-=> Mocking Richard to Rick <=-
  (Mock, mOck, moCk, mocK)
 RSJ> Actually the point is that they have subjected X number of virus
 RSJ> scanners to the same series of tests. They then presented how the
 RSJ> tests were performed and the information gleaned from those tests in
 RSJ> numbers. Their testing may not have sent EVERY virus against these
 RSJ> scanners, but in the test sequences {that they posted along with the
 RSJ> kind/number of each virus} you can determine how each virus scanner
 RSJ> reacted to the test scenerio. These test results are readily available
 RSJ> on the net for anyone to view.
i disagree... the actual point is that very few people out there have
any kind of reasonable grasp of statistical analysis and would take a
test like virus bulletins to be just as informative as one that used
50-75% of all known viruses... the basic problem with the virus bulletin
test is that they have a huge degree of uncertainty that they don't even
try to represent... in a test using 400 viruses when there are 8,000
known, a scanner that detects 80% of the 400 samples can have an actual
detection rate anywhere between 99% and 4%... there most certainly
wasn't enough information in the publication for the average user to
make that calculation even if s/he did have the requisite knowledge of
statistics... 
... ow that you know that i know that you know that i kn...
                   
--- Maximus 2.02
---------------