A Knight+Bishop is much stronger than a Rook. Officially it's only one point difference (3+3 vs 5) but in fact the advantage is larger than the simple count would suggest.
Knight and a Bishop - for a Rook
Sort:
Mar 12, 2019
0
#3
It can depend on the stage and position in the game, a Rook can generally be better in the end game than a Knight and Bishop as the rook is very active covering a lot of squares, however, in the early and middle game stages a Knight and Bishop are generally better as the rooks are not yet as active. Having that extra active Knight/Bishop in the middle game may lead to better tactics and a better position controlling more central squares.
According to Leela chess Zero game play (where she knows 34 million positions/patterns in her neural network brain), she usually trade off Kt+B vs R+P (and keep rook in endgame).
It looks like the rook in endgame seemed slighter better (obviously two minors are better in opening and middle game though)
Previously human generally agreed as 3.25 for bishops and 3.15 for knight.(5 for rook) So most human and human programmed software (like stockfish) prefer to keep two minors in all phases of the game . (Stockfish evaluation of rook value in endgame become higher but not high enough to exchange with two minors, so stockfish usually keeps two minors in endgame)
I look at it this way, a knight can move in ways a rook cant. And a bishop can move in ways a rook cant. So sometimes either a knight or a bishop are better than a rook. So to get BOTH a knight and a bishop for only a rook seems like almost always a good idea.
Mar 20, 2019
0
#8
In a recent game i managed to exchange my bishop for the opponents rook. I felt i made a good deal, but his minor pieces became very active while both my rooks were passive. That cost me the game.
The value of the pieces are relative and highly depends on the position and phase of the game.
Mar 20, 2019
0
#9
Thanks! I have learned a lot from you guys. I appreciate your advice.
I had a game N & B v. R and but queens also. The other bloke had better pawns 5 to my 3. Open board, no center pawns. I had a chance to force queen exchange and decided not to. I think that was wrong, N+B vs R is probably good but N+B+Q vs R+Q if coordinated, maybe the R is better? Does that make sense?
The Bishop and Kn is your creative pieces, so in the opening and middlegame those will be way more active and the pieces to make the attack and defense strong. In the endgame with few pieces the rook and pawn will be easier to mate with.
According to GM Laury Kaufman ' imbalance.
1.R+B+N much stronger than R+R (redundancy penalty between 2R but synergy bonus between R and minors)
2. B+N slightly stronger than R
@sdrosen, Q+ minor should be like condition 1.
@sdrosen, I have personal experience where I won a game in a particular position is extremely flawed conclusion!
1. I vs someone ( candidate bias) ---> elo bias ( to remove it, test Stockfish vs Stockfish)
2. won a few games( sample bias)----> sample is too small to do statistical conclusion ( do at least 200 games in 1 min bullet SF vs SF , 40,000 games is standard testing though)
3. In a particular game (positional bias)----> to remove it, do in balanced condition like that.
Current world best two engines like Stockfish and Leela both agreed, Q+R is losing.
Mar 26, 2019
0
#14
B and N often need a coordinator, a further rook. In the pure endgame a rook is often as strong as both minor pieces who lack coordination. So if both parties have additional heavy pieces the two minor pieces are preferable.
Here an example what I mean: the two minor pieces struggle without coordination:
http://www.chessgames.com/perl/chessgame?gid=1552975
I think you should exchange ur rook for two Knights . Knights are cool!☯
BL4D3RUNN3R wrote:
B and N often need a coordinator, a further rook. In the pure endgame a rook is often as strong as both minor pieces who lack coordination. So if both parties have additional heavy pieces the two minor pieces are preferable.
Here an example what I mean: the two minor pieces struggle without coordination:
http://www.chessgames.com/perl/chessgame?gid=1552975
I think 34. Ng4??? is the serious blunder to me and also illogical in wasting time with Kt.That probably turned winning into draw or lose.
It is the problem with human games where human plays serious blunders left and right especially in endgame that turned the result of game upside down frequently.
I dont need to check with SF , this position looks a forced, easy win for white by 34. Bf5!!!
For N+B vs R with an extra pawn, it seems pretty much a stalemate
Mar 26, 2019
0
#18
It all depends on the dynamics of the position.
Case 1: A fairly complex middle game with dark squared weaknesses and good outposts for the knight.
A rook without open files would easily be dominated by a dark squared bishop and a knight.
Case 2: An endgame in which white has a distant passed pawn supported by a rook vs Black’s Knight + Bishop and a pawn that is nascent.
The minors would have a hard time stoping the pawn.
Case 1: A fairly complex middle game with dark squared weaknesses and good outposts for the knight.
A rook without open files would easily be dominated by a dark squared bishop and a knight.
Case 2: An endgame in which white has a distant passed pawn supported by a rook vs Black’s Knight + Bishop and a pawn that is nascent.
The minors would have a hard time stoping the pawn.
For confirmation, I checked with SF and I am happy that SF agreed with 34. Bf5! ( +1.4 at depth 62)
For serious analysis, I will need 6 men EGTB and 1 billion position search per move( 3 mins per move for slow pc) or if I run SF for 1 hour non- stop for that position, SF may probably show EGTB win score +123.
ChungYin wrote:
For N+B vs R with an extra pawn, it seems pretty much a stalemate
You dont need to say, " It seemed". It is already 100% proved with brute force search up to 7 men EGTB". It is draw!! To win the game, the stronger side B+N need more pawns!! (eg, 3 pawns in both sides) . Chess is 100% solved in any position up to 7 men (pieces) .
https://syzygy-tables.info/?fen=4k2r/4p3/8/8/8/8/4P3/1NB1K3_w_-_-_0_1
I am feeling uncomfortable, with this type of exchange. A rook for me, is much better to play with, than a knight and a bishop - although there is a one point difference in score. Is it because I am still a low rating player, that make me feel this way? Will the exchange make sense later, as my chess evolve? What do you think?