Prevx1 detection rates falling down?

Discussion in 'other anti-malware software' started by ako, Mar 23, 2007.

Thread Status:
Not open for further replies.
  1. Seer

    Seer Registered Member

    Joined:
    Feb 12, 2007
    Posts:
    2,068
    Location:
    Serbia
    Hello. :)

    Oh well, if it's not a test but rather a statistic insight, then I obviously misinterpreted it. I am not an expert, far from that, and as I said not even a Prevx1 user (hell, I haven't even see the software) so I cannot tell how relevant this statistic is regarding Prevx. And I will not go into that anymore. But I believe I really wanted to point out how funny an average user's reaction is to these tables, tests, statistics, whatever. Generally, after such review (take AV-comparatives or Matousec i.e) if his beloved security software doesn't rate well, Joe immediately start questioning himself whether or not his security is strong enough, and accordingly (in most cases) start with various deinstallations of current software and installations of "better" one. So, valid or not, statistic or test, it draws significant attention and seriously affect overall usage of security software. I've been using NOD32 for over a year now, and never considered removing it, no matter what the tests or statistics say (it isn't so shiny in this statistic either). Even if NOD doesn't rate excellent with every test, I don't think that my security is worse than it was a year ago.
    So, Blue, don't hold nothing against me, it wasn't my intention to criticize this compilation of statistics (I know nothing of it as you see), but rather users' literal comprehension of it. I ranted before on this same subject, and I suppose I will tend to do it also in the future whenever I see the lack of common sense. Sorry if my sentences don't look very polished, but english is not my native language, and it seems I have some struggle on my hands here.

    My regards,

    EDIT: Oh yes, I had to change my avatar accordingly. It says everything about my attitude for tests and ratings. It's a pity Wilders doesn't allow it bigger than 60x60. :)
     
  2. BlueZannetti

    BlueZannetti Registered Member

    Joined:
    Oct 19, 2003
    Posts:
    6,590
    The Seer,

    I believe you have captured the reaction of the average user running across these types of results. I'm sure that they would maintain - if it's fundamentally flawed, why is the information gathered and disseminated? It's a good question.

    In this case I used the terminology of running statistic to emphasize a couple of points:
    • There is no way to rerun the test if questions of methodology arise. There's no redo of alternate approaches to allow one to perform a sensitivity analysis.
    • It's easy to weave a scensrio in which a product lags others by a few hours and always shows as missing in this evaluation. Is that a real difference? For most I'd say no, for some I say yes.
    • Look at the tabular data. Because of products entering the fray at different times, some have been challenged with only (as of now) 161 samples, some have seen ~ 450, others ~ 900-1000, and the bulk around 1980. In other words the statistics reflect a varying testbed. It turns out, as of now the most "effective" product, Webwasher, has only seen 161 samples. Others have seen a testbed 10 times larger. This discrepancy will diminish over time, but will never be fully remedied. For the average user, is the 84% for Webwasher (135 out of 161) really better than, for example, the 46% from KAV (911 out of 1987)? Who knows at this point.....
    This is a perspective I wish more people held....
    Your English is fine. As for being critical of tests/statistical compilations/or even my postings, please continue to do so since it is an important activity. That's how we all improve our knowledge in this forum, myself included.

    Cheers,

    Blue
     
  3. tobacco

    tobacco Frequent Poster

    Joined:
    Nov 7, 2005
    Posts:
    1,531
    Location:
    British Columbia
    I would like to add this.

    I've uploaded at least 5 samples that shows Prevx failing detection but upon running this malware immediately afterwards, Prevx grabbed them all by the throat:thumb:
     
  4. Longboard

    Longboard Registered Member

    Joined:
    Oct 2, 2004
    Posts:
    3,238
    Location:
    Sydney, Australia
    @Blue Zanetti

    I'm just trying to integrate your assessment of the data as presented vs the apparent number of threats exposed /utility.

    How does this fit with Prevx and NAV disappearing off the chart.?
    PrevX in particular going from close to "best" to close to worst.

    If Prevx started at close to the top and has since then been exposed to the same samples as the other tools: what to make of falling performances ?

    Is this a reliable, closer to real, scenario of zeroday threats or immediate response tmes to recently seen malware.

    We all accept that at some level some antimalware tools are more effective than others.

    Is this a reasonable compilation of real time effectiveness V sig databases or not?

    Regards.
     
  5. Seer

    Seer Registered Member

    Joined:
    Feb 12, 2007
    Posts:
    2,068
    Location:
    Serbia
    Hello Blue. :)

    After your last post, I really don't feel that I need to add anything more on this subject. Thank you for encouraging my rants and I would only add that I am very pleasantly surprised to see the patience and kindness of staff (and members for that matter) here at Wilders. Very rare place you have here indeed. So, until the next time,

    My best regards :thumb: ,
     
  6. aigle

    aigle Registered Member

    Joined:
    Dec 14, 2005
    Posts:
    11,164
    Location:
    UK / Pakistan
    Thanks. Any link for it?
     
    Last edited: Mar 25, 2007
  7. lucas1985

    lucas1985 Retired Moderator

    Joined:
    Nov 9, 2006
    Posts:
    4,047
    Location:
    France, May 1968
  8. aigle

    aigle Registered Member

    Joined:
    Dec 14, 2005
    Posts:
    11,164
    Location:
    UK / Pakistan
    Thanks.
     
  9. Dwarden

    Dwarden Registered Member

    Joined:
    Apr 11, 2003
    Posts:
    177
    Location:
    Czech Republic
    i wonder why Avast! is missing in this statistics ...
     
  10. ako

    ako Registered Member

    Joined:
    Nov 16, 2006
    Posts:
    667
    It is there (>20). Follow the link there.
     
Thread Status:
Not open for further replies.
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.