Page Index Toggle Pages: [1] 2 
Topic Tools
Hot Topic (More than 10 Replies) Wikipedia article about Chess Piece Relative Value (Read 16645 times)
Stigma
God Member
*****
Offline


There is a crack in everything.

Posts: 3277
Joined: 11/07/06
Gender: Male
Re: Wikipedia article about Chess Piece Relative Value
Reply #24 - 09/03/14 at 22:14:57
Post Tools
FEN is sort of the equivalent of PGN, but for positions instead of game scores - a text format designed to be readable by both programs and humans, and open to all programs.

http://en.wikipedia.org/wiki/Forsyth-Edwards_Notation

Most chess programs have a function for creating FENs. For example, the "Copy position" function in Chessbase will produce a FEN string of the current position, which you can then paste into a text file, web form, other program etc. etc.

Also if, you look at a pgn file with "games" that don't start from the starting position (exercises, game excerpts, endgame positions etc.), you will notice that the PGN format uses FEN to specify the position.
  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
motörhead
Full Member
***
Offline


Here comes the bass, thunder
in the guts...

Posts: 226
Joined: 10/09/08
Re: Wikipedia article about Chess Piece Relative Value
Reply #23 - 09/03/14 at 18:22:48
Post Tools
GMTonyKosten wrote on 09/03/14 at 09:19:42:
motörhead wrote on 09/02/14 at 22:08:02:
[White Kf3, Re1, Nd3, Ne5, Ba2, b2, c2, f4, g3, h2
Black Kg8, Rd8, Nc6, Ng6, Ba7, b6, c7, d7, g7, h7 
(I'm not able to create those pretty diagrams, soory)


Just click on the little board icon, and then paste the fen code, thus:

* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
*

when it appears that Black's king is in check... Roll Eyes


As Stigma gave it: I went wrong categorizing the pawns with B for Bauer (= pawn) 
Sorry about that. Stigma's diagram gives the situation right.

Of the topic:
What is this fen code? How do I get it from a normal list of piece placements?
I know, I show up as being old school...
  

A walk trough the ocean of most souls would scarcely get your feet wet.
Back to top
 
IP Logged
 
Stigma
God Member
*****
Offline


There is a crack in everything.

Posts: 3277
Joined: 11/07/06
Gender: Male
Re: Wikipedia article about Chess Piece Relative Value
Reply #22 - 09/03/14 at 15:45:22
Post Tools
GMTonyKosten wrote on 09/03/14 at 09:19:42:

Just click on the little board icon, and then paste the fen code, thus:

[...]

when it appears that Black's king is in check... Roll Eyes


Let me make a wild guess that the bits on a2 and a7 were meant to be pawns (B for Bauer in German or bonde in Scandinavian languages) rather than bishops. It's easy to forget something when "translating" piece letters.

That would give us this position instead:

* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
*
  

Improvement begins at the edge of your comfort zone. -Jonathan Rowson
Back to top
 
IP Logged
 
Bonsai
God Member
*****
Offline



Posts: 622
Joined: 03/13/04
Gender: Male
Re: Wikipedia article about Chess Piece Relative Value
Reply #21 - 09/03/14 at 10:40:00
Post Tools
Marc Benford wrote on 09/01/14 at 11:30:45:
After thinking about it for a while I came up with these penalties and bonus for pawns:

An isolated pawn gets a penalty of -0.25
A passed pawn gets a bonus of +0.5
A doubled exchangeable pawn gets a penalty of -0.25
A doubled inexchangeable pawn gets a penalty of -0.5

And concerning doubled pawns: when you have two pawns on the same file, only one of them is considered doubled and gets a penalty, the other one is not considered as a doubled pawn and doesn't receive a penalty for this.

Using this system we find all the exact same values as in the Wikipedia article (except for the h3 pawn which the Wikipedia article say is worth 0.33 while with my system we get a value of 0.25 which is pretty close anyway).


Way too simplistic, it all depends on a lot of factors including dynamic ones.

If we e.g. think of , here white's doubled pawn is clearly strength, because it supports the center and can hold up black's counter-play while white attacks on the kingside. I am not saying the variation is necessarily great for white, just that this is surely better than 3...dxe4 4.f3, which is just not good. Part of the reason for that is that white has got a nice stable center due to the doubled-pawns so that c7-c5 does not really threaten anything immediately.

I guess a lot of Nimzo-Indian Sämisch variations are like that, too, if white can use that stability of his center to build an attack black can be in huge trouble (and the passed pawn if anything is an asset), but if he does not build up any dynamism the doubled-pawns will turn out to be a huge liability (much worse than any -0.25 evaluation).
  
Back to top
 
IP Logged
 
GMTonyKosten
YaBB Administrator
*****
Offline


Mr Dynamic?

Posts: 3192
Location: Clermont-Ferrand
Joined: 12/19/02
Gender: Male
Re: Wikipedia article about Chess Piece Relative Value
Reply #20 - 09/03/14 at 09:19:42
Post Tools
motörhead wrote on 09/02/14 at 22:08:02:
[White Kf3, Re1, Nd3, Ne5, Ba2, b2, c2, f4, g3, h2
Black Kg8, Rd8, Nc6, Ng6, Ba7, b6, c7, d7, g7, h7 
(I'm not able to create those pretty diagrams, soory)


Just click on the little board icon, and then paste the fen code, thus:

* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
* * * * * * * *
*

when it appears that Black's king is in check... Roll Eyes
  
Back to top
IP Logged
 
motörhead
Full Member
***
Offline


Here comes the bass, thunder
in the guts...

Posts: 226
Joined: 10/09/08
Re: Wikipedia article about Chess Piece Relative Value
Reply #19 - 09/02/14 at 22:08:02
Post Tools
Marc Benford wrote on 09/01/14 at 11:30:45:
After thinking about it for a while I came up with these penalties and bonus for pawns:

An isolated pawn gets a penalty of -0.25
A passed pawn gets a bonus of +0.5
A doubled exchangeable pawn gets a penalty of -0.25
A doubled inexchangeable pawn gets a penalty of -0.5

And concerning doubled pawns: when you have two pawns on the same file, only one of them is considered doubled and gets a penalty, the other one is not considered as a doubled pawn and doesn't receive a penalty for this.

Using this system we find all the exact same values as in the Wikipedia article (except for the h3 pawn which the Wikipedia article say is worth 0.33 while with my system we get a value of 0.25 which is pretty close anyway).


Sorry, but to me this seems to be to artificial. I can't use this as i don't think this way.
As Smyslov-Fan asks: What are you heading for?

May I cite John Nunn ("John Nunn's Chess Course", p. 116)? "Doubled pawns are in general a slight weakness, and players will normally expect some sort of compensation before accepting them [...] There are also cases in which doubled pawns are almost irrelevant."
Sounds clear but isn't. Take the opening-variation 1.e4 d6 2.d4 Nf6 3.Nc3 e5 4.dxe5 dxe5 5.Qxd8+ Kxd8 6.Bc4 Be6 7.Bxe6 fxe6: Black's doubled pawns in the e-file are by far not pretty, but they cover important central squares. This variation isn't clear after all. And that shows that doubled pawns can't be judged in Quarter-pawns manner. It's just to difficult and depends on the very position.

Axel Smith ("Pump up your rating", p. 25) gave a nice example:
White Kf3, Re1, Nd3, Ne5, Ba2, b2, c2, f4, g3, h2
Black Kg8, Rd8, Nc6, Ng6, Ba7, b6, c7, d7, g7, h7
(I'm not able to create those pretty diagrams, soory)
Here it's worse to double the pawns on g6, despite this looks most ugly (as the g6/g7-pawns hold up White's 3:2 at least as good as the g7/h7-pawns). Better is 1.Nxc6! doubling the pawns in a seemingly more compact way, but keeping the 3:2 majority on the King's side in a more flexible shape.
1... dxc6 2.f5 Nf8 3.Re7 Rd7 4.Re8 with a huge advantage for White acc to Smith. 

After all: It all depends...
  

A walk trough the ocean of most souls would scarcely get your feet wet.
Back to top
 
IP Logged
 
Smyslov_Fan
God Member
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: Wikipedia article about Chess Piece Relative Value
Reply #18 - 09/01/14 at 15:43:44
Post Tools
Just one question: Why is it good to mimic Wikipedia when it comes to chess evaluations? They are hardly the definitive source on chess matters. 

As others have pointed out, the computer programmers have created a very refined system for giving numeric evaluations to the various pawns and structural weaknesses. Unless you're interested in programming, humans don't need this sort of rubric. Humans can tell the relative worth of pawns in a specific position without resorting to counting tenths or hundredths of pawns.
  
Back to top
 
IP Logged
 
Marc Benford
Full Member
***
Offline


I Love ChessPublishing!

Posts: 104
Joined: 07/17/13
Re: Wikipedia article about Chess Piece Relative Value
Reply #17 - 09/01/14 at 11:30:45
Post Tools
After thinking about it for a while I came up with these penalties and bonus for pawns:

An isolated pawn gets a penalty of -0.25
A passed pawn gets a bonus of +0.5
A doubled exchangeable pawn gets a penalty of -0.25
A doubled inexchangeable pawn gets a penalty of -0.5

And concerning doubled pawns: when you have two pawns on the same file, only one of them is considered doubled and gets a penalty, the other one is not considered as a doubled pawn and doesn't receive a penalty for this.

Using this system we find all the exact same values as in the Wikipedia article (except for the h3 pawn which the Wikipedia article say is worth 0.33 while with my system we get a value of 0.25 which is pretty close anyway).
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1240
Joined: 05/17/10
Gender: Male
Re: Wikipedia article about Chess Piece Relative Value
Reply #16 - 08/30/14 at 10:58:20
Post Tools
Agreed. What you really need to know is not how much doubled pawns are worth, but what to do in various configurations once you or your opponent has them, and therefore how to recognize where they are good and where not.  That is knowledge for a human being.

I can also recommend Euwe and Kramer, The Middlegame: Book One and Silman, Reassess Your Chess. Botvinnik's work, for example Botvinnik's Best Games, Volume 1, contains a virtual seminar on how to play when you have weak pawns, since he often voluntarily assumed them as a way to beat his opponents. He also knew how to evaluate a position better than anyone else--and how to communicate its essence to human beings, namely in language.

  
Back to top
 
IP Logged
 
motörhead
Full Member
***
Offline


Here comes the bass, thunder
in the guts...

Posts: 226
Joined: 10/09/08
Re: Wikipedia article about Chess Piece Relative Value
Reply #15 - 08/26/14 at 22:57:41
Post Tools
Marc Benford wrote on 08/25/14 at 14:52:31:
What I am asking for is somekind of average values of the pawns depending on if they are connected, isolated, doubled or passed.



What I am trying to figure out is, on average:
- What penalty should an isolated pawn receive?
- What penalty should a doubled pawn receive?
- What bonus should a passed pawn receive?


The newest source of quality and in depth analysis dealing with doubled pawns is "John Nunn's Chess Course" which is "based on the games of World Champion Lasker". 
Not only the very (but short) chapter there is relevant but too all the games given where Lasker played the Ruy Lopez exchange. 
It offers a lot, at least to my feeling after browsing through the book.
  

A walk trough the ocean of most souls would scarcely get your feet wet.
Back to top
 
IP Logged
 
motörhead
Full Member
***
Offline


Here comes the bass, thunder
in the guts...

Posts: 226
Joined: 10/09/08
Re: Wikipedia article about Chess Piece Relative Value
Reply #14 - 08/26/14 at 22:44:19
Post Tools
Uhohspaghettio wrote on 08/26/14 at 00:00:01:
motörhead wrote on 08/25/14 at 19:18:00:
As it was the case here. It seemed to me bing based on reasonable ground. Berliner is a reasonable source, at least as reasonable as Chernev or Reinfeld...


lol, I think you are guessing. Berliner certainly is not as good a source as them. Berliner was an IM and world champion at correspondence chess and an accomplished computer programmer. He was not known as an influential theoretician.

He tried translating his findings in engine programming to chess. It never took off or found favour, nobody of note advocates it afaik, though to give him due respect it was probably worth a shot. 

Furthermore that's on wikipedia as a brief outline of his ideas and theories. If you really want to follow up on it then why don't you actually buy his book? In his book I'm sure it's explained properly, he did not put it public domain.   
   


Thx for the insightful post. You think, I’m guessing. Ah… And as you think so, it seems your thoughts turn to become the truth in your head, isn’t it?!

Sorry, but you’re wrong. Just look at the topic given here and then look what you’re dealing with.

The topic was & is doubled pawn structures, basing on a Wikipedia text and a book by Berliner. And you? You’re discussing about the author Berliner, his career, the content of his book up to the opening topics in it (as kylemeister did).

That means you fail by far  to meet the topic. That’s boring for those interested in the very topic.

For your thought about me guessing: I own a copy of that Berliner book. I know of the discussions on the book all to well. But at least his book was published by Gambit, not known to publish every rubbish dealt to them.

And to me not all the content is flawed. As not every remark by Chernev and Reinfeld is superficial (I own books of them too). To me it seems that Berliner was surely the better and more insightful chess player than the both. But besides any comparisons: If you look at the writings with an open mind, you can even profit from a second or third or even lesser rated resource. Just don’t believe them slavishly. Think for yourself.

And to my feeling that’s just what Marc tried.

But now look what happened to his questions. 8 of the 11 posts do not touch the topic the slightest but deal with behavior of the participants.

But as ReneDescartes put it:

ReneDescartes wrote on 08/25/14 at 20:15:04:
This philosophical aspect of the question is in fact the only interesting thing about it, and is worth explaining.


And I just ask to respect this. We all would have more from it without a backseat driver behavior. It just keeps away newbies from the forum when they are treated as offensive as here (“Can you explain WHY? Your questions are all just bizarre.”…).
  

A walk trough the ocean of most souls would scarcely get your feet wet.
Back to top
 
IP Logged
 
kylemeister
God Member
*****
Offline



Posts: 4989
Location: USA
Joined: 10/24/05
Re: Wikipedia article about Chess Piece Relative Value
Reply #13 - 08/26/14 at 03:38:24
Post Tools
Plus, by getting Berliner's book you will learn that the Modern Benoni loses to the Penrose (Bd3 and Nge2), the Gruenfeld is devastated by that Rc1 and Kf1 line of the Classical Exchange, the Slav is rolled up by f3 in the Exchange, the Saemisch Panno is pushed off the board by an early Rb1 and much much more!

John Watson's review in (I think) Inside Chess was pretty satisfying.
  
Back to top
 
IP Logged
 
Uhohspaghettio
God Member
*****
Offline


I Love ChessPublishing!

Posts: 515
Joined: 02/23/11
Re: Wikipedia article about Chess Piece Relative Value
Reply #12 - 08/26/14 at 00:00:01
Post Tools
motörhead wrote on 08/25/14 at 19:18:00:
As it was the case here. It seemed to me bing based on reasonable ground. Berliner is a reasonable source, at least as reasonable as Chernev or Reinfeld...


lol, I think you are guessing. Berliner certainly is not as good a source as them. Berliner was an IM and world champion at correspondence chess and an accomplished computer programmer. He was not known as an influential theoretician.

He tried translating his findings in engine programming to chess. It never took off or found favour, nobody of note advocates it afaik, though to give him due respect it was probably worth a shot. 

Furthermore that's on wikipedia as a brief outline of his ideas and theories. If you really want to follow up on it then why don't you actually buy his book? In his book I'm sure it's explained properly, he did not put it public domain.   
  
Back to top
 
IP Logged
 
ReneDescartes
God Member
*****
Offline


Qu'est-ce donc que je
suis? Une chose qui pense.

Posts: 1240
Joined: 05/17/10
Gender: Male
Re: Wikipedia article about Chess Piece Relative Value
Reply #11 - 08/25/14 at 20:15:04
Post Tools
It is only computer programmers who employ and can benefit from such numeric measures. However, programmers, depending on the program, assess not only the value of pieces, but also that of every open line on the board, every controlled file, every attacked square, every weakness in the king position, and so on, numerically. Numerico-logical methods are the only way that computers can make a decision, or do anything at all. Even those ultra-precise values can vary, in order to vary te program's style.

Numeric methods are ineffective for humans. Education for beginners with "Reinfeld" piece values hide this fact and make you think that if only a more refined Reinfeld method could be designed, it could estimate ("on average") whether it is worth it to double a pawn. Yet, in reality, a human thinks "what are the most striking features of this position," "how can I overcome this obstacle," "does this method of attack usually succeed in positions that look like this," and "what is his plan." Counting in chess among humans is only for babies.

You would be wise not to issue orders to other members of the forum. This philosophical aspect of the question is in fact the only interesting thing about it, and is worth explaining.
  
Back to top
 
IP Logged
 
motörhead
Full Member
***
Offline


Here comes the bass, thunder
in the guts...

Posts: 226
Joined: 10/09/08
Re: Wikipedia article about Chess Piece Relative Value
Reply #10 - 08/25/14 at 19:34:23
Post Tools
Marc Benford wrote on 08/25/14 at 14:52:31:
Thank you motörhead for your useful answer.

Firstly, when you said "one of the pawns is to be evaluated as "normal" and thus should be counted with the value 1. And only the second pawn on that certain file, the doubled pawn, should be counted with the lesser value", I agree with you, but the question is: which one of the two pawns is the "normal" pawn which should be counted with the value 1? and which one of the two pawns is the "sick" pawn which should be counted with a value of less than 1?

Secondly you said that the first f pawn and the first h pawn are each worth 1, which I believe is not correct. I think a pawn should be worth 1 only when it's connected. If a pawn is isolated, it should be worth slightly less than 1.
The first f pawn and the first h pawn are isolated. The second f pawn and the second h pawn are isolated and doubled.

Thirdly I don't think the value of the e pawn should be 1. It should receive a penalty for being isolated, and a bonus for being passed. And the bonus should probably be bigger than the penalty, so the e pawn should be worth slightly more than 1.



For the others who believe my questions are perfectly useless: if you don't like this thread you can just ignore it.
(I am not that stupid, I know that the value of every pawns and pieces are entirely dependant on the position.)
What I am asking for is somekind of average values of the pawns depending on if they are connected, isolated, doubled or passed.



What I am trying to figure out is, on average:
- What penalty should an isolated pawn receive?
- What penalty should a doubled pawn receive?
- What bonus should a passed pawn receive?


I think an average bonus or malus isn't possible. It all simply depends to much on the other factors influencing the position. So you can only judge it from position to position.
The rank reached, vulnerabilty, promotion, defensive tasks, connection, mobility of own/contrahent's pieces etc. etc.
  

A walk trough the ocean of most souls would scarcely get your feet wet.
Back to top
 
IP Logged
 
Page Index Toggle Pages: [1] 2 
Topic Tools
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo