Why would I do anything? Whether it's *possible* that Mews turns out better than Willander (of course it is) is not the same thing as whether it's *likely* that Mews tuns out better than Willander, which is a ludicrous statement. .
I could go and create a system to rank the Best NHL Players taking various stats and multiplying by various ratios to get a number for each player. And if it spits out that Kiefer Sherwood is a better player than Auston Matthews, it means that I've created a Very Bad Model where I've weighted hits incorrectly and I should go back to the drawing board.
Likewise, if you devise a prospect rating system and it spits out that Henry Mews is a better prospect than Tom Willander, you've created a bad model. End of story.
Of course posters here are wrong all the time. Even the best NHL drafting teams only get the best player taken within a 10-pick range of their selection maybe 20% of the time. That doesn't mean that bad math is somehow clarifying anything.