Knight vrs bishop

The bishop is good for long range (think army sniper), the knight is the oddball you need and can be quite a surprise in dense battles. While the rook is like pretty handy. All have diff purposes that they excel at. Chess is a wargame so the Bishop would have been thought of as an archer type piece good from a distance but pretty weak close up while knights were good for close in fighting against infantry while rooks were meant as support.

In response to the first paragraph:  I'm a bit surprised about your evaluation, it really depends on the position!  Perhaps the openings you play rely on such attacks or strong bishops?

a,b,c and d round it out, ok.  But you act like those situations are rare?  I think players are more apt to preserve bishops because on the outside chance you survive until the endgame (or an otherwise open position) with them both it will likely go very well for you.  But during the course of a game, a,b,c and d pop up...  again these are not rare situations.

Not to beat a dead horse, but knights make excellent attackers.  If you're a dragon fan you might smirk, but I'm talking in general terms not specific openings :)  Often I feel like my bishops are a bit lame.  Their mobility is great but they're bound to one color complex.  Knights take a bit longer to get around (so in open games can become lame) but once they "get there" are much more useful IMO.

Anyway Kaufman did that analysis of a bunch of master game to come up with the reltaive values from scratch and the knight and bisohp ended up within 1/64th a pawn from eachother or something like that.  You can read it here under THE METHOD heading http://home.comcast.net/~danheisman/Articles/evaluation_of_material_imbalance.htm

Yes, that rook has problems.

The close equality should not come as a surprise, because pieces that are close in value tend draw their values towards each other. The reason is that treating the theoretically superior piece as more valuable (i.e. constantly having to avoid trading with its inferior counter-part) suppresses its practical value, and might easily suppress it more than the theoretical value difference if the latter is small. So it would be better to handle it as if they were equal (i.e. not avoid trading). Which in practice means they will be eventually traded, which makes them equal.

The same principle is responsible for the fact that although Q is more valuable than R+B, 3Q is actually inferior to 2Q+R+B. There the option to 'rescue' the value of the theoretically superior piece by switching to a trading strategy does not work, because 1-vs-2 trades are not easy to force.

So the only right answer seems to be:
It depends on the position.

You need to learn in which positions bishops are better and in which knights are better. But at lower ratings knights are often better because they create chances for forks that people seem to miss.

Let me give you an example. Max Euwe's Practical Chess Lessons 5 has an entire chapter on this subject.

That makes no sense.

Try to trade bishop for a rook, or even better queen.

Example of a knight better than a bishop:

Example in the endgame

I'm sorry but these "guidelines" are simply of no use.
You need to evaluate the position and then make a decision. A guideline is only handy when a significant amount of positions can be treated that way. That means like 95% of all positions, which simply isn't the case.

It really depends on what kind of endgame. Look at the game I posted. That is an endgame where a knight is much more valuable.

Of course. Thanks for the correction.

At least I can say that a pawn is never better than a queen.

Thanks for the nice analysis, Loomis!

Wow, great post ReasonableDoubt!

From an amateur’s point of view, bishops perform well in open positions and knights in closed positions.  Also a bishop is able to pin a knight down when there are two squares open between them.  Pawns are also able to do this to some degree if they are well placed.   On the other hand, a bishop’s fate is to be stuck on the same colored square for the entire game, and this is where the knight can take advantage of the situation.  But a bishop pair is very powerful, especially when the stages of the endgame are reached. 

Also, a bishop is a powerful weapon protecting a pawn's promotion square from a distance.  On the flipside, the knight can be used as a wonderful pawn blockade, and its tentacles reach over to other squares as well while blocking the pawn.

Knights are at home on support squares, and bishops are quite happy when there are no pawns blocking their way, unless they can sneak in from behind a line of enemy pawns and gobble them up.

I found the post from ReasonableDoubt very interesting, but one has to understand positional concepts of chess more clearly in order to fully grasp the ideas behind his post.

I am a huge Jeremy Silman fan, and I would very strongly recommend his book “The Amateurs Mind”.  It explains the bishop vs. knight concept (as well as a variety of critical positional chess concepts) in a clear and concise manner so that we patzers can understand what a far strong player like ReasonableDoubt is teaching us.    Mr. Silman also has an award-winning endgame book, as well as two other books on positional chess.

But FINISH reading “Amateur’s Mind” before moving on to the next ;-)