chess game online

Chess Game Online

Share and you will discover!
Nutty name, real passions
[ Sign up | Log in | Guest ] (beta)
grunfeld 56 ( +1 | -1 )
Knight vs. Bishop Forgive me if this is an old topic.

I know the conventional wisdom is that a bishop is slightly more valuable than a knight but it seems to me in the end game a knight is more powerul. It may not have the reach of a bishop but a bishop is relegated to either black or white squares where a knight is not.

It seems to me that would make it more valuable. A King could sit on the color opposite of the bishop and be untouchable by a bishop but that's not true of the knight.

Hope I won't be slammed by those saying "duh, how long have you been playing this game anyway".
ironbutterfly 19 ( +1 | -1 )
K v. B Doesn't it depend significantly on how many pawns are on the board? On a crowded board the N can be far superior (if there are pawns on the B's color), but on a relatively empty board the B has much more sweep, don't you think?
More: Chess
grunfeld 28 ( +1 | -1 )
K v. B Thanks for the reply Iron. You could be right. But my argument hinges on the fact that you can evade a Bishop forever by sitting on the opposite color but you can't hide that way from a Knight.

I guess it would depend a lot on what other pieces you have to corner the King.
ionadowman 144 ( +1 | -1 )
grunfeld... This is an old topic... but always worth revisiting. Personally, I go along with conventional wisdom, but with this caveat: the bishop - even the bishop pair can be overrated. Consider this position, with White to play:


As White what would you play here? Would you keep the bishop pair? Or would you try this: 1.Bxf6 gxf6 2.Bxc6 bxc6 3.Ne4 (to protect the c2-pawn)? Black has now the bishop vs knight, but look at those pawns! When something like this turned up in a tournament game several decades ago, I didn't hesitate, and secured a pretty easy win. I've had several games on GK in which I've had N vs B or even 2N vs 2B and the knights have done OK. Ironbutterfly is right: to decide whether to exchange B for N or not does depend on other considerations, especially pawn structure, but the overall dynamics of the position ought to be taken into account. In general, however, the bishop is better on an open board (possibly enough to offset one doubled pawn pair, but not two!); the knight better when the bishop's mobility is restricted (especially true in "closed" positions).

Just one point in addition. In an ending in Q+B+pawns vs Q+N+pawns (pawns equal), it is (generally) a little better to have the knight, even on an open board. The Q and N complement each other rather better than Q and B, the N covering squares the Q cannot reach.
ccmcacollister 496 ( +1 | -1 )
grunfeld ... As you say ... and that is why in answer to the Question of which is better, the pawn structure is by far the greatest single determining factor. In a purely N vs B ending + pawns.
[The matter is extremely position dependent, so let us bar from discussion any position that is just a forced win from the start, or has a great number of other pieces. In general, the more Other pieces there are upon the board, the greater that factor becomes, even in comparison to pawn structure.]
King position is of course also very important in the pure B vs N ending.
Some factors that will favor a Bishop:
*Having opponent pawns upon its color which are locked there aor indefensible, or simply tie the K aor N to their defense.
*Pawns on both wings.
*No good outposts for the N on the color opposite the B's.
*Having a passed Rook pawn(s) vs the N. It is well known endgame theory that Rook pawns (a and h pawns) are the ones generally hardest for a lone N to stop from progressing, or will take it furthest from action elsewhere if it does get a blockade in time.
*A bishop can temporize against the N side, but the N cannot drop tempos unless involving a pawn from the second rank; or a restricted position of the opp K ... wherein N's K can triangulate, but B's King cannot triangulate due to some restriction against using some adjacent SQ due to the pawn structure or N SQ coverage making it impossible for the triangulation if it would place in check or is occupied SQ.
*Finally, Bobby Fischer usually has the B side, and you would have the N~! haha :)
But seriously, to see the techniques for the B-side, see Fischer endings of this type, as he might be considered the Professor of Bishop V Knight wins.
And some to favor the N:
*playing vs multiple broken pawn islands, being able to jump to attack one to the others
*Good N outposts, especially opp the B's color, or from which it attacks a pawn, or especially attacks multiple pawns from there.
*The B side having unpassed pawn, or pawn chain with its base, upon the "Rook 3" square of the Bishop's same color.
*Doubled pawns to play against, especially if isolani(s). {tho this might also help a Bishop to win a rearward doubled-isolani-pawn fixed upon its color, for eg.} And N's love positions with a lot of holes aor SQ's in front of isolani or backward pawns ... whereas the B might also use such SQ's but can eventually be driven from them by the N, or traded for there.
*A SQ(s) where K & N can blockaid an enemy passer with impunity since it is opposite the B's color. (much as in an Opp Color B ending).
*This is definately an imo to me, but it occurs to me that the Question of "Who would win the K&P ending, if the Pieces were removed from the board" is going to be quite significant, since the side having it can place their piece as desired but the other side
will be hindered from using any Squares that could trade to that ending, into a loss. So ... I would guess here that in general, having a won K&P ending situation involved,
may tend to favor the N side significantly more than the B-side .... as I think the N would have more chances to chase or trade off the B than the B cornering the N to do so.
If anyone knows otherwise, I'd enjoy hearing the rationale. It's not a factor I've really thought about till now. Also to hear of more circumstances favoring one or the other side, or exceptions. Or if I'm completely wrong [...okay bogg I asked for it, make me bleed arterial humility ... :) ]
There are things like protected passed pawns, or safe advanced pawns, etc that may favor either side that has them. {There is a general rule that less advanced pawns are defensively safer from capture in general, and more advanced pawns are offensively more dangerous ... if they can be maintained or utilized anyway}
On a personal point: The endings I first chose to specialize in were somewhat K&P since that is numenal; moreso R+P's and N's+P's. The first is the most common ending after K&P, while the latter however, while often drawish to the well versed, seems to be one of the most poorly played endings in general, by the most players. So it has resulted in many wins and was well worth the effort. Certainly more wins from N vs N endings than I've had in B vs B ending, where it is entered with even material. {But in B vs N ... of all the games I see, it tends to favor the B side more in a majority. And as mentioned Fischer games have much to tell.} I beleive this is simply because pawn
structures upon entering a minor piece ending probably are B-favorable more often than otherwise. EG more with multi wings than pawns to one side {which are drawish anyway ... rather than especially N favorable. And that seems to be a large part of the difference. Of course the side with pawn structure unfavorable to their side of the argument will try to avoid trading down into that ending, we would expect.}
Regards, Craig A.C.
schnarre 47 ( +1 | -1 )
I've tended to favor the Bishops personally, but that's me. They are the only piece aside from the Queen that can span the board on the diagonals; their shape can allow them to hide well among pawns in OTB play (I've seen a fair amount of that in casual games). If a Knightshould be in one of the corner squares, such as a1 then a Bishop can be parked at a4/d1 effectively immobilizing it.
It's a trade off in any case, but ultimately I've found it to come down to which of the two one can use best..
ironbutterfly 8 ( +1 | -1 )
Thanks, iona and craig, for looking into the question so deeply!
ganstaman 208 ( +1 | -1 )
General thinking is that bishops are stronger in endgames despite their color-disadvantage. Key reason is that, maybe ironically, they stay on the same color always. So a bishop can block a pawn (or at least attack a square in its path) while moving all about the board (staying on 1 of 2 diagonals, but parking anywhere on them). The knight, on the other hand, can move away but then must return to the square in front of the pawn if it wants to keep up the blockade. It can never drift too far away.

The general strength of anything is how much it can attack/defend -- what is its mobility? A bishop is strong when it has scope, freedom, a clear path unblocked by unmoveable pawns. A knight is strong when it can hop all over the place and hit on squares other pieces can not. Rooks have little value in the opening, but once things start to clear up and they actually start to attack squares, they become important. That's one reason why rooks are powerful in the endgame -- they actually have room to maneuver and squares to attack.

It never hurts to repost this: ->

Or quote it: "Let's start with the age-old question of bishop vs. knight. The conclusions are clear and consistent: although the average value of a bishop is noticeably higher than the average value value of a knight, this difference is entirely due to the large value of the bishop pair. In other words, an unpaired bishop and knight are of equal value (within 1/50 of a pawn, statistically meaningless), so positional considerations (such as open or closed position, good or bad bishop, etc.) will decide which piece is better.

This applies regardless of whether there are few or many pieces on the board.... Although single bishop and knight are equal against each other, my research confirms Capablanca's claim that the bishop is a bit better than the knight when fighting against a rook or (in the endgame) against multiple pawns."
grunfeld 10 ( +1 | -1 )
Thanks everyone for taking time to offer such helpful responses.

It's much appreciated.
ionadowman 97 ( +1 | -1 )
Great posting Craig! ... I've learned a fair bit from your comments (especially that remark about rook pawns). Generally, I've relied on "feel" in assessing endgames, so it is useful to get solid info.
Your remark about looking at potential pawn endings arising from the exchange of minor pieces in a B+Ps vs N+Ps ending is well made. Apart from anything else, one needs to take into consideration such exchanges, as such an exchange might represent one's only winning chance, or, on the other hand, might compromise one's chances of a draw.
I also agree with your observation that if one has a minor piece plus pawns vs rook plus pawns, your chances of holding are (generally!) better with the bishop. R+nP vs B+(n+1)P is extremely hard to win if the bishop side has no positional weaknesses! (I'm also inclined to think the B complements the R better than a N does, but bear in mind N+R can without help checkmate a king; B+R can not).
ionadowman 125 ( +1 | -1 )
Leonard Barden's... ...Guardian Weekly column (8 June) has a piece on Alexander Morozevich, described as "today's equine defender". Opening the article, Barden mentions Mikhail Tchigorin as one who often favoured the knights over the bishops, and I recall that Capablanca mentioned Frank Marshall having a prediliction for the knights... The game N Short vs Morozevich is a fine example of the knights' domination over the bishops when conditions are right.

I've seen many games in which the knights looked stodgy against the flashy bishops, and others in which the bishops might as well have been glued to the board, so lacking in mobility have they been, whilst the knights seem to have been equipped with rocket packs. I've just recently annotated a GK game aiel vs ionadowman in which the contest between 2B vs 2N is a feature. But perhaps this game isn't such a good example, having other, maybe more important, motifs! I hope this game is available: I've made it public, but it doesn't appear on the "all annotated games" list on my machine...

ionadowman 63 ( +1 | -1 )
I've discovered... ... that the game aiel - ionadowman is on the annotated games list - at page about 7. Now, I sat on this game as a private listing until it had all the annotations I wanted to make, so it appears that the games are listed according to the date at which they are scored for annotation, not the date when they are posted on the public list. Pity. I have a 100-mover still waiting for annotations to be completed. No doubt by the time it bursts upon an awestruck public, it will be buried on p.14... :-(
ccmcacollister 145 ( +1 | -1 )
ION ... I believe that I may have a solution to that Annotated Game posting delay problem you mention. Thus I am posting it here so that (1) Others can see it, if it really does work, and (2) So that others can see it and tell us if it Does NOT WORK at all !
My idea is this: I believe that you should be able to go ahead and continue as you are, adding notations at the rate you wish. Then when it is done and ready to be posted to PUBLIC, use the feature that lets you Export/Download{I forget which it is called} a PGN copy of the game. Then browser COPY that PGN formatted copy of the Game+Notes that the Annotate section will make for you. Then DELETE the Annotated Game totally.
After deleting, start a New Annotation using the same game, and paste your PGN GAME+NOTES copy that you made onto the IMPORT function page of the ANNOTATE section. And I think that it should be simply to SAVE it, then OPTION to PUBLIC & SAVE ... so that it will then have the Present Date as its creation ... and so place it upon PAGE ONE !? Would it not?
Of course, if it were me, before DELETING the original annotated game, I would take the COPY and immediately save it by pasting it into a PM to myself. [ or to e-mail or notepad, etc. Whatever method you use to save texts here. I like to PM myself, with an e-mail copy to me, as my preference.]
Does this sound like it makes sense to anyone!? ...and will work !?
ccmcacollister 37 ( +1 | -1 )
ps ... Thanks for the kind words, also~! :))
AND YES ... I did see the Aiel game appear in the public games. I've heard some people say tho, that sometimes the Public display has not taken effect immediately for them, at the time requested. But perhaps later that day?! Don't know if that still holds or not, however.
ultimatt 218 ( +1 | -1 )
Replying to post 3 This is the post I'm replying to:

Thanks for the reply Iron. You could be right. But my argument hinges on the fact that you can evade a Bishop forever by sitting on the opposite color but you can't hide that way from a Knight.

I guess it would depend a lot on what other pieces you have to corner the King.

While the bishop can only move to half the squares, and thus the knight can hide from the bishop by sitting on a square of the opposite color, a knight simply sitting around won't do much good. You have to remember that a knight always moves to a square of the opposite color. What this means is that a bishop can, given the right situation (namely an open board), absolutely dominate a knight. Consider the following position (no other piece shown, for simplicity):

(I don't know how to use fen code, so it's just a URL for an image)

Notice that no matter where the knight moves, the bishop can take it. A knight can never do this to a bishop.

Now, yes, this is only one example and it would take a bit of luck to get in such a position, but it is an example of how a bishop can be better than a knight in the end game.

I tend to think that knights are better in the opening and middle game, and that bishops are better in the end game. A bishop and queen are more effective than a knight and queen I think (of course this depends on individual positions, but I'm speaking generally here). The ability to connect the two pieces is important in the end game. Note that if a knight defends a piece, that piece cannot be defending the knight (unless it is another knight). And vice versa, if a piece is defending a knight, then the knight cannot possibly be defending the piece (again, unless it's another knight). This restricts possibilities quite a bit.

One of my biggest reasons for favoring knights earlier in the game is for their forking ability. That potential doesn't exist as much in the end game, because the pieces tend to be farther away, and with lots of open space, can move far away if a fork is imminent. So, yes, I think bishops are better in the end game (generally) and knights are better in the opening and middle game (generally).
ultimatt 18 ( +1 | -1 )
Oops Shoot. The URL for that image I mentioned is here:

kewms 31 ( +1 | -1 )
For what it's worth, I just won an endgame in which I was a full piece down. I had to sacrifice my bishop to stop a passed pawn. That left him with an extra knight, but my pawns just ran away from it.
board #5950797
ionadowman 136 ( +1 | -1 )
ultimatt... ... shows one of the great strengths of the bishop vis-a-vis the knoght:
The paralysis of the knight is complete. But the knight can be effective against undeveloped or cramped pieces, e.g.
If it were White to play, 1.Kxd8 leaves him with a winning endgame. But it's Black to play. The bishop can run ... but where can he hide? Wherever the bishop goes, the knight attacks it with a discovered check, and the bishop falls.
In this fairly closed ending, it is the knight that dominates:
1.Nd6ch. If 1...Kd8 2.Nf7ch simply picks up the h-pawn. The same holds for 1...Ke7 2.Nf5ch
1...Kd2 2.Nf5 h5 3.Ng7 and the h-pawn is again a goner.
1...Kf1 2.Nc8 picks up the a- or b-pawn.
In the meantime, Black's bishop has no targets.
Right then: retreat to the K-side
Finally, a more elemantary endgame, in which the knight defends effectively against an advanced flank pawn:
White is a piece for pawn ahead, but Black's extra pawn is one square from promotion. White's knight is defending the queening square, and Black would like to drive away the knight. But watch what happens. 1...Kd3 2.Nb4ch or 1...Kd5 2.Nb4ch and the the Black king has no other avenues by which to advance.