Avast! virus/trojan detection speed/quality

[Arup]

Vlk,

Its your humility in admitting that will keep me glued to Avast till you guys pry me off with a crowbar:)

Keep up the good work and my next year, Avast will kick butt.

[Juanjo]

Vlk, you don't know how I am glad of reading it, like we say in Spain "it was time to catch the bull for the horns" in this aspect. I hope that this hard work starts noticing soon.

[jujubee]

Dwarden,

1. about the detection rates and speed of updates: I sort of agree that avast! is currently not in its best form but we're working hard to change this. We have just hired 3 more virus analysts, are actively working on a way to greatly improve the process of sample submissions etc... However, all those things take some time. I'm hoping 2006 will be a pivotal year in this sense - you guys will see a dramatic improvement in the detection rates as well as reaction speeds - and avast! will return to where it was in the late 90's - on the very top.

First of all, I appreciate the honesty in admitting that Avast still has some work to do. I don't think companies like Symantec would ever admit that (and they are far, far from perfect)

I too am hoping for better detection rates. I feel that Avast is a high quality product, but I can't yet find myself wanting to buy the Pro version since I am not yet convinced that its detection rate and virus database is as good as other companies.

I know that Symantec can detect about 70,000 viruses, PandaSoftware over 90,000 viruses, and Kaspersky nearly 157,000 viruses!!!

How many viruses can Avast detect so far?

[TAP]

I know that Symantec can detect about 70,000 viruses, PandaSoftware over 90,000 viruses, and Kaspersky nearly 157,000 viruses!!!

How many viruses can Avast detect so far?

See this thread
http://forum.avast.com/index.php?topic=17856.msg151968#msg151968

Do you know how many viruses NOD32 or AVG can detect? I think the number of viruses in antivirus database tells nothing about the overall efficiency of antivirus in the real world.

[alanrf]

jujubee

If that is to be the basis of your judgement then I would assume that you also must believe the Symantec users get infected more than twice as much as Kaspersky users and I rather doubt that is true.

Surely the major issues are earliest recognition and deployment to ensure protection of users from current threats - not whose database contains the most antique viruses. I have to concur with TAP; how much do you (or I) know what those numbers really mean? 

I wish there was a reliably independent league table of infection rates of users of the competing antivirus products.   

[TAP]

The important things we need that are speed of releasing signature, accuracy of detecting *all* dangerous malware that are still circulating in the wild and the overall reliability/stability from avast! rather than the biggest database. Of course, effective-proactive detection is also great to have. 

[Dwarden]

well my main post was oriented not to get some totals number to "shock" and catch users ...

i was mainly interested in  Alwil's new method for submitting virus/trojan/exploit/spyware/malware examples
to Alwil team and it's fastest addition to Avast!'s VPS if positive one found!

also it was about getting some feedback or info about file(s) submitted (like clean, damaged/nonfunctional malware code, false alarm, not clean(soon to be added to VPS) and so on) ...

i don't care if it's automated email answer or human written one or some website based enhanced scanner with upload ability ... but it must return needed answers ...

it's all about this situation "i sent You week ago piece of malware identified as Trojan.Java.ClassLoader.f [/i] by 18 antiviruses ..." and my question is "Is it really dangerous? (seems so), Is this positive match? (seems so) When it appear in VPS (so far unknown)? If not then why etc...

so far Avast! reacts very well to main threat now it's about smaller evil ...

and as it was said Alwil work on solution ...

[jujubee]

jujubee

If that is to be the basis of your judgement then I would assume that you also must believe the Symantec users get infected more than twice as much as Kaspersky users and I rather doubt that is true.

Surely the major issues are earliest recognition and deployment to ensure protection of users from current threats - not whose database contains the most antique viruses. I have to concur with TAP; how much do you (or I) know what those numbers really mean? 

I wish there was a reliably independent league table of infection rates of users of the competing antivirus products.   

Hmm, it seems like the numbers of viruses that the other companies are reporting maybe due to over-counting. That's news to me. I always thought those numbers represented unique individual viruses.

When I used Symantec Antivirus, I sometimes would check the number of viruses that were in its virus list. I would very quickly scrolldown the list of viruses that it showed, and it seemed to me that every virus presented was its own unique virus. But then again, I obviously did not run through the entire list to see if 70,000 viruses were actually listed on it.

I too agree that the more important quality in a virus scanner is the ability to detect ITW viruses. Before I downloaded AVG and Avast, I checked to see that they were ICSA certified and when their most recent Virus Bulletin award was (AVG - Dec 2005,  Avast Oct 2005). I don't really know much about Virus Bulletin, but I'm assuming its a trusted authority on the detection ability of virus scanners. Anyways, I saw that both programs did well recently and that gave me some assurance that they have good detection that is atleast up to par (and maybe even better?) than the rest

[TAP]

Hmm, it seems like the numbers of viruses that the other companies are reporting maybe due to over-counting. That's news to me. I always thought those numbers represented unique individual viruses.

I don't know but I don't think it is over-counting, I think it's a difference of counting method.

For instance, according to the latest av-comparatives.org test August 2005, AntiVir says it has 202,710 viruses in its database,  Dr.Web says 82,894 and I think avast! has approximate 46,000 virus units in its database at that time but AntiVir, Dr.Web and avast! almost have the same overall detection rate level (zoo malware) and also, avast! almost have the same overall detection rate as Trend Micro.