One issue I see is.. what exactly is a brilliant move?
In the currently used system, the best move leaves the evaluation unchanged. A good or excellent move drops it slightly, an inaccuracy drops it to a noticable amount, and mistakes and blunders drop it even more. This is easy to calculate, by just viewing the centipawn loss.
Brilliant moves however are something else; they are the best move according to the engine, but one that's hard for a human to find. A computer can tell if a move loses centipawns or not, but as of now there is no good way for computers to easily find if a move is hard to find for humans or not. Besides, a move that looks brilliant to a 1500 might be completely natural to a 2500.
I see many have asked you to write the code yourself. This could be because no one is willing to write the code for this, though I could just be completely wrong here...
One issue I see is.. what exactly is a brilliant move?
In the currently used system, the best move leaves the evaluation unchanged. A good or excellent move drops it slightly, an inaccuracy drops it to a noticable amount, and mistakes and blunders drop it even more. This is easy to calculate, by just viewing the centipawn loss.
Brilliant moves however are something else; they are the best move according to the engine, but one that's hard for a human to find. A computer can tell if a move loses centipawns or not, but as of now there is no good way for computers to easily find if a move is hard to find for humans or not. Besides, a move that looks brilliant to a 1500 might be completely natural to a 2500.
I see many have asked you to write the code yourself. This could be because no one is willing to write the code for this, though I could just be completely wrong here...
given how many times this request has been rejected that very few people are interested. Just idea keeps popping up as new players show in site.
given how many times this request has been rejected that very few people are interested. Just idea keeps popping up as new players show in site.
For 10 years I have been asking for definitions for "complexity" and "brilliant move" and I still don't have a definition about what we are talking about. Actually, no human moves are brilliant:
https://youtu.be/e47VomRFhAU
For 10 years I have been asking for definitions for "complexity" and "brilliant move" and I still don't have a definition about what we are talking about. Actually, no human moves are brilliant:
https://youtu.be/e47VomRFhAU
One additional move annotation that could be useful is an annotation to indicate that, according to Stockfish, the move that was played is the best move in the position AND substantially better than all other possible moves in the position.
One additional move annotation that could be useful is an annotation to indicate that, according to Stockfish, the move that was played is the best move in the position AND substantially better than all other possible moves in the position.
@MartianBlitz said ^
One additional move annotation that could be useful is an annotation to indicate that, according to Stockfish, the move that was played is the best move in the position AND substantially better than all other possible moves in the position.
Can you show us an example of any other software product which does this? Also, what about obvious moves?
@MartianBlitz said [^](/forum/redirect/post/vrjcByIr)
> One additional move annotation that could be useful is an annotation to indicate that, according to Stockfish, the move that was played is the best move in the position AND substantially better than all other possible moves in the position.
Can you show us an example of any other software product which does this? Also, what about obvious moves?
The brilliant move is a strategic turning point, like a rainbow that appears and disappears in the sky.
Maybe the community should try experimenting by building a python script to find these chess moments, search for swingy positions where one move changes the evaluation sharply. Blunders are the inverse of brilliance because they also cause sharp swings in evaluation. So I believe a brilliant move produces a large positive swing, while a blunder produces a large negative swing.
How many brilliant moves do you think could exist in a game? Probably as many as there are blunders. Assuming the blunders are punished.
Conversion and Resourcefulness: How often do you come back from lost positions. Well I would assume that takes a brilliant move or a blunder on the part of your opponent and you have to be lucky enough to find the brilliant moves. Are you "resourcefulness" enough to create a brilliant move? Well, in our chess tutor/insight the question is answered.
Don't go scoring brilliant moves by whether it is a pv 1 move. Score it by how much the evaluation changes because of that move. Engine play doesn't do large cp swings, but humans do. Brilliance is about catching the moment. Seeing a pot of gold at the end of the rainbow.
The brilliant move is a strategic turning point, like a rainbow that appears and disappears in the sky.
Maybe the community should try experimenting by building a python script to find these chess moments, search for swingy positions where one move changes the evaluation sharply. Blunders are the inverse of brilliance because they also cause sharp swings in evaluation. So I believe a brilliant move produces a large positive swing, while a blunder produces a large negative swing.
How many brilliant moves do you think could exist in a game? Probably as many as there are blunders. Assuming the blunders are punished.
Conversion and Resourcefulness: How often do you come back from lost positions. Well I would assume that takes a brilliant move or a blunder on the part of your opponent and you have to be lucky enough to find the brilliant moves. Are you "resourcefulness" enough to create a brilliant move? Well, in our chess tutor/insight the question is answered.
Don't go scoring brilliant moves by whether it is a pv 1 move. Score it by how much the evaluation changes because of that move. Engine play doesn't do large cp swings, but humans do. Brilliance is about catching the moment. Seeing a pot of gold at the end of the rainbow.
@TMCA_JIVIN_VARGHESE said ^
I still question why people want brilliant moves so badly...
They want to flatter the brilliant moves with other chess friends
Simply no. You haven't read the full request, nor bringing anything to the table. Move on please.
@TMCA_JIVIN_VARGHESE said [^](/forum/redirect/post/2bKCHaOj)
> > I still question why people want brilliant moves so badly...
>
> They want to flatter the brilliant moves with other chess friends
Simply no. You haven't read the full request, nor bringing anything to the table. Move on please.
Brilliance is not "did Stockfish choose this move?" Brilliance is "did this move cause a large favorable change in the position?"
Brilliance is not "did Stockfish choose this move?" Brilliance is "did this move cause a large favorable change in the position?"
@Toadofsky said ^
One additional move annotation that could be useful is an annotation to indicate that, according to Stockfish, the move that was played is the best move in the position AND substantially better than all other possible moves in the position.
Can you show us an example of any other software product which does this? Also, what about obvious mo...
I don't know of any software product that does exactly that. I think what chess.com classifies as "Great Move" comes close. But I would not label it "great move" or use an exclamation mark, because the method will include obvious moves, but maybe obvious moves can be filtered out.
@Toadofsky said [^](/forum/redirect/post/O8kropn7)
> > One additional move annotation that could be useful is an annotation to indicate that, according to Stockfish, the move that was played is the best move in the position AND substantially better than all other possible moves in the position.
>
> Can you show us an example of any other software product which does this? Also, what about obvious mo...
I don't know of any software product that does exactly that. I think what chess.com classifies as "Great Move" comes close. But I would not label it "great move" or use an exclamation mark, because the method will include obvious moves, but maybe obvious moves can be filtered out.
Let me quote a GM ....
https://www.youtube.com/watch?v=YxnWXIvoaTA