Brilliant Move

Honestly it is a good idea but not the best addable feature

I just want to add my two cents, once on Chesscom I saw it evaluate taking a queen that an opponent very obviously hung as "brilliant". But, I suppose taking it was the only good move in the position, but no human player with more than 10 seconds on the clock would have missed it, let alone call it "brilliant".

In the meantime, I'll continue to consider that my moves are brilliant when they make my opponents immediately resign.

@Tribe11 said in #25:
> I just want to add my two cents, once on Chesscom I saw it evaluate taking a queen that an opponent very obviously hung as "brilliant". But, I suppose taking it was the only good move in the position, but no human player with more than 10 seconds on the clock would have missed it, let alone call it "brilliant".
>
> In the meantime, I'll continue to consider that my moves are brilliant when they make my opponents immediately resign.
200 rated players exist

@JakeLiewChuen said in #26:
> 200 rated players exist
Of course, but a brilliant move shouldn't be subjective based on your rating, just like a blunder isn't. The point is that even on chesscom, the algorithm for deciding if a move is brilliant doesn't always coincide with what a human being would call brilliant.

It has.

An interesting (and obviously far from perfect) way of defining Brilliant Move could be a move that was at the same time evaluated as BAD by Stockfish at depth 12, for example (!! just an example, guys), and evaluated as GOOD at the usual 17-18 depth analysis. Afaik a depth 12 is eventually reached during the depth-18 analysis, so *in principle* we wouldn't be consuming sensibly more processing and energy with this approach, however, hacking Fishnet to include it might be a little bit too ambitious without a reasonable ground :>
(And btw this topic came to my attention exactly because @thibault was here xD)

A chess employee explained brilliant move in forum. It's a move, engine considered not as the best with a low analysis deep immediately and then recognizes, it was the best for the position evaluation. Many people there post their games with one or more brilliant moves, mostly you could in theory start a long combination to win material or checkmate opponent, which is for most players impossible to recognize.
But I think its not reliable, many games were posted, in which brilliant moves later in deep analysis even were mistakes.

My only brilliant I got was in this game (I am white)
[Event "?"]
[Site "?"]
[Date "2021-11-25"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[App "ChessIs Android app"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 {Opening: Ruy Lopez (Spanish Opening)} a6 4. Bxc6 Qf6
5. Ba4 *

My friend was dreaming that day, lol, my bishop survived. Of course Ba4 is the best move, but not brilliant.

I guess this feature is just a gimmick which mostly is useless.