Can it be absolutely quantified? I haven't seen many but to me they always seem to be outside the bounds of traditional back and forth aggression / counting principles. Like putting a piece where traditionally it wouldn't obey the expected threat / counter threat / attack / defense / pin / gambit but something that attacks or defends in a more special, situational way. Especially when under traditional math it would be a bad move, like something that would otherwise be underdefended, but in that specific situation considering the larger scenario it's the best move. I could be incorrect.
What is a brilliant move?

The engine's definition of a brilliant move is not the same as our plain-language definition.
To the engine, a move is brilliant if it gives a better result in all variations than the move that the engine had considered to be "Best".
How would that happen? It's because the engine can only calculate a certain number of half-moves ahead, so some results are hidden "beyond the horizon" until more moves are played and the actual state of the game becomes clearer.

To the engine, a move is brilliant if it gives a better result in all variations than the move that the engine had considered to be "Best".
How would that happen? It's because the engine can only calculate a certain number of half-moves ahead, so some results are hidden "beyond the horizon" until more moves are played and the actual state of the game becomes clearer.
Really? That would seem like a weak engine, to me. This is the only game I know where I got a "brilliant move" (as a free account I have limited access to analysis) and it was part of the standard response to the fried liver attack:
My move 9 as black. I would have thought the chess.com engine would either know the fried liver defense or would be able to look far enough in to see the value of this move?
Hello people, I was just wondering what classifies as a brilliant move. How does the engine determine what is brilliant and what is just a best move?