"But there is clearly a strategy to confuse the opponent"
Yeah. Psychology. This is perhaps an example of why I don't always find it easy to take the luck people seriously. It's not even that they're wrong -- probably no game can reward a person's skill 100% proportionally since it can't know whether you "intended" to play the right move. But it seems like a lot of these people will just decide something is luck when really it's just a type of strategy. Even breathing on the opponent, not something you should do, that's not luck, that's a conscious choice made to try to win I guess. Trying to confuse the opponent, while not a purely scientific approach, is as much a strategy as anything. But I guess when the strategy isn't in the right format a lot of people want to just call it luck. Or if people don't like why they lost, then that reason is now luck. Other reasons they are ok with, but if some move pissed them off, now it's luck.
Simply not expecting something to happen doesn't in itself mean anything since humans can choose to feel that way whenever they want. I was surprised that my opponent played a line I studied last night... ok? Cool? No one really cares? I hope you can sleep at night despite being surprised? Your hands still made the moves; you still used your knowledge base. You may have felt surprised too; maybe you were hungry; maybe you were lonely... I'm not really sure why I would care. We can just "decide" that we are interested in these things enough to put them into the pile of "luck," yet we are putting in conscious decisions into this pile alongside with mindless coin flips, which seems rather strange.
I mean, I don't know if I will win the next chess game I play; does that mean if I think I will lose but then I win that it was all chance and probability? If I think it's luck does that make it luck? :)

Probability regarding results and impersonal factors in a sample of chess players is entirely different from luck in a chess game. If he was talking about probability distributions and referred to it as the sort of luck one experiences in a game of chess, he was talking gobbledigook. Whether or not luck exists in chess on an individually specific basis is surely the only question that concerns us, since probability distributions are just depictions of samples and if he really meant that, it's non-contentious and entirely6 uninteresting.