...pawn per move

I read the game evaluations of players/users and evidence is given in (x)pawn per move.How does this work?
Thanks.

Originally posted by jb70
I read the game evaluations of players/users and evidence is given in (x)pawn per move.How does this work?
Thanks.

What? Read where? Are you talking about engines analyzing games and evaluating position in units of pawns? This is done by most all engines if you have one.

Please explain what you're looking for in more detail.

P-

Originally posted by Phlabibit
What? Read where? Are you talking about engines analyzing games and evaluating position in units of pawns? This is done by most all engines if you have one.

Please explain what you're looking for in more detail.

P-

Here are basic worths of pieces by Pawns

Pawn = 1
Knight= 3 or 2.5 in some's view
Bishop = 3
Rook = 5
Queen = 9
King is invaluable

Thus, if you trade your Bishop for a Rook, you are up 2 pawns.

P-

Originally posted by Phlabibit
Here are basic worths of pieces by Pawns

Pawn = 1
Knight= 3 or 2.5 in some's view
Bishop = 3
Rook = 5
Queen = 9
King is invaluable

Thus, if you trade your Bishop for a Rook, you are up 2 pawns.

P-

Sorry.I mean the position not just material.

Originally posted by jb70
Sorry.I mean the position not just material.

That I don't know how to count outside of seeing how many squares Black covers on rows a,b,c, and d vs. how many squares White covers on rows e,f,g, and h.

Perhaps someone will tell us both what you're looking for.

P-

Originally posted by jb70
I read the game evaluations of players/users and evidence is given in (x)pawn per move.How does this work?
Thanks.

Well, if for example Fritz says that the position is +1.00 then white is up the positional equivalent of a pawn. 3.00 would be a minor piece, 5.00 a rook. These are NOT exact evaluations but rather approximations based on what the computer thinks the chances for each side are. 1.00 (or -1.00 for black) is usually the value at which one side is considered winning- below that and a player may have a better position but they are not yet winning.

Originally posted by randolph
Well, if for example Fritz says that the position is +1.00 then white is up the positional equivalent of a pawn. 3.00 would be a minor piece, 5.00 a rook. These are NOT exact evaluations but rather approximations based on what the computer thinks the chances for each side are. 1.00 (or -1.00 for black) is usually the value at which one side is considered winning- below that and a player may have a better position but they are not yet winning.

Any idea HOW it's evaluated? Just based on Best Move Lines? There must be a formula, but I suppose it's probably beyond a human being able to figure out.

P-

I wrote a chess program back in 1986. I no longer have the source code, but as far as I can remember it tried to score each position using the following factors...

Centre control - for attacking squares near the centre
King safety - for keeping pawns and pieces near the king
King attack - for attacking squares near the enemy king
Material
Mobility - I think I scored this at 1/32 of a pawn per possible move.
Pawn structure - passed, isolated, doubled, backward
Piece attack - for a direct threat to win material

...plus some endgame-only factors so that it could manage the simplest rook and queen mates. It didn't have the speed or memory to look more than two or three half-moves ahead.

Modern chess engines have much larger openings databases than mine did, and also have endgame databases. Are there any other completely new factors? Or is it just down to processor speed?