Isn't this closer to the K-MAPS concept?
@TotalNoob69 said in #2:
> Isn't this closer to the K-MAPS concept?
Please explain as I am not sure what you mean ... is it related to some sort of logic map
> Isn't this closer to the K-MAPS concept?
Please explain as I am not sure what you mean ... is it related to some sort of logic map
@CMKingscrusher-YouTube can I join?
@Kingscrusher-YouTube said in #3:
> Please explain as I am not sure what you mean ... is it related to some sort of logic map
It was something I saw in an Anna Cramling video. I thought it was a more general concept. The idea being you evaluate a position by "K-MAPS" :
- King safety
- Material
- piece Activity
- Pawn structure
- Space advantage
I wanted to add a way to visualize this in LiChess Tools, but I never got to it.
> Please explain as I am not sure what you mean ... is it related to some sort of logic map
It was something I saw in an Anna Cramling video. I thought it was a more general concept. The idea being you evaluate a position by "K-MAPS" :
- King safety
- Material
- piece Activity
- Pawn structure
- Space advantage
I wanted to add a way to visualize this in LiChess Tools, but I never got to it.
@GiannaMuramira said in #4:
> @CMKingscrusher-YouTube can I join?
If you mean my team then sure thing.
> @CMKingscrusher-YouTube can I join?
If you mean my team then sure thing.
I fully agree. This is an excellent blog thanks so much!
@TotalNoob69 said in #5:
> It was something I saw in an Anna Cramling video. I thought it was a more general concept. The idea being you evaluate a position by "K-MAPS" :
> - King safety
> - Material
> - piece Activity
> - Pawn structure
> - Space advantage
>
> I wanted to add a way to visualize this in LiChess Tools, but I never got to it.
You have got me thinking with this question - many thanks :). In the opening stage of a game - the 1st 10 ply (half moves), we could consider it more of a boolean flag (YES/NO) for the particular elements of :
King safety FALSE (King unsafe because in center - e.g. two exploitable diagonals potentially to the King on e8 - a4-e8 diagonal and h5-e8 diagonal)
Central square control FALSE - the four central squares have not been controlled yet in the start position of chess
Development of pieces FALSE - pieces on the 1st row
But sure in the middlegame there are other elements generally to consider. But that that is not to say in some openings there are some very early pawn structure decisions.
I just think because those particular elements are FALSE in the opening stage of the game, they are really high priority to set to TRUE in effect. And then you are off to a decent middlegame where other elements are key.
In the endgame, the passed pawn elements become more important because of the probably of queening becomes so much higher. And the King safety priority generally drops because the probability of being checkmate reduces.
The stage of the game gives the context in general for the priority of the elements to consider.
> It was something I saw in an Anna Cramling video. I thought it was a more general concept. The idea being you evaluate a position by "K-MAPS" :
> - King safety
> - Material
> - piece Activity
> - Pawn structure
> - Space advantage
>
> I wanted to add a way to visualize this in LiChess Tools, but I never got to it.
You have got me thinking with this question - many thanks :). In the opening stage of a game - the 1st 10 ply (half moves), we could consider it more of a boolean flag (YES/NO) for the particular elements of :
King safety FALSE (King unsafe because in center - e.g. two exploitable diagonals potentially to the King on e8 - a4-e8 diagonal and h5-e8 diagonal)
Central square control FALSE - the four central squares have not been controlled yet in the start position of chess
Development of pieces FALSE - pieces on the 1st row
But sure in the middlegame there are other elements generally to consider. But that that is not to say in some openings there are some very early pawn structure decisions.
I just think because those particular elements are FALSE in the opening stage of the game, they are really high priority to set to TRUE in effect. And then you are off to a decent middlegame where other elements are key.
In the endgame, the passed pawn elements become more important because of the probably of queening becomes so much higher. And the King safety priority generally drops because the probability of being checkmate reduces.
The stage of the game gives the context in general for the priority of the elements to consider.
hii
@Kingscrusher-YouTube said in #8:
> You have got me thinking with this question - many thanks :)
...
> The stage of the game gives the context in general for the priority of the elements to consider.
I actually implemented the feature yesterday :) With the help of Grok who wrote the functions. Now I need to debug and refine them.
To be sure it has not considered the stage of the game, However, I don't think it matters. I mean, if the king safety is low, but you're in a phase of the game where that doesn't matter, you just ignore it. The way I implemented it is you see a "K-MAPS" text and each letter has a color which is greenish or reddish based on values from -1 to 1.
As for what the LLM did in the code, I don't have to go into details, but it seems to have used:
K - pawn shields, open files near king, enemy threats
M - obvious
A - number of possible moves in the position
P - isolated, passed pawns
S - number of controlled squares in the opponent's side (with bonus for central squares)
> You have got me thinking with this question - many thanks :)
...
> The stage of the game gives the context in general for the priority of the elements to consider.
I actually implemented the feature yesterday :) With the help of Grok who wrote the functions. Now I need to debug and refine them.
To be sure it has not considered the stage of the game, However, I don't think it matters. I mean, if the king safety is low, but you're in a phase of the game where that doesn't matter, you just ignore it. The way I implemented it is you see a "K-MAPS" text and each letter has a color which is greenish or reddish based on values from -1 to 1.
As for what the LLM did in the code, I don't have to go into details, but it seems to have used:
K - pawn shields, open files near king, enemy threats
M - obvious
A - number of possible moves in the position
P - isolated, passed pawns
S - number of controlled squares in the opponent's side (with bonus for central squares)