The thing is: for what you say is happening to actually happen - that would REQUIRE the developer to implement some mid-level-AI-functionality into their game to analyze what's good and what's bad for one player and to process player-information to find out which player to favor. That's many-many hours of work to do. So why would that be happening? Why so much effort? To advantage one random dude over some other random dude? Why?
Good players (and that is universal to many games and not just Marco Polo) play in a way that is flexible towards randomness. They set up themselves in a way that more random events are good outcomes for them and less random events are bad outcomes for them. When they are ahead, they take less risks. When they are behind, they take risk in order to at least have a chance to get lucky.
So yes - they do get more favourable outcomes than worse players. But that's not because the randomness favors them but because they set up their game-position well.
Good players (and that is universal to many games and not just Marco Polo) play in a way that is flexible towards randomness. They set up themselves in a way that more random events are good outcomes for them and less random events are bad outcomes for them. When they are ahead, they take less risks. When they are behind, they take risk in order to at least have a chance to get lucky.
So yes - they do get more favourable outcomes than worse players. But that's not because the randomness favors them but because they set up their game-position well.