At high levels, if you play a trap line which your opponent HAS prepared for, you may lose. If they haven't prepared for it, you are more likely to win.
Your results depend on an unknowable element - whether your opponent has prepared or not. Therefore, chess (as played vs real opponents in reality) has randomness. I'm not saying the pure game has randomness. Obviously it doesn't. And games like poker have randomness even under perfect play, while with perfect play there is no randomness is chess.
Is this an invalid definition of random? Is randomness only really random if it is generated during the event?
My answer is no. Imagine you are playing poker - a random game, right?
Now imagine that the seed used in the RNG was secretly generated and stored a year ago. And you didn't know that, and you played a tournament thinking you were playing "random" poker hands. Now that you know the seed was old, do you now think that the poker you were playing was not random?
I doubt it. Therefore, randomness means "you don't have a way to know a hidden variable which partially determines your results in a competition". In the chess case that variable is the state of your opponent's preparation. Is duplicate bridge not random? The hands are preset and the same for every table (for comparability) - but from your POV they're still random.
Does it matter that you can sometimes figure out opponent preparation by looking at their past games? Well, imagine that occasionally (but not always) the dealer flashes the opponents cards. Does that mean the actual poker tournament is not random? No. It would only not be random if you had full knowledge of every hand, before the tournament. The fact that there are occasional leaks of secret information in both the poker and chess does not mean that they are not random.