Using AVG with no apologies

i believe there are no apologies necessary for using AVG, or for a thread being closed. i made a point in another thread how there seems to be more arguing now than when there were vs threads and it was closed, so it's ok. i am asking again though, can AVG which is not a bad av improve its heuristic detection? i saw the post about how AVG is doing well in detection on the Jottie website so it must be pretty good already. i think AVG with ThreatFire and the new Online Armor Free version along with AVG AS or SuperAntiSpyware for on demand would be a great combination.

 

Wordward,

Until any product reaches 100% legacy and 0-day detection, there's always room for improvement. It really comes down to the technical expertise of their staff and the how well matched that expertise is to emerging threats.

On a more specific note, understanding of course that any static test result is subject to a multitude of qualifiers and caveats, look at trending behavior in fixed test protocols over time. For example, let's say we look at the trends from the last 5 tests at www.av-comparatives.org looking at the results which don't include DOS or other OS malware:

• On-demand comparative (old to recent for AVG only):83.3, 90.47, 91.69, 96.37, 97.75% (AVG Anti-Malware)
• Retrospective comparative (old to recent):4, 4, 7, 3, 8%

So, the general trajectory is upward in both areas. As with many products, retrospective results lag by a substantial margin given that best in class results for that test now hover in the 50-70% range with a large number of contenders above 25%. Obviously, that means there can be plenty of room for improvement.

That's just one way to look at things.

Blue

 

sorry recent closed threads makes me hold back somewhat. i think what i am asking is this. after seeing AVG AM do so poorly in the AV Comparatives pro active/retrospective tests compared to almost all the other AV's, can AVG rebound and regain any lost confidence in it's heuristic detection? also does anyone think Grisoft may have seen the results and will step up and try harder now to improve its heuristic detection? they seem to be trying to improve the polymorphic virus detection with new program updates, but i guess the heuristic improvements take longer.

 

The recent run of closed threads shouldn't hold members back from making well reasoned points, hopefully those threads will illustrate that some avenues tend to leave a reader following the thread as unaware of the subject matter as when they started. For example, volleys of sound bite opinions back and forth with no expansion as to the rational origins of those opinions doesn't help anyone.

Grisoft doesn't need confidence in the detection, they need results. When I see a relatively static result like that in any technical subject area and a simultaneous increase by a reasonable fraction of the competition, I always wonder if they're answering the wrong questions in their development efforts. In other words, are their solutions geared to the proper challenges out there? I don't know the answer to that question, but it's also important to understand in detail the limitations of the retrospective test protocols.

The reality of market dynamics is that they must, although realisitically there are a number of mitigating factors that can improve real world performance well beyond that predicted by those types of test results. For example, a very aggressive strategy with respect to deployment of new signatures or addition of a bare bones HIPS-like module that locks down the bulk of current attack points in the event something is undetected.

The hard things always do....

Blue

 

Improving the detection of polymorphic viruses means improving the scanning engine. A better engine might lead to better proactive detection.
AVG 8.0 should bring the full power of the Ewido engine. If Grisoft manages to keep a low amount of FPs, a Advanced checkmark in retrospective tests can be achieved.

 

AVG is trying to improve the polymorphic as well as heuristic detection of its products since last year. And if you read the detailed polymorphic results in the PDF, you will indeed see an improvement in terms of percentage numbers of polymorphic malware detected by AVG (but only an 100% detection for each polymorphed variant of that particular polymorphic virus is counted as a reliable detection). Therefore, I will not say Grisoft is just sitting there doing nothing, but improving both the heuristic and polymorphic detection is a long process and results is not usually seen immediately.

So, the answer is YES, AVG can and will improve, but any significant improvement will only be with the release of version 8.0

 

avg certainly wont do anything more with 7.5,

and like every new version, people want to see an improvement.

good luck to ya with 8.0, im a fan of avg and have been for many many years, so i'll be hoping this works out for ya and everyone else.

-- i agree, that more arguing happens now

 

thank you everyone. i like the looks and simpleness of AVG, but realize Avira and perhaps Avast would be a better choice for now. just hoping it will keep improving and that it is enough protection with ThreatFire and the new free version of Online Armor for now.

 

I have AVG pro on 2 pc's and Avira Premium on 2. I trust them both.

As far as heuristics, I never saw AVG flag anything as "heur" several months ago. Now when I see VT results, it fairly often does. If you also use a HIPS, there is no reason not to feel safe with it. You must also keep in mind how good AVG AM did the latest demand test - it actually earned advanced. That said, I think it's a product on it's way up. I'm looking forward to version 8.

 

If you run Threatfire with it, you should be fine... that's a good combo for some extra protection.

 

thank you Kerodo.

 

well,

i really like AVG, i aint sure about the free versions detection, but i think the software is fantastic.