Knight vs Bishop summary(/essay)

Sort:
tarius78

Hello all:

I've just responded to another forum on this 'hot-button' issue in the chess world, and my answer proved much longer than I thought. Still it was not complete, but long enough that I feel it needs its own posting on a new comment string. Many people have been very one-sided on this issue, and that is clearly the wrong approach. Please consider my summary below and provide feedback and/or additions and necessary:

[WARNING: THIS IS BASICALLY A MINI-ESSAY - SO GRAB A CUP OF JAVA!]

A few things/facts that need to be kept in mind (with knight vs bishop):

1)Every game is different. This means, different contexts and space (closed/open), different material availability (e.g. # of each piece), and different dynamic (the time situation for each position - tempo issues).

2)What makes a piece of value or not is primarily whether or not it is being used properly and/or is well situated. A bishop, even if one were to value it higher than a knight, is worthless on its original square when all the action is happening elsewhere... Therefore knowledge of how to use each piece and where to place it (like the general strategy of placing knights outposted on the 5th rank) will make the difference.

3)Bishop pairs are of great importance. You'll see in many analyses that when one loses their bishop pair (only have 1 left) compared to another oponent with both bishops and other material being equal in 'points', then the bishop pair player usually has the advantage barring some great positional dsicrepancy. The reason for this, as was mentioned early when it comes to the # of squares accessed by knights vs bishops, is that with both bishops around, the dissadvantage of the bishops compared to knights is basically cancelled out. Indeed, if we are to survive the midgame, the board will inevitably open up so having both bishops in the endgame is a clear advantage.

4)Mating possibilities must be considered. Should we have a choice between 2 knights , 1 knight and 1 bishop, or 2 bishops for an endgame against a lone king, every player knowledgeable in the basic mates knows that the possibility of mating increases proportionally to the presence of bishops (that is to say they have been listed here from least to most possible). As such, the mating difficulty decreases according to the above order. Indeed, 2 knights is impossible without a deliberate blunder by the oponent, knight and bishop is the hardest of the basic mates to accomplish in the 50 move period, and 2 bishops is much easier, though still a mild challenge. This is important in considering the value of the pieces.

5) Knights are known to be of great use in blocading in the endgame. So as a defensive tool they are great for blocking off  pawns that would otherwise march to promotion or to remotely secure the queening square. Because of the knights unique attack pattern, it is great for harassing the oponents king. Few other pieces can consistently target the king like the knights, as often the availalble squares to a king are limited by the attacked squares of the knight, and those that aren't can be directly attacked by the knight on the next move. This unique attack pattern also makes it hard for the king to both protect the queening square and to attack the knight in many situations. For a bishop, the king need only stay on the opposite colour. So even in the endgame, depending on the situational needs, the knights may be more valuable.

6)The bishop can form a battery with the queen, whereas the knight can not, nor can the knight 'connect' with any other piece strongly except for itself. As such. Though once again, this is more relevant when the board is open enough for batteries to be of greater value.

7)Lastly, the knight is the most common choice for underpromostion because of its unique abilities.

Conclusions:

-If you are relying on minor pieces for mating, or if it seems that it may come down to that in the endgame, then clearly the bishop is of greater value.

-If you have a choice between having a pair of knights and a pair of bishops then choose the bishop pair. If you have already lost 1 bishop, the relevance is far less, and indeed more often than not, you should elect to keep the knight.

-It matters which pieces your oponent has left as well - also do you want a balanced or imbalanced situation? If imbalance is sought, usually I choose piece diversity over strength in numbers of one kind. More possibilities that way..

-Overall there is far more potential with the knight. If you need to keep your options open, or are the kind of player who delays castling until the last possible second (to keep the oponent guessing at which side you'll castle) then choose the knight over the bishop.

-If the game will be a very open one, and/or you are racing towards the endgame (and especially if any or many major pieces have already been exchanged) choose to keep the bishops, i.e. they will be of greater value.

-If you still have you queen, then the bishop can be sacrificed well before the knight - no one else moves like the knight does!

-Knights can get in and out of the action/melee better than any other piece. When the situation calls for up-close-and personal - definitely stick to the knight like as the game approaches the midgame and things are more closed off. In otherwords, if you are aiming to mate in the midgame the knight is stronger usually. If you are trying to mate in the endgame, the bishop will be of greater value more times than not.

 

Hope these points clear up some of the issues, and have helped out. I welcome and debate/banter :)

Torkil

Thank you for this great overview, this is much more sensible than most other comments I have read concerning the knight-bishop issue.

A few remarks:

- As you have observed correctly, the knight is a valuable piece in tactical melées, due to its unique forking ablities. Yet I wouldn't agree that it is the more valuable piece for a middlegame kingside attack.

- The knight is sometimes referred to as the "best defender of its king". Obviously this is due to the fact that for defending a piece usually doesn't have to have a long range. In this context the following basic example may be of interest:

 

 

 

 

 

 

 

 

 

 

 

 

 

The knight blocks the direct attack on the king, while at the same time it covers the squares e3 and g3, thus preventing any direct checks. Patterns like this have helped me defend sharp positions more than once (with some other material on the board as well, of course).

- Also you have adequately noticed the knight's value for blockading pawns. This is extremely true in the middlegame where there are usually other pieces around the pawns in question, which in turn can be attacked by the knight. In the endgame the knight is stil a good blockader but tends to get into difficulties with distant passed pawns, where his little range may come to tell. In cases like that (minor piece endgames with few, scattered pawns), it is often the bishop side which is superior.

- You point out the value of the bishop pair, quite truely so. Thus you reach the conclusion that you should maintain that pair during exchanges, if possible, but that if you have only one bishop left, it is of little importance which piece you swap off. While this is generally true, an important extension of this may be the question if the opponent still has got the bishop pair, and which of your pieces is likely to be able to force an exchange against one of these bishops. Making this question a priority is likely to save some difficult games.

- Some players maintain that in endgames the Q+N constellation has got the advantage over the Q+B team. As I see it, this has not yet been proven.

Anyways, in his superb book "Secrets of Modern Chess Strategy" John Watson devotes quite some space to minor piece issues and claims that the correct use of the minor pieces is a major part of the strength of today's grandmasters. I would recommend this book to everyone who is interested in further details on this issue.

I'm looking forward to your comments on this ;)

Cheers,

Torkil

Omicron

Hi, thanx to both for your interesting aproaches on this issue. I've been meaning to post a thread on knight Vs Bishop but haven't had the time to do so in a decent way. I'll try to post a game of mine tomorrow, that I wanted to share, where I decided to stick to bishops in a very tough endgame. I thought of myself as pro-knight kind of player, but after that game I definitely understood It's very important to discern what minor piece suits better a given position.

tarius78

Thank you Torkil, for your insights - very useful indeed.

You are absolutely correct about the value of the knight as a defender. As I tried to allude to in my mini dissertation, when on defense, the knight can sometimes out-preform the bishop against very strong pieces (mostly due to its access to both coloured squares). Its range, of course is its dissadvantage.

As stated on Chessvideos.tv in the endgame simulator, the Queen vs Knight and King, is more challenging than the Queen vs Bishop and king. I would state that given the battery preformance possibilities, the queen and bishop together instead of queen and knight together on the offense is more powerful (battery, plus makes a wall and can penetrate pawn  lines easily).

Moreover, I will certainly consult the book you mentioned - seems quite useful and informative! Can't wait to take a peek..

Thanks again...

likesforests

tarius78> Indeed, if we are to survive the midgame, the board will inevitably open up so having both bishops in the endgame is a clear advantage.

The bishop pair is an advantage, but the board doesn't inevitably open up, even in the endgame. It's up to the side with the bishops to work to open it up. In many, many endgames the side with the knight pair kept it closed and won.

tarius78> Some players maintain that in endgames the Q+N constellation has got the advantage over the Q+B team. As I see it, this has not yet been proven.

That Q+N beats Q+B is usually attributed to Capablanca. There are legitimate reasons to disagree, but to be fair at least state the source.

Another point not mentioned in favor of the knight is its trickiness. At least one (verified) NM on here says it's the best piece to have while behind. Tongue out

Overall, nice mini-esssay. Well spoken.  :)

tarius78

Thanks likesforests for the feedback. When I said that the board will innevitably open up, I think that covers more than 85% of games, unless a mate or clear material advantage surfaces during the midgame. That's why I said 'survive' the midgame, meaning that both players will enter the endgame with good chances/prospects relative to each other.

Second of all, I wasn't actually the one to make the original claim that Q+N beats Q+B if you re-read, you'll see that Torkil said that, not me. What I maintained was that (as verified by the endgame simulator on www.chessvideos.tv endgame simulator which ranks the various mating or drawing problems in order of difficulty) Q against a N was more difficult than Q against a B for mating. Try it yourself, you'll see what I mean. So I agree with you, Knight is the best piece to  have when behind.

Regarding trickiness, I chose to use the word 'unique' regarding its pattern since those well accustomed to the game of chess will recognize this otherwise 'tricky' pattern quite clearly. Just takes practice/experience, like everything else :)

Thanks for the Capablanca source though ;)

likesforests

tarius78> When I said that the board will innevitably open up, I think that covers more than 85% of games,

The board doesn't inevitably do anything. It's up to the players to make the board favorable fo their pieces. See Lasker vs Chigorin. Regarding the 85% statistic, Chessbase says the Bs win 48% of the time. They have an edge, but it's not 85% overwhelming.

tarius78> Second of all, I wasn't actually the one to make the original claim that Q+N beats Q+B... Thanks for the Capablanca source though ;)

I simply read your mini-essay and commented on it, as you requested. I quoted that line from your essay. I see you removed it now in its entirety.

GM Flear thinks Q+N and Q+B are roughly equal. Capablanca thinks Q+N is better. I'm not too interested in how chessvideos.tv rates the Q+B vs Q+N mates, given I've done both in the past few games and they're quite easy.

tarius78> those well accustomed to the game of chess will recognize this otherwise 'tricky' pattern quite clearly

It depends what you mean by "accustome to the game" and "quite clearly". I've used 'tricky knight moves' against 1900s before, and it was an NM who said it was good piece to have when behind due to its trickiness.

Knight vision isn't good or bad, your skill can be rated on a continuum.

Try this:

  1. Close your eyes.
  2. Visualize a rook moving from b1 to a8. Name all the squares it crosses.
  3. Do the same for a knight.

I bet most see the rook path much faster.

Or:

Get a board and place a White knight on a1 and Black pawns on c3, f3, c6, and f6. Start the clock. Try to get the knight to a1, b1, c1... h1, then h2... to a2, then a3... h3... until you have toured every square on the board. What's your time?

Torkil
likesforests wrote:

Regarding the 85% statistic, Chessbase says the Bs win 48% of the time. They have an edge, but it's not 85% overwhelming.


Ok, as I see it nobody claimed that the bishops win 85% of the games. Tarius wrote that he thinks the game opens up in the endgame in 85% of the cases, which is still difficult to prove or refute, as he didn't exactly define what he actually means by "opening up". I suspect this statement refers to the fact that pieces and pawns are exchanged on the way to the ending, thus clearing lines for the remaining pieces, although I tend to agree that locked pawn structures can arise in more than 15% of all games.

 

Right, so it may have been Capablanca who first stated the supposed superiority of N+Q versus B+Q, I didn't know that.

However, considering the source I mentioned (J.Watson: SoMCS), I got the impression that this assumption is shared at least by a large part of modern top players. Watson goes quite a long way to put that statement into doubt, with what seem to be good points to me. As this seems to be a complicated matter, I would like to refer you to his book rather than try to explain the whole thing here - not sure if I could.

likesforests

Torkil> However, considering the source I mentioned (J.Watson: SoMCS), I got the impression that this assumption is shared at least by a large part of modern top players.

I'll take a look next time I'm at the bookstore. I have Flear's Practical Endgame Play--beyond the basics. He spends a chapter on Q+B vs Q+N and seems convinced they're roughly equal, even after studying Capablanca's games. It's an intesting topic. :)

GeneralMiller

I have always had a preference for knights over bishops, but have found that it is imperative to use them well together in most situations.

MrZugzwang
Gonnosuke wrote:

Excellent post.  If you've not read it, I highly recommend IM Kaufman's "Evaluation of Material Imbalances".  The section on the importance and value of the bishop pair is especially enlightening and reenforces some of what you mentioned.


 As a longtime fan of Larry's it's nice to be able to now say "GM Kaufman" *(since he won the World Senior Championship.)

tarius78

I have read this most excellent article/study by Kaufman, and yes, I must say quite informative! There are a couple of shaky points when he fudges around the fractions a little , but by and large this article provides some very good guidelines about evaluating material and imbalances...

When it comes to the bishop/knight contraversy, he is somewhat elaborate, but makes a slight error in my judgement of focussing too much on the "redundancy" of all the other pieces besides the bishops, in that they will cover the same squares, whereas the bishops are mutually exclusive. In fact, what I concluded (and as I have for sometime now) is that a general characteristic is that knights have greater defensive potential, and bishops have greater offensive potential. I will clarify: this 'redundancy' mentioned, is a negative spin on an otherwise desirable trait in many situations: batteries. The ability to connect pieces is just as strong a factor as the lack of redundancy of square control mentioned by Kaufman. They are just two different strategies. Let me expound that idea:

1) When pieces can connect to each other and form batteries (like Rook and Queen, or R-R, or Bishop-Queen) they gain great strength because the back each other up and thus protect each other. This also allows for some great and devastating attacks (hence the name) and in a great deal of situations, it is a great asset, strong and versatile.

2)When pieces can not connect, as a trade-off, they can attack more squares, since they are exclusive in this manner. The bishop pair is the prime example, as stated by Kaufman. They can not support each other like the other 'like' pieces, so instead they form a nice 'double' barrier (as exploited in the 2bishop vs K basic mate) which launches formidable and controlling attacks. Their potential range also makes them great 'snipers' of distal unprotected or weakly protected pieces, breaking formations and sneaking behind enemy lines through diagonals. As a matter of fact, for several years, I've reffered to them as my 'assasins' as they have for a long time been my favorite piece to initiate exchanges with, and many beautiful combinations are initiated with bishop sacs or uneven exchanges (how many combos have you seen which start with or involve the famous Bxf7+!). Indeed, this common trend of using bishops for exchanges and gains in position is noticed at the grandmaster/master level by Kaufman in this same article. As such, the bishops are overall a more aggressive piece, and generally more one-sided in that they specialize in attacks. (Though obviously not limited to them!).

 

So overall, I would conclude that depending on your style of play, bishops will be more valuable if say you are a more agressive player in general. I think that the most patient and longer-game preferring individual will prefer the knights, hence for defensive play. We should be able to correlate these trends with a preference for open, vs. closed games perhaps...? However, like in most things, the advantage goes to those who know how to change gears most effectively. As such, to be most versatile, and given the ability to recognize what the situation calls for, optimal use may be found for both piece types, given the opportunity, which will inevitably come. Afterall, as Louis Pasteur said:

"Chance favours the prepared mind".

tarius78

Thank you Tonydal, for supporting my claim about space inevitably opening up. That was what I was trying to say earlier, that most positions are open, not closed.

Yes you are right about the contraint given closed positions. I would say though, that one must just be more patient in closed positions, with pieces other than the knight that is, and shift them around pro-actively preparing for the 'inevitable' opening up of the board. When the position is closed, the game simply seems to go into a sort of 'slow-motion' phase, where each small detail will count greatly for the 'speeding up' phase to come. It's a chance to set things up and accumulate small advantages in my opinion. So I don't really see how that particular fact boosts the value of bishops however. If one can exploit the knights to their full potential when other pieces are slowed down, then the contraversy rages on... But it should not be such a contraversy, for as I have just stated in my last post, it is a matter of being observant and opportunitic - having your 'finger on the pulse of the game' as it were. As likes forest mentioned in post #6 above, one can 'work hard' to keep the position closed, or at least maximize the advantage it gives you, if it does indeed provide one, or at the very least prolong this 'slow-down' mode to increase chances of an advantage 'breakthrough'.

Regarding the game starting open, yes, again, undeniable. But many players are quick to taint the center with pawns, and knigt-pawn barriers, whereby, it is common to have to wait to put the bishops to proper use. Perhaps that is one of the reasons for the old rule of thumb " Knights before bishops" .

Then again, Tonydal, perhaps you are generally more agressive (?)... A hypothesis to be tested.

tarius78

I see... well, can we then claim that as black, the knights are generally speaking preferable than the bishops, since it would be in the better interest of black to seek a closed, 'clogged up' position, to enchance chances of a draw, following your line of thought?...

This is indeed new thinking on the subject: please, tonydal, everyone, explore/discuss/expound - I (as I'm sure are many others) are interested!

erikido23
tarius78 wrote:

I see... well, can we then claim that as black, the knights are generally speaking preferable than the bishops, since it would be in the better interest of black to seek a closed, 'clogged up' position, to enchance chances of a draw, following your line of thought?...

This is indeed new thinking on the subject: please, tonydal, everyone, explore/discuss/expound - I (as I'm sure are many others) are interested!


 Why does everyone else want to draw(and in boring fashion) when they play with black.  I like to crush everyone no matter which side of the board I am on. 

 

As one of my friends says -I love blood on the board, whether its mine or yours.  

tarius78

Hehe, good post, good comment, and good point! I'm often the same way, action action action I say! (Hence my affinity to the KG opening, esp accepted!).

However,  here I was just exploring a valid point that Tonydal brought up, not that it's necessarily my preference, but it may make for good chess! Of course - we all seek the win, but let's face it, esp at higher levels, when playing black, you hope to draw given a formidable oponent, and otherwise keep your eyes open for tactical oversights on the other players part, and stay opportunisitc with a killer instinct if the chance should present itself..

Generally speaking, the slight edge you have a white can be capitalized on by keeping the pressure on black, especially in the beginning, but this may have a lasting effect throughout the game. One great way of keeping pressure is by being very aggressive, and following the thread of this discussion and previous points made, agressive=bishop use (see my 'assassins' comments earlier!). If this approach is adopted, it is only natural that black will have to think 'defence', and hence, again according to the ideas explored earlier, may want to hang on to those useful knights.

No need to mention that offence may make for great defence, but I find the knight more flexible in this department, despite greater mobility of the bishop. I think that that is one of the main differences between the pieces: uniqueness. The unique pattern of the knight, as comapred to the 'redundancy' of the bishop, granted the queen's presence, can not be overlooked. Since, nothing can replace a knight, except another knight, it is always sad to let it go, particularly when it's your last one! So one good rule of thumb I think is "Try to always hang on to at least 1 knight, if possible/reasonable, or all other factors being equal." This uniqueness (highlighted by its direct attack motif) gives the flexibility of being able to hide behind other pieces like pawns and still be effective, yet also overcome all other obstacles and jump right into the mix. The '3-dimenionality' of the knight is invaluable, as long as it is appreciated. This factor is what causes the debate to rage on, and why the answer "bishops are better, slightly, overall" is not so cut and dry!

No I think that the last few posts are on to something here: defence vs offence, open/closed, what other pieces are left, and 'redundancy' vs uniqueness vs 'connectivity' of pieces issues - these are the true guides to a proper answer on the issue, a balanced, measured, and comprehensive approach.

Keep the ideas coming, I feel we're making progress!

MrCheesec4ke

https://www.chess.com/blog/MrCheesec4ke/knights-vs-bishops-the-battle-of-the-minor-pieces-1

LaraPeter

www.chess.com/forum/view/general/knight-vs-bishop-summaryessay

LaraPeter

Knight vs Bishop summary(/essay) - Шахматные Форумы - Chess.com

igorfernando3

Да, тема довольно интересная. Недавно наткнулся на тему, тоже интересную, но довольно сложную. Поэтому мне очень помог сервис написания эссе    https://www.essaywritinglab.co.uk/     С их помощью работа получилась интересной и полной, а главное с высокой уникальностью.