http://chart.av-comparatives.org/chart1.php?chart=chart6&year=2016&month=9&sort=1 
FP-28
https://www.av-comparatives.org/wp-content/uploads/2016/10/avc_fdt_201609_en.pdf-99.9% 
Real- World protection tests http://chart.av-comparatives.org/chart1.php?chart=chart2&year=2016&month=9&sort=1&zoom=2
They have the numbers wrong.
They count FileRep and Win32:Evogen (SUSP) as FP while it they are not.
Saying something is not downloaded/used much is not a FP.
Saying something is suspicious is not a FP.
So the test shows that avast had 0 FP’s and not 28 as they claim.
And how do they calculated the presumed number of users affected ?
The amount of users that are likely to use one of the samples ?
The amount of users a av has ?
or…
They better fix the security problems/risks they are having.
https://www.ssllabs.com/ssltest/analyze.html?d=www.av-comparatives.org
http://retire.insecurity.today/#!/scan/003ebd2a28c8f23a5589a61593e0e31997c9ae55607ca9291b2c281fa8776ee6
So the test shows that avast had 0 FP's and not 28 as they claim.1. > Win32:Dropper-gen [drp] 1. > Win32:Malware-gen
so not 0 FP
I overlooked those two, but still not even near the 28 FP’s as they claim.
I disagree, I get those EVO-GEN & FILEREP False positives ALL THE TIME. And they’re EXTREMELY ANNOYING, they disrupt my work and/or delete clean files, or break downloads and I have to recover them from the vault and/or disable avast so I can finish what I am doing. THEY ARE FALSE POSITIVES, don’t try to deny it.
I don’t know where they always manage to pull these numbers from, but avast! has by far the least FP’s I’ve ever seen. It’s one of main reasons why I’m using it. Qihoo 360 on the other hand always receives good scores in tests, but whenever I tried it I had so many false positives I just always ended up uninstalling it. And same goes for many other that “beat” avast! here.
I think in the case of these FileRep and Win32:Evogen (SUSP) detections avast needs to give more detailed information on the screen as to what they actually mean. So as not to scare the horses so to speak.
Previously on detections there has been a more information link, unfortunately that was almost worse than no information. e.g. Avast has saved your ass, etc. but no actual useful information on the detection/malware name/type.
+1 Avast Evo-Gen sucks. 8)
I wouldn’t say that Evo-Gen sucks. It’s a very powerful feature that maybe needs more refinement. Maybe recent acquisition of AVG will in some way help that with vastly larger user base and also their machine learning capabilities.
jvidal, they are not false positives as nothing harmful is detected.
avast is just saying “Hé look, there is a file that I haven’t seen much (yet)”
A car is not bad just because not many people are buying/driving it.