Hmm, I think it's unrealistic to look at it that way. As an analogy, you can take chess.
One of the best AI chess engines at the moment is called stockfish (ELO rating Approx ~3.5k). It beats any human player by a significant margin but even then, the top players in the world (Grandmasters have ratings of ~2.5k) can still beat it 3-5 times every thousand games. The point being, even the best of the best can lose a game here and there for whatever reason.
I think it's more pragmatic to measure your success in those terms, scaled down of course, and see if you can regularly come out on top at the stakes you play.
__________________
|