Bishop pair verse Knight pair

[Kevin O'Rourke]

Bishop pair verse Knight pair

I’ve not had this happen very often but occaitonally it crops up. Assuming there are pleny of pawns left and perhaps each side has a rook or 2, is the Bishop pair the strongest thing in the world in this situation? i.e. taking on 2 steeds? Is it a position you should strive to achieve or is it not as strong as one might have hoped?

Lastly, is there something to be said for the knight pair as a formidable force in chess? I don’t think I’ve ever heard the term knight pair used.

[soheil_hooshdaran]

I saw Glenn Flear dedicated a section (or chspter) to 2 knights versus Bishop+Knight in his practical chess endgame.
The knight don't have color-blindness like one bishop

[Kevin O'Rourke]

I must admit that I would find it difficult with the 2 knights against a bishop pair.

There’s something about a bishop pair that could spell trouble ahead.

[Barry Sandercock]

Better to be left with a King and two Bishops against a lone King, rather than a King and two knights if you are hoping to checkmate !

[Kevin O'Rourke]

Oh god.. I wouldn’t know how to do that. Can I please have one pawn left? 

[Barry Sandercock]

No, afraid not.

[Jonathan Bryant]

I don’t think I’ve ever heard the term knight pair used.

Probably because a second bishop fixes the obvious problem you get when you only have one bishop. Another knight is just more of the same.


Some time ago Jan Gustafson did a broadcast and responded to a similar question by saying he preferred bishops to knights and also that he preferred rooks to knights for the same reason - in most cases they’re simply the better piece.


Incidentally with regard to mating with two pieces I much prefer to have one knight and one bishop than two bishops. I find it easier - probably because I"ve practiced BN v K much more than BB v K.

[Matt Mackenzie]

Mating with two bishops really *is* easy compared to B/N, though.

[Jonathan Bryant]

Mating with two bishops really *is* easy compared to B/N, though.

I just can’t do it. I get a block at the last little stretch.

BN v K on the other hand, is essentially remembering three positions. Absolutely impossible if you don’t know what you are doing, I grant you - you couldn’t win it by accident like I probably could BB v K in some situations - but if you do know it’s quite straight forward.

But as I say, if I actually practiced BB v K I’d get it too.

Not that I’ve had either in a game.

[Joey Stewart]

Two bishops vs knight might be a theoretical draw but is tough to defend in practise - whenever I have had to win it in blitz games the knight almost always ends up getting isolated from the king and picked off. Certainly, a position you could justifiably push for the full 50 moves without any legitimate complaints of lack of progress.

[IM Jack Rudd]

Two bishops vs knight might be a theoretical draw but is tough to defend in practise - whenever I have had to win it in blitz games the knight almost always ends up getting isolated from the king and picked off. Certainly, a position you could justifiably push for the full 50 moves without any legitimate complaints of lack of progress.

It's actually a theoretical win, but some of the wins take more than 50 moves.

[Stewart Reuben]

Two knights protecting each other are sometimes called 'conjoined'. It used to be Siamese, but the language has moved on.
b1 and d3 or similar are sometimes called the 'blind spot' for the knight as it takes 4 moves to get from one to the other.

Fritz 13 opinion. Normal position White +0.39.
White without 2 knights against Black without 2 bishops. +0.63
White without 2 bishops against Black without 2 knights. -0.17.
That suggests 2 bishops against 2 knights is worth quarter of a pawn.

[Kevin O'Rourke]

Nothing worse than having a Knight in the corner of the board.

[Geoff Chandler]

Stats in pure 2B's v 2 N's endings greatly favour the Bishops.
The difference is large enough to draw a firm conclusion
(although of course the actual position on the board is the one that really matters.)

Last time I checked on a 6 million+ database it was something like
3,000 wins for the Bishops and 1,000 for the Knights.

The trick appears to chop a Knight for a Bishop at the opportune moment.
Which is the same for having the two Bishops in any line up v Bishop and Knight.
Often you see in the chosen example one of the Bishop being exchanged to force home the win.

"Nothing worse than having a Knight in the corner of the board."

Except if it is mate. A pretty rare occurrence a Knight giving mate from a corner square.

I know Morphy had one. this is the bones.

White has just played Ng6-h8.

5bnN/5kpp/8/8/8/8/1B6/4R1K1 w - - 0 1

And a cornered Knight Mate is hard to spot coming, especially (perhaps) in time trouble.

Bologan - Tkachiev, Tan Chin Nam Cup 2001.

3r4/b3RN2/p4pk1/Pr5R/3P2P1/2n4P/5P2/6K1 b - - 0 38

Black played 38...Rxd4 (move number 38 is where I'm getting my time trouble guess from, it looks like 38...Ra8 holds on). 39.Nh8 mate.

'Siamese Knights'

I like that.

Sneaking into chess literature for naming such knights protecting each other is 'Spassky's Concept'.

First hinted at in RDK's 'Learn from the Grandmasters' (an excellent book and needs a Vol II.)
and recently in 'Improve Your Chess Pattern Recognition' by Arthur Van De Oudeweetering
where he uses the same Timman v Balashov, Sochi 1973 game to describe 'Siamese Knights.'
titling the game 'Spassky's Concept' (page 237).

'Spasssky's Concept' might catch on, a bit like 'Philidor's Legacy', 'Alekhine's Gun' and 'Horwitz Bishops.'

[PeterFarr]

I think it was in Mihai Suba's book "Dynamic Chess Strategy" that he gave a quip about Knights being stronger on the edge of the board because they don't waste so much of their energy on too many squares.