Because Leela (like fastchess mentioned above) has two parts: A neural network predicting good moves, and a tree search exploring the moves suggested and evaluating the resulting positions (with a second net).

If the prediction (policy) net had a 100% accuracy, you wouldn't need the tree search part at all.

Got it. Thanks for clarifying. Let me restate.

Part one of Leela ranks several chess moves. Part two picks among those.

60% of the time part 2 chooses the #1 ranked move.

That works :-)

One addition: The second part can be run for an arbitrary amount of time, gradually improving the quality of the returned move.

The 60% figure comes from the training games, which are played very quickly, and so don't have a lot of time for refining, thus increasing the prediction accuracy.

In real games, tcec-chess.com/ this "self accuracy" would probably be a bit lower.

You haven’t mentioned any nondeterministic behaviour, therefore Leela is supposed to predict it’s own moves with a 100% accuracy.

It's not non-determinism, it's partial information. The NN part guesses the best move that will be found by search X% of the time. If you just ditched the search part, Leela would be faster and lose out on (1-X)% of the better moves.