Stop trusting AI so much. You might as well just use magic-8ball.com
Why bring AI into this? Let's eliminate computers completely from the equation. You are watching a chess game on TV and suddenly Eric Rosen comments that a move was brilliant. Are you going to shout "bullshit!" ? See, no AI. The concept of a brilliant move IS REAL, because people use it to communicate something. Why not try to capture it, somehow?
That being said, does anyone know of a large database of serious games with professional annotations of the move quality?
That being said, does anyone know of a large database of serious games with professional annotations of the move quality?
ChatGPT @mcgoves said in #21:
> Stop trusting AI so much. You might as well just use magic-8ball.com
:)) Asked ChatGPT his feelings about this... Here’s the answer
The idea of implementing a “brilliant move” annotation system on Lichess sparks an interesting debate. Here’s an analysis of the pros and potential challenges:
⸻
Advantages of a Brilliant Move Annotation System
1. Recognition of Creative Play:
Such a system would highlight bold or unexpected moves, rewarding players for creativity.
2. Enhanced Learning:
Identifying and studying these moves can help players understand advanced tactical and strategic ideas.
3. Increased Engagement:
Recognizing brilliant moves could motivate players to explore more complex lines and invest more in their improvement.
⸻
Challenges and Considerations
1. Subjective Definition:
What is considered “brilliant” can vary from player to player and context to context, making standardization tricky.
2. Technical Complexity:
Developing an algorithm that accurately detects such moves would require significant resources and careful validation.
3. Lichess Philosophy:
Lichess promotes a clean, educational experience. Adding such features might shift the focus away from its core mission.
⸻
Future Perspectives
Recent research — including studies from the University of Toronto — is exploring how AI can identify creative chess moves.
(Source: echecs-et-strategie.com)
These advances could eventually make it easier to integrate a brilliant move system into platforms like Lichess.
⸻
Conclusion
Integrating a brilliant move annotation system on Lichess could enrich the user experience by highlighting creativity and offering additional learning opportunities.
However, it requires thoughtful design to ensure it aligns with the platform’s core values.
> Stop trusting AI so much. You might as well just use magic-8ball.com
:)) Asked ChatGPT his feelings about this... Here’s the answer
The idea of implementing a “brilliant move” annotation system on Lichess sparks an interesting debate. Here’s an analysis of the pros and potential challenges:
⸻
Advantages of a Brilliant Move Annotation System
1. Recognition of Creative Play:
Such a system would highlight bold or unexpected moves, rewarding players for creativity.
2. Enhanced Learning:
Identifying and studying these moves can help players understand advanced tactical and strategic ideas.
3. Increased Engagement:
Recognizing brilliant moves could motivate players to explore more complex lines and invest more in their improvement.
⸻
Challenges and Considerations
1. Subjective Definition:
What is considered “brilliant” can vary from player to player and context to context, making standardization tricky.
2. Technical Complexity:
Developing an algorithm that accurately detects such moves would require significant resources and careful validation.
3. Lichess Philosophy:
Lichess promotes a clean, educational experience. Adding such features might shift the focus away from its core mission.
⸻
Future Perspectives
Recent research — including studies from the University of Toronto — is exploring how AI can identify creative chess moves.
(Source: echecs-et-strategie.com)
These advances could eventually make it easier to integrate a brilliant move system into platforms like Lichess.
⸻
Conclusion
Integrating a brilliant move annotation system on Lichess could enrich the user experience by highlighting creativity and offering additional learning opportunities.
However, it requires thoughtful design to ensure it aligns with the platform’s core values.
If I were mean I would focus on this paragrapb:
> Developing an algorithm that accurately detects such moves would require significant resources and careful validation.
But I am a nice guy.
> Developing an algorithm that accurately detects such moves would require significant resources and careful validation.
But I am a nice guy.
@TotalNoob69 said in #22:
> Why bring AI into this?
I don't know. But you did.
> Let's eliminate computers completely from the equation. You are watching a chess game on TV and suddenly Eric Rosen comments that a move was brilliant. Are you going to shout "bullshit!" ?
I won't. But a GM might. And heck, Eric might call a good move by a beginner brilliant, but not the same move by a club player. Not only is brilliancy dependent on the expert, but also on the player.
> Why bring AI into this?
I don't know. But you did.
> Let's eliminate computers completely from the equation. You are watching a chess game on TV and suddenly Eric Rosen comments that a move was brilliant. Are you going to shout "bullshit!" ?
I won't. But a GM might. And heck, Eric might call a good move by a beginner brilliant, but not the same move by a club player. Not only is brilliancy dependent on the expert, but also on the player.
> @corvusmellori said in #9:
> So, would 1. b8=Q be a BRILLIANT move in this position? The evaluation gain jumps at depth 8.
> lichess.org/analysis/standard/8/1P6/5k2/8/8/8/p4K2/8_w_-_-_0_1
I think you correctly understood the general idea.
Depth 8 was just an example. I didn't want to prejudge what an appropriate depth level would be. There have been suggestions to set the depth level according in relation to ELO points (lower depth level for lower rated players and higher depth level for higher rated players).
> So, would 1. b8=Q be a BRILLIANT move in this position? The evaluation gain jumps at depth 8.
> lichess.org/analysis/standard/8/1P6/5k2/8/8/8/p4K2/8_w_-_-_0_1
I think you correctly understood the general idea.
Depth 8 was just an example. I didn't want to prejudge what an appropriate depth level would be. There have been suggestions to set the depth level according in relation to ELO points (lower depth level for lower rated players and higher depth level for higher rated players).
> @TotalNoob69 said in #10:
> We could discuss possibilities and maybe try to implement some. Refusing to discuss it is not productive at all and it proves nothing. Rather than fighting to prove others wrong, let's focus on how they could be right.
@TotalNoob69 : Thank you for chiming in and bringing some positivity into this conversation. IMO this was the first nice and positive contribution to this discussion (aside from mine of course ;-P).
To others here: I really don't like the general tone here, badmouthing an idea without any arguments, just saying it's "terrible", etc. How about arguments ?
IMO it's completely obvious that the idea is can hardly be "terrible" if leading chess websites have implemented it.
I also find it disingenuous to allege that no definitions have been presented or even that definitions are outright impossible. Both is wrong.
I have presented my own suggestion for how to define a brilliant move.
My definition is 0% subjective and 100% objective.
And I have posted a link to the definition on chess.com .
So please stop dealing in falsehoods. As @TotalNoob69 said: Of course you can say you dislike the idea. But this is your personal subjective opinion. Don't act as if you owned the truth. Thank you.
I would appreciate a more grounded, technical discussion.
In particular, I would appreciate suggestions to improve the definition of the move classification.
(Everyone here focuses on brilliant moves, but remember that there are many additional categories not yet discussed here.)
> We could discuss possibilities and maybe try to implement some. Refusing to discuss it is not productive at all and it proves nothing. Rather than fighting to prove others wrong, let's focus on how they could be right.
@TotalNoob69 : Thank you for chiming in and bringing some positivity into this conversation. IMO this was the first nice and positive contribution to this discussion (aside from mine of course ;-P).
To others here: I really don't like the general tone here, badmouthing an idea without any arguments, just saying it's "terrible", etc. How about arguments ?
IMO it's completely obvious that the idea is can hardly be "terrible" if leading chess websites have implemented it.
I also find it disingenuous to allege that no definitions have been presented or even that definitions are outright impossible. Both is wrong.
I have presented my own suggestion for how to define a brilliant move.
My definition is 0% subjective and 100% objective.
And I have posted a link to the definition on chess.com .
So please stop dealing in falsehoods. As @TotalNoob69 said: Of course you can say you dislike the idea. But this is your personal subjective opinion. Don't act as if you owned the truth. Thank you.
I would appreciate a more grounded, technical discussion.
In particular, I would appreciate suggestions to improve the definition of the move classification.
(Everyone here focuses on brilliant moves, but remember that there are many additional categories not yet discussed here.)
<Comment deleted by user>
@weeknights said in #26:
> I think you correctly understood the general idea.
> Depth 8 was just an example. I didn't want to prejudge what an appropriate depth level would be. There have been suggestions to set the depth level according in relation to ELO points (lower depth level for lower rated players and higher depth level for higher rated players).
Why does who is playing the move matter for terms of brilliant it does not for blunders?
It allows 1500s to brag saying i got a brilliant while making it impossible for 2200 to get it. also makes them not consistent.
the other problem is depth is a terrible choice as some openings require the engine to go deeper else the elo bar is way off.
Traxlar is a good example where Bc5 is seen like a blunder until 40 depth or something.
> I think you correctly understood the general idea.
> Depth 8 was just an example. I didn't want to prejudge what an appropriate depth level would be. There have been suggestions to set the depth level according in relation to ELO points (lower depth level for lower rated players and higher depth level for higher rated players).
Why does who is playing the move matter for terms of brilliant it does not for blunders?
It allows 1500s to brag saying i got a brilliant while making it impossible for 2200 to get it. also makes them not consistent.
the other problem is depth is a terrible choice as some openings require the engine to go deeper else the elo bar is way off.
Traxlar is a good example where Bc5 is seen like a blunder until 40 depth or something.