Piece Material Value Analysis

Archive of the old Parsimony forum. Some messages couldn't be restored. Limitations: Search for authors does not work, Parsimony specific formats do not work, threaded view does not work properly. Posting is disabled.

Piece Material Value Analysis

Postby Han » 13 Jul 2000, 15:32

Geschrieben von: / Posted by: Han at 13 July 2000 16:32:52:
All these are taken from my pages. I put this here to start an arguement!
I will attempt to formulate some equations for getting the piece values. Most of them are based on logical reasoning (which you may disagree completely with), while some others are just my preferences :-)
My system assumes that all the differences between the left and right of the inequalities are very small, and therefore are assumed to be the same, which is denoted by a.
P =Pawn, N = Knight, B = Bishop, R = Rook, Q = Queen, a = unknown constant
Inequality Equation Reason
N > 3P ; N = 3P + a Knights are slightly stronger than 3 pawns.
B > N ; B = N + a Generally, bishops are better than knights due to its range.
R < 2N-P ; R = 2N - P - a Generally, 2 minor pieces exert more pressure than a lone rook can.
Q > 3B ; Q = 3B + a The queen attacks laterally too.
Q > R + B+ P ; Q = R+ B+ P+ a This is controversial. But the queen is able to cover all of the diagonals, which the bishop cannot.
Let's solve the equations. P = 100, a= any reasonable value.
In this table, a is assumed to be 10. The biggest value overrides others to satisfy all inequalities.
Piece Expressions Piece Values , a = 10
Pawn P 100
Knight =3P + a =310
Bishop =3P + 2a =320 320 Overrides smaller value.
=3P = 300
Rook =5P + a = 510 510 Overrides smaller value.
=5P - a = 490
Queen =9P + 7a = 970 970 Overrides smaller value.
=9P + 4a = 940
The main idea in this is that a is relatively small, and are therefore quite equal.
Han
 

It is quite messy, so visit my website for a neater version,

Postby Han » 13 Jul 2000, 15:36

Geschrieben von: / Posted by: Han at 13 July 2000 16:36:13:
Als Antwort auf: / As an answer to: Piece Material Value Analysis geschrieben von: / posted by: Han at 13 July 2000 16:32:52:
All these are taken from my pages. I put this here to start an arguement!
I will attempt to formulate some equations for getting the piece values. Most of them are based on logical reasoning (which you may disagree completely with), while some others are just my preferences :-)
My system assumes that all the differences between the left and right of the inequalities are very small, and therefore are assumed to be the same, which is denoted by a.
P =Pawn, N = Knight, B = Bishop, R = Rook, Q = Queen, a = unknown constant
Inequality Equation Reason
N > 3P ; N = 3P + a Knights are slightly stronger than 3 pawns.
B > N ; B = N + a Generally, bishops are better than knights due to its range.
R < 2N-P ; R = 2N - P - a Generally, 2 minor pieces exert more pressure than a lone rook can.
Q > 3B ; Q = 3B + a The queen attacks laterally too.
Q > R + B+ P ; Q = R+ B+ P+ a This is controversial. But the queen is able to cover all of the diagonals, which the bishop cannot.
Let's solve the equations. P = 100, a= any reasonable value.
In this table, a is assumed to be 10. The biggest value overrides others to satisfy all inequalities.
Piece Expressions Piece Values , a = 10
Pawn P 100
Knight =3P + a =310
Bishop =3P + 2a =320 320 Overrides smaller value.
=3P = 300
Rook =5P + a = 510 510 Overrides smaller value.
=5P - a = 490
Queen =9P + 7a = 970 970 Overrides smaller value.
=9P + 4a = 940
The main idea in this is that a is relatively small, and are therefore quite equal.
Han
 

Re: Piece Material Value Analysis

Postby Dann Corbit » 13 Jul 2000, 19:30

Geschrieben von: / Posted by: Dann Corbit at 13 July 2000 20:30:50:
Als Antwort auf: / As an answer to: Piece Material Value Analysis geschrieben von: / posted by: Han at 13 July 2000 16:32:52:
How much are these two bishops worth?
3k4/5b2/3p4/2pPp3/1pP1Pp1p/pP3PpP/P5P1/1B1K4 w - -
Pieces have a static value but also a dynamic value. A pawn or piece can have it's value multiplied or driven to zero.
How much is this pawn worth?
8/1P6/6k1/8/2K5/8/8/8 w - -
More than 100 centipawns, I would say. In fact, I will trade a rook for it in a heartbeat.


my ftp site
Dann Corbit
 

Re: Piece Material Value Analysis

Postby Han » 14 Jul 2000, 10:31

Geschrieben von: / Posted by: Han at 14 July 2000 11:31:56:
Als Antwort auf: / As an answer to: Re: Piece Material Value Analysis geschrieben von: / posted by: Dann Corbit at 13 July 2000 20:30:50:
How much are these two bishops worth?
3k4/5b2/3p4/2pPp3/1pP1Pp1p/pP3PpP/P5P1/1B1K4 w - -
Pieces have a static value but also a dynamic value. A pawn or piece can have it's value multiplied or driven to zero.
How much is this pawn worth?
8/1P6/6k1/8/2K5/8/8/8 w - -
More than 100 centipawns, I would say. In fact, I will trade a rook for it in a heartbeat.
In my system, it is only the static value that is being calculated. Only in the process of the game, then the dynamic material values come in. And anyway, you have selected a too extreme example to state your point:-)
Han
Han
 

Re: Piece Material Value Analysis

Postby Dan Andersson » 14 Jul 2000, 18:34

Geschrieben von: / Posted by: Dan Andersson at 14 July 2000 19:34:52:
Als Antwort auf: / As an answer to: Re: Piece Material Value Analysis geschrieben von: / posted by: Han at 14 July 2000 11:31:56:
How much are these two bishops worth?
3k4/5b2/3p4/2pPp3/1pP1Pp1p/pP3PpP/P5P1/1B1K4 w - -
Pieces have a static value but also a dynamic value. A pawn or piece can have it's value multiplied or driven to zero.
How much is this pawn worth?
8/1P6/6k1/8/2K5/8/8/8 w - -
More than 100 centipawns, I would say. In fact, I will trade a rook for it in a heartbeat.
In my system, it is only the static value that is being calculated. Only in the process of the game, then the dynamic material values come in. And anyway, you have selected a too extreme example to state your point:-)
Han
There is nothing especially extreme about these positions, about one in ten games I play feature these kinds of abnormal (?) positions in the post game analysis. As to evaluating material, I have found that a bonus based on number of pieces work really well. For example:
if linear
(noOfWhitePieces-noOfBlackPieces)*positiveConstant
or discrete
PieceBonus[noOfWhitePieces-noOfBlackPieces]
Dan Andersson
 

Re: It is quite messy, so visit my website for a neater vers

Postby Tom King » 14 Jul 2000, 22:57

Geschrieben von: / Posted by: Tom King at 14 July 2000 23:57:07:
Als Antwort auf: / As an answer to: It is quite messy, so visit my website for a neater version, computer chess :-) geschrieben von: / posted by: Han at 13 July 2000 16:36:13:
All these are taken from my pages. I put this here to start an arguement!
I will attempt to formulate some equations for getting the piece values. Most of them are based on logical reasoning (which you may disagree completely with), while some others are just my preferences :-)
My system assumes that all the differences between the left and right of the inequalities are very small, and therefore are assumed to be the same, which is denoted by a.
P =Pawn, N = Knight, B = Bishop, R = Rook, Q = Queen, a = unknown constant
Inequality Equation Reason
N > 3P ; N = 3P + a Knights are slightly stronger than 3 pawns.
B > N ; B = N + a Generally, bishops are better than knights due to its range.
R < 2N-P ; R = 2N - P - a Generally, 2 minor pieces exert more pressure than a lone rook can.
Q > 3B ; Q = 3B + a The queen attacks laterally too.
Q > R + B+ P ; Q = R+ B+ P+ a This is controversial. But the queen is able to cover all of the diagonals, which the bishop cannot.
Let's solve the equations. P = 100, a= any reasonable value.
In this table, a is assumed to be 10. The biggest value overrides others to satisfy all inequalities.
Piece Expressions Piece Values , a = 10
Pawn P 100
Knight =3P + a =310
Bishop =3P + 2a =320 320 Overrides smaller value.
=3P = 300
Rook =5P + a = 510 510 Overrides smaller value.
=5P - a = 490
Queen =9P + 7a = 970 970 Overrides smaller value.
=9P + 4a = 940
The main idea in this is that a is relatively small, and are therefore quite equal.
Hey, Han, what's the URL of your website?
Tom King
 

URL of Website :-) {Sorry for not putting a link up} (no tex

Postby Han » 15 Jul 2000, 10:54

Geschrieben von: / Posted by: Han at 15 July 2000 11:54:40:
Als Antwort auf: / As an answer to: Re: It is quite messy, so visit my website for a neater version, computer chess :-) geschrieben von: / posted by: Tom King at 14 July 2000 23:57:07:
Han
 

URL of Website :-) {Sorry for not putting a link up}

Postby Han » 15 Jul 2000, 10:55

Geschrieben von: / Posted by: Han at 15 July 2000 11:55:58:
Als Antwort auf: / As an answer to: Re: It is quite messy, so visit my website for a neater version, computer chess :-) geschrieben von: / posted by: Tom King at 14 July 2000 23:57:07:
Url is www.chess1.homepage.com
I accidentally clicked on the button too many times!
Han
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: No registered users and 29 guests