Chess Piece Valuation

Originally posted by Fat Lady
I think the most useful thing about using the usual 1,3,3,5,9 point system is that it gives the player a pretty good indication about what is going to happen in the endgame if the game gets that far.

You give a dangerous example Fat Lady.
In endgames , an active rook might worth 7 and a passive one might worth 3, an active knight might worth as much as a rook and a passive bishop might worth less than a pawn.
Doing a wrong exchange in the middlegame with pieces that have the same numerical value can easily lead to a lost endgame.

I still think that the numerical values of the pieces are only an indication of their initial strength and nothing more.No good player actually ever uses them.No one counts points when he is about to exchange or sacrifice pieces.

Originally posted by e4chris
I've been looking round on the net for an explanation of the chess piece values, how are 1,3,3,5,9 derived? and i can't find one.

These values aren't rigorously derived - there's no point as they position dominates over material values of pieces. The pawn is chosen as the basic unit. I think that the other values of the pieces come from how many pawns you need to force one through in a pawn vs. lone piece ending. You can give checkmate with 2 minor pieces against a king from an arbitrary position only if one of the pieces is a bishop - so they are preferred to knights and are sometimes valued as 3 1/2. 5 probably overvalues the rook so it's sometimes quoted as 4 1/2, since experience shows that two minor pieces are normally better than a rook and a pawn. It's more meant to be a rough guide for beginners to get a handle on what things are worth so they don't swap their queen for a knight thinking they're equally powerful, rather than a rigorous mathematical method of evaluating the strength of a position.

I think can see where Fat Lady is coming from.
In an endgame a lone Rook will usually win v a lone Knight or Bishop.
RHP DB stats bear this out.

Games lone Rook v Knight or Bishop = 33,028

Rook won = 24,779
Drawn = 3,432
Rook lost = 4,811

These losses and draws are the exceptions that ALL players must be aware of.
It is never an automatic win. (no game is).
Some are quite humerous (and scary).

scottcrockart - terrapin333 RHP 2007



Rook v Knght + Rook v Bishop. The exceptional exception.

Everyone knows. (well everyone should know) that two pawns on the 6th
usually beat a Rook. Here.


White to play cannot stop the pawns. One must promote.

Swap that Rook for a Knight or a Bishop and the game is drawn.

Agree that the values seem to be based on 3 pawns for a piece and it makes sense in most cases, they don't seem to be mathematically derived. But Staunton gives them to a few decimal places, i wonder how he got them. I had CM8000 a while back and you can enter values you wanted it to use for each piece eg make its own queen worth 10 so it wouldn't swap it, and a saitek which seems to score the position based material just 1,3 etc, plus or minus a pawn for position. doesn't seem like either have a more advanced way of calculating it

agree with greenpawn that the charts basically show the need for control of the center... good examples...

edit, though if i was going to write a computer program i might try and make it value the knights 1/2 a point more. i think there value is diminished a little because its a headache to think 3 moves ahead with the knight but easy with the bishop.

A chess piece is worth what you can get out of it. A pawn that gives a mate is worth more than 4 queens doing nothing.

you still need to calculate the exchange...

I think the piece valuation is a good guide for general exchanges, but I do not usually trade a bishop for a knight unless I get positional compenstaion. Trading a like pieces to get rid of an opponents piece that is centrally placed or is in an attacking position against one's king is usually beneficial. However, it mainly depends on the present position and the resulting position after the exchanges that determines if I exchange or not.

Originally posted by RJHinds
I think the piece valuation is a good guide for general exchanges, but I do not usually trade a bishop for a knight unless I get positional compenstaion. Trading a like pieces to get rid of an opponents piece that is centrally placed or is in an attacking position against one's king is usually beneficial. However, it mainly depends on the present position and the resulting position after the exchanges that determines if I exchange or not.

It's more than a question of whether or not in general to exchange a bishop for a knight --

If my opponent tends to rely on his knights, favoring his knights over bishops in the middlegame, not only would I swap a bishop for a knight, but certainly if it's a closed position, I would not hesitate to sacrifice a rook for a knight.

Originally posted by YourWorstKnightmare
It's more than a question of whether or not in general to exchange a bishop for a knight --

If my opponent tends to rely on his knights, favoring his knights over bishops in the middlegame, not only would I swap a bishop for a knight, but certainly if it's a closed position, I would not hesitate to sacrifice a rook for a knight.

Well, I would hesitate to do that as an even trade. I would have to be really convinced that the position called for it.

The post that was quoted here has been removed

Is making a disparaging remark all you have to contribute to the topic of this thread? Don't you have at least one thing you could say about the valuation of chess pieces?

it has been said that another way to evaluate pieces as forces (and to see what they require spatially) is to evaluate pieces as analogous combat weapons,
where a knight is an attack helicopter, a bishop a cruise missile, a rook is a tank, a queen a battleship.

I've done this table to show the 'board value' If you put every piece black and white on every square, then give that square the values of the total number of moves for all the pieces ... if that makes sense.... If you add up all the boards i first posted.. This is what you get, normalised to 1

1.00 1.06 1.09 1.09 1.10 1.09 1.06 1.00
0.93 1.06 1.12 1.12 1.12 1.12 1.06 0.93
0.95 1.11 1.22 1.22 1.22 1.22 1.11 0.95
0.95 1.11 1.22 1.27 1.27 1.22 1.11 0.95
0.95 1.11 1.22 1.27 1.27 1.22 1.11 0.95
0.95 1.11 1.22 1.22 1.22 1.22 1.11 0.95
0.93 1.06 1.12 1.12 1.12 1.12 1.06 0.93
1.00 1.06 1.09 1.09 1.10 1.09 1.06 1.00


I like this chart i think it shows where to put your pieces. If you look at a rook it has the same value on every square of the board, but in practice they are most useful pointing at the centre.

Originally posted by e4chris
I've done this table to show the 'board value' If you put every piece black and white on every square, then give that square the values of the total number of moves for all the pieces ... if that makes sense.... If you add up all the boards i first posted.. This is what you get, normalised to 1

1.00 1.06 1.09 1.09 1.10 1.09 1.06 1.00
0.93 1.06 1.12 1.12 ...[text shortened]... ue on every square of the board, but in practice they are most useful pointing at the centre.

The values of the squares and pieces seem to vary depending on the location of the King and other pieces.