By definition: a rating system difference of 200 points is wrong 25 percent of the time.
Actually it's wrong 100% of the time. The higher rated player has an expectation of .75 which never happens in a single game. Which is about the same as saying a broken watch is more accurate than a watch that loses a 1/365 of a second each day. The broken watch is correct twice a day, while the other watch is correct once every 8,640 years.
I would also argue that the game of chess doesn't exist without humans to recognize it as a game of chess. Thus, this 'pure' chess that supposedly exists independent of the humans who play it is a fallacy. As humans are subject to a seemingly infinite number of 'outside circumstances' that affect their mental state at the moment of play, chess seems almost entirely a game of luck.
Edit add: certainly it seems to function much as a non-linear system with a great deal of chaos within a finite number of choices. Perhaps the best great chess players can hope for is to minimize the chaos within the game.
Craps is basically all luck.I can't think of too many games that are so luck oriented.
Backgammon has more skill required still, but the dice causes there to be enough luck involved, it has as much chance at changing the outcome of a game, as skill or lack there of, ever does.
Chess on the other hand, is almost completely skill oriented. You can't just push random pieces, while your opponent has a strategy and practically think you have a chance to win. You will lose everytime.
Ok,ok ... I'll give the odds the chance that you might pick enough best/really good moves, against a competent opponent, to actually have a realistic chance to win. You can waste your life trying to prove me wrong and I might actually win the lottery if I start playing now before I die.
Hence, chess isn't a game of luck, though luck, whether good or bad, can have it's effect over the outcome, but much less often than a game like Backgammon. The more skilled player is almost always going to win.