Clutch plays exist. But do clutch players?
certain players will exceed more often then not when the game is on the line. I will concede it is vague statement; but again, there will never be a valid away to quantify it
That's the core contradiction of the guy's argument. It's not that hard to figure out "when the game is on the line" and then quantify how the player did in those situations and compare it to his normal level of performance and what other players do in those situations. If clutch is a true ability and they really do "succeed more often than not," such an analysis would show it. But are the likes of Ortiz and Jeter raising their games in those situations or are they just the same great hitters they always are? Guys who can't handle pressure don't get to the top of the mountain.
The number of true game-on-the-line situations, as opposed to more generic clutch situations that we have shorthands for like RISP, close-and-late, etc., is sufficiently small that the sample-size argument can be invoked against drawing a conclusion one way or the other. I think that's what Bill James argued in a followup to his earlier work -- he said we may never be able to prove clutch, but we can't necessarily disprove it either.