lichess.org
Donate

Explainable chess analysis

@BananaBeaver said in #6:
<1/ and 2/>
> Note how 2/ combines with 1/ : you also want to know if there is another threat...

Thanks, @BananaBeaver for the tip! This will help see obvious threats but for a long and/or deep plan, IMO we need to combine SF analysis and its visual hints to understand the ideas.

Anyway, by "Explainable chess analysis" I meant something easy to understand, such as

(Examples)
- The idea is to attack weak pawn at c3
- The plan is to pin the knight at e5 while trying to overload the isolated pawn at d5
@Professor74 said in #8:
> But what about the other valid continuations?
There is more than one good/best move at any time.

Human thinks differently from engines.

Can I ask what model have you used and how you trained it?

> I sincerely suggest you seek professional chess advice for these analysis.

We have a GM coach who helps us with analysis on a weekly basic.
The problem is, I can't reach out for help anytime and my question is if there is a tool, a service, or a way to help us with that.
@Toscani said in #7:
> If the position is novelty, than the only person that can say the plan is the person that played the move. Chess is not solved.

@Toscani

Have you tried decodechess?
Though it is still in the infancy stage, I support its idea.
@vuhung said in #12:
> There is more than one good/best move at any time. Human thinks differently from engines. Can I ask what model have you used and how you trained it?
> We have a GM coach who helps us with analysis on a weekly basic. The problem is, I can't reach out for help anytime and my question is if there is a tool, a service, or a way to help us with that.

Thanks for fueling the discussion. We all know there are many search schemes using chess engines. There are various types of simulation models using hundreds of engines on different platforms. On the other point you touched on, I suggest you focus on taking advantage of the advice your GM coach provides you. Not many chess players are fortunate enough to have a GM as a coach. I wish you success!
There are quite a number of Move by Move Chess Books & watching Chess Videos helps also ... These explain every move By Irving Chernev Logical Chess Move by Move ... @vuhung Neil Mcdonald The Art of Logical Thinking in Chess ... Understaning Chess Move by Move by John Nunn
@ThunderClap said in #16:
> There are quite a number of Move by Move Chess Books & watching Chess Videos helps also ... These explain every move By Irving Chernev Logical Chess Move by Move ... @vuhung Neil Mcdonald The Art of Logical Thinking in Chess ... Understaning Chess Move by Move by John Nunn

@ThunderClap

Thanks for the comment.

I am looking for a tool that explains any situation, not only openings,
which probably only done by not-yet-developed tool.
@Professor74 said in #14:
> Thanks for fueling the discussion. We all know there are many search schemes using chess engines. There are various types of simulation models using hundreds of engines on different platforms.

Can you please shed some keywords on search schemes and simulation models?

> On the other point you touched on, I suggest you focus on taking advantage of the advice your GM coach provides you. Not many chess players are fortunate enough to have a GM as a coach. I wish you success!

Thanks for the suggestion! Yes, we do.
@vuhung said in #17:
> @ThunderClap
>
> Thanks for the comment.
>
> I am looking for a tool that explains any situation, not only openings,
> which probably only done by not-yet-developed tool.
@ThunderClap said in #16:
> There are quite a number of Move by Move Chess Books & watching Chess Videos helps also ... These explain every move By Irving Chernev Logical Chess Move by Move ... @vuhung Neil Mcdonald The Art of Logical Thinking in Chess ... Understaning Chess Move by Move by John Nunn Every single move EXPLAINED from the opening middlegame endgame ... If you are asking for an AI that will exist in the future then in the meantime' !? Read a Chess Book or10
<Comment deleted by user>

This topic has been archived and can no longer be replied to.