Brilliant Move

@liam-beckettla said in #1:
> Hello, I thinks it would be a good idea for lichess to add a brilliant move in the computer analysis section, kind of like what Chess.com has now.
Go on chess.com if you want.

Brilliant moves could be converted in centipawns, by using the median value of the game.
If the median centipawn of the game was 25 centipawns, then any move below the median would be a briliiant move.

White cannot be getting brilliant moves if it already has the advantage !
Blacks brilliant move that gained equality or won the game could !!

A brilliant manouver is when the odds have turned around and it was not helped by errors from the opponent.
Having an overall comment of a game is more valuable than searching for the brilliant moves of a game.

Example: "Average Centipawn loss: 25. White played brilliantly with a centipawn deviation of: 25.6
That game was better than 85.0% of our analysed Blitz games."

So a KR versus KR end game where every move has zero centipawn loss would be a brilliancy for both players?

No game gets turned around without errors from the opponent. Chess is drawn. Nobody wins games unless the other people blunders.

Nobody has any idea how to define brilliancy so that it only detects moves where the person was thinking brilliant thoughts before making it.

Everyone is defining brilliancy so that it might call moves by people playing their first game of chess (or moves played by stockfish level 1) a brilliant move. What's the point of doing that?

Take a good look at this game and see the turning point without blunder, errors or imperfections in the middle of the analysis graph. Sure there are other faults during the game. The brilliance is when you can turn the game around and not when the game is already in your favor.

Look at move : 21 to 26. The tables turned and nothing was indicated as imperfections, errors or blunders.



There are so many games like this that others have created.

This is where the brilliant move makes the difference. The player making a draw or a fortress to survive the round are also brilliant moves.

The best move for an engine 10 deep may not be the best move for an engine 20 deep and it could even change at 30, 40, 50 deep. Sometimes short term plans are better than long term plans. The engine does not know when to make a short term plan and when to make a long term plan. The engine might find a quicker increase of material in one line, but is the line safe and fool proof from second move responces or others future inperfections?

An unexpected second best move from the opponent could disrupt the 50 move deep line.

Blunder/Brilliant or Brilliant/Blunder are inversely proportional. The middle of the scale must be perfect 0 centipawn moves.

Example: (-25 CP = Blunder) ....... (0 CP) ........ (+ 25 CP = Brilliant)

There is nothing brilliant about a zero centipawn move especially when it's a forced move.
So for the move to be brilliant it must be a strong move. You have to gain the upper hand. Turning th table around may need a sound sacrifices that gains the initiative and forces the opponent to play reactively.

The link has a definition ....
en.wikipedia.org/wiki/Chess_annotation_symbols#!!_(Brilliant_move)

Brilliant is a quality move, that has a long term gain, rather than short term tactical gain.
The player is at a loss and must find the means to solve the problem by a clever long-term chess strategy that will steer the game in their favor.

This makes it very hard to use centipawn values to see a long-term plan. Engines probably still don't see the horizon of a game. They don't know what you will play next and probably assume you will play the best move which in the long run might be favorable for the engine.

If you are happy with you brilliant moves, keep them to yourself, cause there will always be someone waiting to pop your bubble or criticize that it's not a brilliant move or not as much as another players game.

The best thing that could happen is to never see in chess notations the marks of brilliant "!!". Imagine an AI saying that a brilliant move does not exist. Each move has it's value, nothing brilliant about one being worth more than another.

Sarcasm: Now that was a brilliant blunder !!

I had a centipawn loss of 1, after 41 moves. I've never had that before.

@derekw said in #46:
> I had a centipawn loss of 1, after 41 moves. I've never had that before.

Congrats. I'm also always happy to get low CPLs. However, it is much easier to achieve this, if the opponent blunders heavily early and plays on.

I.e., there needn't be a single brilliant move in the game for that, just a blunder (e.g. the queen, just as in the example).

The brilliant move would be the only move the keep the current evaluation of the position, while being hard to find. Taking the queen for free wasn't hard to find. Therefore, it can not be labeled a "brilliant" move.

If the game was drawn, such a CPL would be much more impressive, as not even Carlsen gets this often.

@eclipsR said in #3:
> Something ive never done
120???

@eclipsR said in #48:
> 120???
That's not enough, lets make it to the million territory

I think yes, it is a good idea. But I may say a good move can also be added.