From a Scientific/stats point of view, his "tests" are quite meaningless, there's no scientific repeatable methodology, the viewer can't tell the force being applied with the mallet nor the angle the knife is being struck, whether its going through knots in the wood, the hardness of the metal the tip is going through, whether the particular knife he is using has a poor heat treat or is in the 95th percentile of heat treats by that particular maker, or whether theres a bubble in the steel. Without out proper, repeatable methodology and a sample size large enough to have significance these aren't "tests" but rather entertainment to some, pointless destruction to others. I've watched almost all his videos, I find them entertaining, but nothing else. Yes, it can be compared IMO to a crash test of a car, but are you going to drive a car that has a test sample size of one? The only car we tested, the driver survived one head on collision, the car's safe, put it out on the market. Another example, you meet one oregonian who's a total ass, would you then assume everybody from oregon is a total ass, and that because the one californian you met was awesome, all californians are awesome? Large sample size= more reliable information.