[OhNoMyqueen]

> When watching the WC games, I've seen it happen that a move wasn't considered as a top move by the engine, but once played the engine realizes it's actually crushing.

You don't need to watch WC game to witness that. Any position sharp enough will have swinging evaluations at the search horizon, and even amateurs can find a twist in their favor if they're lucky enough (to have their 8-moves ahead flash of genius actually work 20 moves ahead by complete chance).

In decent engines the heuristics used to prune the search tree range from exact (e.g. alpha-beta prunning) to conservatively innexact (low chance to actually miss something), and unless you find a systematic winning or drawing pattern (like the examples in OP's article), you can't beat the engine with your "intuition".