PC Mag test ; unexpected result for BD and PCAV

Does the second part of the malware removal chart show cumulative scores or something else?
At the bottom it shows; PCAV 1.5 Free Edition compared with other recent and relevant products.

However, the malware removal scores seem odd.
Comparing 'PCAV 1.5 Free Edition' and 'ZoneAlarm AntiVirus + Firewall 2012';
All malware; 91% - 5.9 / Rootkits; 100% - 4.1 / Scareware; 100% - 9.5 (PCAV)
All malware; 79% - 6.0 / Rootkits; 100% - 6.7 / Scareware; 100% - 7.5 (ZA)

How in the world can 100% rootkit removal, lead to such different scores?

 

The only reason Comodo and Outpost did so well is probably because of their HIPS. However, the people testing probably know what to choose for the alerts, while the average user would probably allow everything, leading to lower scores. As far as AdAware, IDK. Even the free version is supposedly better than all the other AVs? Unlikely.

I don't think I trust their tests much. I mean it says Norman did better than BitDefender lol.

Also, does the free AdAware have GData's engine? If so I might switch Prevx for it and use it alonside Avast! if that's okay. My teacher uses AVG and AdAware together with no problems.

 

I would choose Prevx before Ad-aware anyday

 

I would too but I have the free version so it only detects.

 

Pity PC Mag doesn't post dates on these charts.

The one in this thread looks like it was from Jan.

This one is from the March lineup for security suites:
http://www.pcmag.com/image_popup/0,1871,iid=289190,00.asp

 

It's (whatever it is) so heavy it made my desktop fold into itself and now all my cables disappear into an infinite negative time singularity.

 

While I respect Neil quite a bit and know for a fact he has more experience in malware analysis and AV testing than most "independent labs" out there, the main problem I see with these tests is that the "malware testbed samples" are recycled only once per year. So obviously a product which is tested in month-0 after renewing the testbed is at a serious disadvantage over a product tested at month-10 for example, as the second product will have a 10 month advantage over the first product to come across the sample and add it to its signature database.

 

It uses Vipre. Your teacher is lucky for now.

These test are always interesting for me, so thanks.

Is there one page where all the results and updates are located?

 

+1 "Pbust" totally agree with you.
PC Mag test are always like this

 

If this is true, these tests are worthless.

 

thank you
but i dont trust this test i dont know why my feeling

and
where is norton internet security ?

where is avira ?

what about maleware bytes ? why its poor ?

 

Sorry to hear you respect a tester that uses a test model which is so fundementally flawed.As posted above how can any results be trusted ?

He will know that and yet still publishs data into the public realm.

Sorry this guy might be employed to do a job but he blatently is negligent as a tester

 

Malwarebytes isn't good against old and infrequent malware.

 

Neil has been employing real-world test (aka whole-product test) methodology for many years before AV-Test and AV-Comparatives started doing that. Not only that but he was doing real-world tests as a one-man shop as opposed to AVT or AVC who have 20 to 40 man orgs. I see it more of a limitation of resources than a trust issue. Blame PCMag.com corporate for not providing Neil with more resources. I'm certain if Neil had the same resources as a AVC or AVT his tests would rock.

 

I always though this

 

If this is true it actually brings up a bigger problem. If I were to take 100 samples today that I can confirm very poor AV detection on and then retest them even 1 week later every big name AV is going to nail them. If there are many AVs in this test not detecting after half a year then these samples are confirmed FPs/legit adware/damaged samples and wont ever be detected for the correct reasons.

 

I don't know about Outpost,but don't they tested just Comodo AV and not the whole suite?

 

yes you right; Comodo AV is mentioned not the IS, unless they made a typo mistake. anyway Comodo AV is not as bad as before, it is not perfect but becomes more and more efficient.

 

I guess but the fact of the matter is Neil is presenting data to the public that has been gathered from a fundamentally flawed test model.

Please tell me as an industry expert(from Panda AV) whether you consider this an acceptable practice ?

Just my 2cents but lack of resources does not in anyway justify the presentation of flawed data.If Neil does not have the time/resources or expertise for accurate and unbiased testing he should stick to product reviews and leave testing to the experts.

 

I didn't say that, did I?

 

For clarification, two charts are up at PCMag.com. One for BLOCKING, one for REMOVAL.

1. http://www.pcmag.com/image_popup/0,1871,iid=303596,00.asp

2. http://www.pcmag.com/image_popup/0,1871,iid=303597,00.asp

 

Comodo AV still comes with Defense+ unless you disable that.

 

yes i know, i thought they disable it if not rather test CIS