1% detection rate. That is not good in my opinion. And there are some other things I have problems with, basically in the main AV engine. How it scans some files and how I can’t tell if a scheduled scan was done or not. What I will do if I deiced on G-data is keep my eye out on comodo. I will still submit to comodo but now will also submit to avast and bitdefender. ;D no big deal, and no I will never talk down about comodo.

I still have to try g-data on my fathers computer that is older and see how it really affects the system performance wise.

Did GDATA signature DB increase too?
I fail to see how 0.9% from MRG 18 could be no good compared to GDATA 0.0% increase from MRG test 16 onward.