hgm wrote:Because there is never more than 1 Bishop, the B-pair bonus cannot be involved.
I seriously doubt there exists somethng like a measurable BN bonus. In all games my engines ever played, I have never seen even a single one end in KBNK. Mating potential hardly contributes to piece value, also for individual pieces. E.g. a curtailed Rook limited to a maximum of 2 steps in any direction has almost exactly the same opening value as a similarly limited Bishop, despite the fact that the former can mate, and the latter not, against a bare King.
There are many possibilities. For instance, it is well known that Q+N is slightly better than Q+B everything else being relatively equal (at least for human games )l. How to make this successfully into an engine is another story.
michiguel wrote:For instance, it is well known that Q+N is slightly better than Q+B everything else being relatively equal (at least for human games )l. How to make this successfully into an engine is another story.
What is said by Capablanca in early 1900's is not necessary true. Petrosyan thought that minor piece cost 3, and a Rook costs 4 pawns, his famous exchange sacrifices was not sacrifices for his own.
mjlef wrote:I was wondering what your experiences were with these:
Rook pair penalty (small penalty for having two rooks)
Knight pair penalty (small penalty for having two knights)
Having a pawn bonus (small bonus for having at least one pawn)
I see several program use these and was wondering if anyone had tested how good they might be?
Mark
I have tested a number of these types of evaluation and I never liked the result. Positions quickly degenerate to endgames.
What do you mean? this type of evaluations are not good because positions with imbalances turn rapidly into endgames? I do not think I get what you meant.
Miguel
The same goes for material balance evaluation. They tend to result in quick equal exchanges.
Nope. That is not what I meant.
What I was saying is that, once you start penalizing for the presence of certain pieces, those pieces tend to get traded early. A few more trades here and there, and you are now in the endgame.
Rook pair penalty does just one thing: it encourages exchange of a pair of rooks when the other side has one rook less (because of an exchange sacrifice). I don't think that it amounts to going down the fast lane to the endgame, as long as opponent tries to keep pieces on the board. and good understading of dynamic compensation ought to lead him to do so even without a rook pair penalty.
michiguel wrote:For instance, it is well known that Q+N is slightly better than Q+B everything else being relatively equal (at least for human games )l. How to make this successfully into an engine is another story.
What is said by Capablanca in early 1900's is not necessary true. Petrosyan thought that minor piece cost 3, and a Rook costs 4 pawns, his famous exchange sacrifices was not sacrifices for his own.
I guess I do not understand what you mean. Petrosian had a particular disregard for rooks but how that relates to Q+B vs Q+N?
mjlef wrote:I was wondering what your experiences were with these:
Rook pair penalty (small penalty for having two rooks)
Knight pair penalty (small penalty for having two knights)
Having a pawn bonus (small bonus for having at least one pawn)
I see several program use these and was wondering if anyone had tested how good they might be?
Mark
I have tested a number of these types of evaluation and I never liked the result. Positions quickly degenerate to endgames.
What do you mean? this type of evaluations are not good because positions with imbalances turn rapidly into endgames? I do not think I get what you meant.
Miguel
The same goes for material balance evaluation. They tend to result in quick equal exchanges.
Nope. That is not what I meant.
What I was saying is that, once you start penalizing for the presence of certain pieces, those pieces tend to get traded early. A few more trades here and there, and you are now in the endgame.
As Pawel said, only if you have imbalances (R+R vs R+something else). If you do not have imbalances, both sides get the penalty (or bonus) and a trade is not encouraged. Going to an endgame is not bad if the imbalance dictates it is the way to go.
I think this may not work because the penalties should be so small than the global effect is masked by other more important things. Bishop pair is an exception because the bonus should be really big and noticeable.
I'd think it'd also make a "rook sac" more likely when one has both rooks, regardless of whether the other side does.
[d] 3b4/8/8/3p4/R3b3/8/8/4R3 w - - 0 1
Rxe4 dxe4 Rxe4
If there's a rook pair penalty, the position is evaluated that much better after the exchange. If the position is near alpha, that could change the PV. Unlikely, sure. But possible
michiguel wrote:For instance, it is well known that Q+N is slightly better than Q+B everything else being relatively equal
As far as I remember this particular imbalance was mentioned by Capablanca. I doubt that anybody famous earlier or later tested or even agreed or disagreed with his opinion.
michiguel wrote:For instance, it is well known that Q+N is slightly better than Q+B everything else being relatively equal
As far as I remember this particular imbalance was mentioned by Capablanca. I doubt that anybody famous earlier or later tested or even agreed or disagreed with his opinion.
Is Capablanca's theorem that a queen and knight are better than queen and bishop true?John Watson cast some doubt on this in his excellent book Secrets of Modern Chess Strategy and came to the following conclusions:
"a) an unusually large proportion of q+N v Q+B games are drawn;
b) most games which are won by either side ... are characterized by that side having one or more rather obvious other advantages."
However,in closed positions with the bishop hemmed in by its own pawns, the knight certainly is superior
I guess noone noticed anything special in the egtb tables with respect to the qb-qn endgames.
)l. How to make this successfully into an engine is another story.