Leo tested the detection of AV1 against 100 samples and 22 samples were undetected. Next, he used AV2, AV3, ... on undetected samples and we could see that those AVs detected a few samples from those undetected samples. He forgot that AV2, AV3, ... could also miss a few samples that were already detected by AV1. But, we did not have a chance to see if it could be true or not, due to an invalid procedure applied in the Leo test. The only useful information from the video is that the Norton Power Eraser tool (not an AV) is very efficient.
This is a good example of when the initial assumption about AV1 makes the test results irrelevant.
The av used for detection are more aggressive
Norton power eraser uses more aggressive huristics then max settings on Symantec and uses some that are exclusive to the scanner (Symantec diagnostics , Norton power eraser is on purpose built to find as much malware as possible)
Hitman pro is far more aggressive then sophos too
Obviously they would detect more and have more false positives then defender
With 100 samples I'm sure the super aggressive scanners would get better results so what as they aren't a realtime av for obvious reasons and if he tested sophos , Norton consumer products on default settings they should miss some samples just like defender
"Power Eraser scans and virus and spyware scans both run an active scan and a full scan.
However, Power Eraser uses the more aggressive High Intensity Detection (HID) scans that employ advanced machine learning (AML) techniques. The AML engine determines if a file is good or bad through a learning process.
The AML engine recognizes malicious attributes and defines the rules that the AML engine uses to make detectio
ns."
Source
Symantec Power Eraser is a free virus removal tool to remove malware and threats from your computer. Power Eraser is provided with the product and does not need to be installed.
techdocs.broadcom.com