Originally posted by razor2007Sounds like you'd spend ages on doing the maths, and little time to actually play the position!
All my information is from an article published in Chess Life in March 1999 by IM Larry Kaufman. It is titled "The Evaluation of Material Imbalances."
In reference to the a & h pawns are worth 15% less, "There is one case which can be treated as positional or material, namely the rook's pawn, which differs from other pawns in that it can only capture on ...[text shortened]... of the side being valued, with the opposite adjustment for each pawn short of five.
Yeah, that's the problem. The reaon this si interesting is because it gives us a basic idea s to the true values of piecwes in different positions instead of a generic 1-3-3-5-9 chart type system.
Through experience we can, in a broad sense, make use of these principles to the point that they become internal.\
It's second anture to me casue I'm a math major, but I understand yuor point.
Originally posted by LittleBearit is interesting how a wonderful player such as yourself counters a static evaluation with an example of a conditional static evaluation ...
Pieces should not be evaluated by a static scoring; pieces have a dynamic value. Such is chess.
e.g, two pawns in adjacent columns, in the six rank, easily beat a rook.
each added condition (of truth) surely brings us closer.
Originally posted by Testriderhave you ever tried to use a knight to stop "a" and "h" pawns queening?
I dont think a knight is worth 3.75 points when you have a d+e pawn, and only 2.75 points when you have a g&h pawn
or have you used a knight to help them to queen?
have you ever had an opponent score a draw by clogging the centre with pawns and holding it all together with a knight against your rook or bishop?
i think i was on track ...
If Little Bear says that the pieces have a dynamic value it’s the value in a given position. Every move can change the value of a piece. Let’s say – based on that static evaluation with points – a pawn has a value of x---points. Next move you sacrifice this pawn to open the position and to improve the the range of your other pieces . The value of the pawn is then what? He disappeared so he has no value anymore. But before moving that pawn’s dynamic value is very important. Another player in the same position would perhaps never sacrifice this pawn, so the dynamic value of this pawn would be less important for him. But if the value of a given piece depends of the strategic, tactical and positional plans (and abilities) of a player this old-fashioned static evaluation with points is just meaningless and will never help anybody to improve his own chess-skills.
Playing CC and OTB since long but I never ever heard one of my opponents or team-mates talking of points.
This point system is of a large value of statistical interest. The values are averages taken form a database of over 900,000 games.
Yes, in some extremesituations these values are enitrely untrue, but it is still a valuable tool for most situations. I play with points all the time, but that's just me.
Without the point system, how would we know that trading a bishop for a rook was a good trade?
Originally posted by razor2007I've always wondered how these values are extracted from a set of games. any ideas? maybe assign a value of x to one type of piece, say a rook. other types might default to reinfeld-values. then run through all games, adjusting x so that the exchanges will make sense (maybe using least-squares?). then do the same one by one for every type of piece, using the new values that were obtained. then iterate this whole thing until piece-values drift less than some preset 'stableness-factor'.
This point system is of a large value of statistical interest. The values are averages taken form a database of over 900,000 games.
sounds a bit ugly, there probably is a better way? maybe some kind of a matrix with a pseudoinverse to get the least squares solution?
Originally posted by razor2007To answer that question, consider how you decide whether to trade compromising your king safety in order to gain a space advantage. Do you use numerical values to do this? No. But yet you manage to compare and choose between them. Same with material… I judge how valuable the bishop is compared to the rook, based on the position, but without needing to give anything a numerical value.
Without the point system, how would we know that trading a bishop for a rook was a good trade?
Remember, a rook is often worth more than bishop, not because some books say something like 5 and 3 respectively, but for the very reasons where such values came from. e.g. a rook can maximally control 14 squares from any location; a bishop can only control squares of one fixed colour, up to a maximum of 13; etc. Beyond beginner level, it’s necessary to forget the values and think of the reasons. Can my rook display its powers in this position and/or does it have the potential to do so later in the game? Same for the bishop. Plus lots of other reasoning. The pieces don’t have given values; they have to justify their worth in every position.
Further, because a lot of our chess thinking is done subconsciously, there are often times when we “feel” a preference but can’t fully explain it. And our “gut feelings” or intuition doesn't have values attached.
What do you guys think about a well bishop posted in the centre against a bishop cramped in its pawn structure with no future of ever getting out? Would points be deducted (the baaaad one) based on the difficulty in activating it? Just wondering since I want to know how these programmers think...
Edit: Well posted bishop...