Jump to content


Friends, we are happy to invite you to our NEW Global Kaspersky Club! Please follow this link www.kasperskyclub.com , sign-in and enjoy our new platform!


Photo
- - - - -

advanced+ on the proactive test


  • Please log in to reply
24 replies to this topic

#16 TJP

TJP

    Young Gun

  • Members
  • PipPip
  • 74 posts

Posted 02 December 2007 - 01:18 AM

No worries, it says 0% doesn't it?
If You are warned the percentage will increase. Too much warnings and you are banned


Well I've got no problems then as I don't plan to be a "keyboard warrior" or libel other AV vendors within my posts.

I just wanted to check with the moderators as this stat magically apppeared.

Cheers.


#17 TonyW

TonyW

    Veteran

  • Members
  • PipPipPip
  • 188 posts

Posted 02 December 2007 - 11:25 AM

(Nod32 at 71& & Kaspersky at 40% )?? :)
How could Nod32 beat Kaspersky at proactive detection= detection of unknown samples without use of signatures??? :) Kaspersky has much superior heuretics than nod32 now & should be the leader & the best in proactive detection instead of nod32!!!! :) . Why is AV comparatives siding with nod32 by the way???? :) Well at least Kaspersky got advanced+ so I am a bit happy. :)

A number of points to consider:

1. AV-Comparatives does not side with any AV vendor.

2. v7 of KL products use a completely new heuristics engine, which is being fine tuned all the time. Previously, KL scored 9% with v6.0.2.614. A one-off test was done with v7.0.0.119, which produced a score of 35%. A big increase from the earlier test, wouldn't you say?

3. Any false positives or missed detections are reported to the AV vendors, and these are already fixed.

4. The tests don't use current signatures or presently tweaked heuristics. They were using the products as of 3 months ago. In the case of KL, had the current or even beta heuristics been used, maybe a higher result would have been achieved. The heuristics have improved a lot since the August test, and are continuing to improve.

5. Don't just rely on heuristics. In KAV/KIS, we're also lucky enough to have the PDM, which when used alone has been shown to provide very strong proactive protection.

#18 ultima weapon

ultima weapon

    Veteran

  • Members
  • PipPipPip
  • 223 posts

Posted 02 December 2007 - 01:48 PM

A number of points to consider:

1. AV-Comparatives does not side with any AV vendor.

2. v7 of KL products use a completely new heuristics engine, which is being fine tuned all the time. Previously, KL scored 9% with v6.0.2.614. A one-off test was done with v7.0.0.119, which produced a score of 35%. A big increase from the earlier test, wouldn't you say?

3. Any false positives or missed detections are reported to the AV vendors, and these are already fixed.

4. The tests don't use current signatures or presently tweaked heuristics. They were using the products as of 3 months ago. In the case of KL, had the current or even beta heuristics been used, maybe a higher result would have been achieved. The heuristics have improved a lot since the August test, and are continuing to improve.

5. Don't just rely on heuristics. In KAV/KIS, we're also lucky enough to have the PDM, which when used alone has been shown to provide very strong proactive protection.



Well that is good to know that Kaspersky is now giving nod32 a run for its money in heuretics. Kapersky will also introduce packer detection in version 8, I saw it in the official forum so It may also help but its downside is the increase of false positives. Nod32 is weak in signature detection & does not have packer detection by the way so it relies too much on heuretics, But from the looks on things maybe nod32 is not the leader in heuretics anymore. Virus Info Tests shows SOPHOS as the leader in heuretics now that what I believe.
http://virusinfo.inf...p?page=testseng
Posted ImagePosted Image

#19 TJP

TJP

    Young Gun

  • Members
  • PipPip
  • 74 posts

Posted 06 December 2007 - 10:11 AM

Anyone care about KAV failing the December VB100 test?

While the sky isn't falling for me, I was disappointed with the 1 missed sample (rather than fail due to a false positive). Mind you, in this test a lot of the top tier anti-virus solutions failed due to false positives and/or missed samples.

Further, does anyone know if Kaspersky Labs adds samples they miss in tests? I have read other AV's do this, just wondering if Kaspersky did the same.

Cheers.


#20 ultima weapon

ultima weapon

    Veteran

  • Members
  • PipPipPip
  • 223 posts

Posted 06 December 2007 - 10:33 AM

First of all you must remember that no security product can offer 100% detection that why I have several scanners in my computer. Test also vary because of different samples. That why A antivirus will do well on some test & some dont.
Check the results 89& , 70& , 91& of samples detected.Have you noticed that there is never a 100%. What does that tell you no security product can catch 100% of the threats.
Posted ImagePosted Image

#21 Sjoeii

Sjoeii

    Kaspersky Club Project Manager

  • Root Admin
  • PipPipPipPipPipPipPip
  • 4591 posts

Posted 06 December 2007 - 11:32 AM

Don't worry about these tests. Please don't believe everything you read. There were more failing. Kaspersky is safe to use
http://twitter.com/kaspersky_club

#22 TJP

TJP

    Young Gun

  • Members
  • PipPip
  • 74 posts

Posted 06 December 2007 - 12:11 PM

Thanks for responding. I think that as we discuss & applaud great test results, we should also discuss the less stellar test results...

Rest assured that I am far from worried about Kaspersky's ability to protect we users - hence my "while the sky isn't falling for me..." in my original post. Kaspersky continues to be classified amongst the top 5 av vendors when one reviews all of the numerous testing results :(

One question remains unanswered: does anyone know if Kaspersky Labs add missed test samples into the database?

Cheers.


#23 Sjoeii

Sjoeii

    Kaspersky Club Project Manager

  • Root Admin
  • PipPipPipPipPipPipPip
  • 4591 posts

Posted 06 December 2007 - 12:14 PM

Thanks for responding. I think that as we discuss & applaud great test results, we should also discuss the less stellar test results...

Rest assured that I am far from worried about Kaspersky's ability to protect we users - hence my "while the sky isn't falling for me..." in my original post. Kaspersky continues to be classified amongst the top 5 av vendors when one reviews all of the numerous testing results :(

One question remains unanswered: does anyone know if Kaspersky Labs add missed test samples into the database?

Cheers.

Not without testing themselves. They always check malware you send in and then add it to the sigs
http://twitter.com/kaspersky_club

#24 ultima weapon

ultima weapon

    Veteran

  • Members
  • PipPipPip
  • 223 posts

Posted 06 December 2007 - 12:48 PM

Thanks for responding. I think that as we discuss & applaud great test results, we should also discuss the less stellar test results...

Rest assured that I am far from worried about Kaspersky's ability to protect we users - hence my "while the sky isn't falling for me..." in my original post. Kaspersky continues to be classified amongst the top 5 av vendors when one reviews all of the numerous testing results :(

One question remains unanswered: does anyone know if Kaspersky Labs add missed test samples into the database?

Cheers.


I own several licenses of security products & everyday I encounter mallware because I own a computer sales & repair shop. I own Kaspersky, Bitdefender, Nod32 & Avira licenses. Considering to buy Panda, still on trial now.. I tested all this products on several computers that my customers brought to my shop. Very rarely does kaspersky missed! It is the best to me & I respect it the most!!!. Those test have no effect on a real world scenario because mallware samples are different for every individual & it is impossible to have a complete library of mallware for testing. I also back the admin statement that Kaspersky offers the best protection.

Edited by ultima weapon, 06 December 2007 - 02:27 PM.

Posted ImagePosted Image

#25 TonyW

TonyW

    Veteran

  • Members
  • PipPipPip
  • 188 posts

Posted 06 December 2007 - 01:09 PM

[size=2][font="Franklin Gothic Medium"]
One question remains unanswered: does anyone know if Kaspersky Labs add missed test samples into the database?

As far as I'm aware, they do add missed samples if they're sent to them. The last time KL failed a VB test was due to KL removing a detection from the database to tweak it, and, unfortunately, this was at same time test was carried out.

At the end of the day, it's the real world situation where any AV is best measured against.