• GM thamizhan
  • | Oct 3, 2011

Dear Grandmasters Arun Prasad/Magesh Panchanathan:

My question may be too simple for you, but here goes: when reviewing my games, 80% of the time I lose because of knight blindness. I simply fail to see the potential my opponent's Knights have, and, by extension, I must by under-using my own knights. Knights, of course, are the most complicated pieces to manoeuvre, and I am intensely aware of my deficit, but I still keep missing knight moves. 

My question is: is there some silver bullet cure for knight blindness? Some form of practice that will switch the lights on, and keep them on? Or is it simply a matter of escalate my fear of knights, and look at all 8 squares each knight has, every move, so as to avoid my recurring 'knightmares'??

Thank you in advance, 

 David Melbourne

Dear David,

Your question is definitely not a simple one, and to be honest, the more specific your questions are, the better you will be able to understand the solution. I am impressed with the fact that you were able zero in on a problem like this. Normally players would find it difficult to recognize minute patterns like ignoring threats along files/ranks/diagonals or even mishandling a particular piece such as a knight or a bishop. Such patterns are very likely to be present in every single chess player's repertoire, but they tend to go unnoticed.

The reason I say it is difficult to identify such patterns is because a mistake in chess can be attributed to a thousand different causes. Lapse of concentration, lack of energy, lack of sleep, bad opening/middlegame/endgame knowledge, bad time management, bad focus, overconfidence and I can keep adding on to the list. For example if you had lost to a knight fork three times in a row in tournament games, but you were in severe time trouble in one game, slept badly in one game and had bad food in the other game, then I would say your problem is all over the place rather than just missing knight forks. You should be watchful for such errors. In any case, now let us discuss some steps involved in finding and eradicating such issues.

Firstly, this minute level of accuracy in determining such a problem can be a great asset and yet be potentially dangerous too. As a precaution, we will talk about why and how one should come to such conclusions about their game, just so we do not spend hours trying to fix a problem that did not exist. We humans are often influenced by random ideas or thoughts that are put into our heads by the surrounding environment. For example, I have seen several parents who tend to identify a pattern in their children's games which according to them contributes to their loss. This is not essentially wrong all the time, but I find about 90% of such conclusions are based on some randomly collected information rather than proper logical methods. I am not trying to say you probably misjudged your problem, but I am trying to say that there are a lot of pitfalls on such roads and we need to watch out carefully to avoid one. In order to avoid this, try to be precise in your analysis and in your conclusion about your mistakes.

Secondly, even if you were to have identified the problem precisely, the fact that you are thinking that you are going to miss knight maneuvers is going to have an impact on the way you are thinking about the knights during any game. Are you thinking about missing knight moves because you miss them in reality or are you missing knight moves because you think that in your mind to begin with? A little confusing yeah! This is the classic chicken or the egg question. It is probably difficult and also pointless to try to find an exact solution to this question, but we can set some guidelines. It is great that you found a problem with the knight maneuvers, but do not dwell upon that long enough to let that affect you during a game.

Think less, you don't solve the problem,

Think more, you add to the problem,

Think just enough, you solve the problem.

What is just enough? That is a million dollar question you should ask yourself. Only you will know how much is right and probably your results will know even more. Practically speaking, striking the right balance between not doing something or overdoing something is the most difficult thing in life.

The last part of my answer and also the most important part is how do you actually solve it if such a problem really exists. Given that you have identified your problem, what you will need to do is to work on your weakness. Identify positions where knight maneuvers are important (This is where a coach or a working partner would come in very handy) and try to work on them. Analyze such closed positions and solve tactics that involve knight maneuvers. That should put you in the right track.

Here is a very good exercise that will help you understand the power of a knight. It is white to play and win in the following position. Try to think and find the solution first before you continue reading the explanation below.


It is a theoretically known fact that a rook and knight cannot win against a rook which implies that if white loses his pawn on c5, then the game would be a draw. It also makes black's target pretty straight forward, he will go after the c5 pawn at all cost. This also implies that if the white knight reaches the square e4, then white is already winning because black has no way of capturing the c5 pawn anymore and the white king will march slowly back into the center and white will eventually win.

The real problem in this position is that the white knight is not able to reach any safe square to protect the pawn. Let us take a look at the solution now.

That was one mighty knight taking control of one whole file from the rook even though it directly was only able to control two squares (f7 and f3), the rest were controlled through the threat of forks.

Another little exercise you can do is to move a knight around a chessboard in your mind. You could just move it at random; you could pick squares and then try to find the fastest route between them; or if you get really good, you could try doing a "knight's tour": passing through every square on the board exactly once!

Now, if you think you have problems with knights, go on and work on more positions and exercises like these and I am sure you will see visible improvements in the knight department.


  • 5 years ago


    Another good way to avoid "knightmares" is to study knight endgames. The absolute best primer, starter, is the 2N v. P endgame (naturally, the ones where the pawn hasn't crossed the Troitzky line, otherwise it's a dead draw).

    This endgame is very rare, however learning how to win it with all possible positions of the pawn and kings will key the mind in to the leaps and hops of the knight(s). It is a variant on knight tours and visualising knight circuits in one's head. Studying knight endgames also has a double whammy: whilst getting better at handling knights, learning endgames will also improve one's play!

  • 5 years ago

    NM BMcC333

    Crossbow, thanks for the comment, somehow I missed it when you made it. I am glad someone appreciated my basic point. Advice for total beginner may not apply to you once you have stopped dropping pieces and begun to try to keep an initiative.

  • 5 years ago


  • 5 years ago


    @ NM BMcC333:  Perhaps, I would be confused if I learned your statements when I was still 5 yrs. old- the time I was beginning to learn the rules of chess. I learned the basic fundamentals and theories of the three stages of chess when I was eight yrs. old-the time I could understand English language. Now that i'm 13, I am trying to study different books written by strong grandmasters. But as you said: "almost all beginner books is a major reason many players can not improve their tactics and feel so confused while analyzing."... I beganto contemplate last night about your statement above, i'm afraid that I missed a lot of things when I was still a beginner! And maybe my chess study is pointing toward the wrong direction.  I think I need to reexamine what I had missed when I was a newbie. Thanks for will help. Mabuhay ka!

  • 5 years ago

    GM thamizhan

    elindauer - nicely done!

    onParole - Do a hundred is definitely a good piece of advice. To register a pattern without having to think involves a lot of training.

    7beaufeet - Your solution would be simple if you play chess with just the knights. The point is not to try to solve using any brute force algorithm over the board, because during the game you have other important things to work on. The point is to use brute force at home so you would not need to think about knight forks specifically. Your pattern recognition will send you an alarm automatically if there is one possible any time soon. For example if you see a Kg7 and Rf8 for black, then even if your knight is no where close to e6, your mind will register that possibility deep inside and when the right time comes you will try to use it automatically. All this is possible only if you have worked enough with the knight at home...

    jps7chess - It is definitely possible to check all the 8 squares each knight can go to in each move, but that is not a practical solution. Like i mentioned earlier, the point is to just get that natural feel for your knights (All pieces eventually) and that comes only with a lot of home work. However, keeping the knight and king diagonaly one square apart is a good suggestion to keep away from knight tricks.

  • 5 years ago

    NM BMcC333

    Sorry Wesley if my advice is confusing, but chess is a complicated game. It is one of the major appealling factors. There are many other games more simple and a few may be more complex. Go is very mind boggling to me but since I learned chess at age 4 and go @24 i have no real perspective.

    Advice like knights move in an L and to another color may be simple and easy to digest, but how helpful are they to an already competent player? If you didnt appreciate the knight's color change that is very useful, but something tells me the original questioner is beyond that. Queen and knight better than queen and bishop is also useful but it is important (for me anyway) to realize i am proposing a new way to think about planning and analyzing, not just tweaking fundamentals, another issue entirely.

    Lev Alburt once told me that everyone under 2400 should go for knights since they are easier to use.

    I am glad someone is having fun with kamakaze chess based on my advice, but attacking chess does not always have to be directed against the king. That said, if you enjoy attacking the king, find pawn structures that let you do that even if it means sacrificing. Another great piece of advice came from Southern Legend Boris Kogan, "It  does not matter what you play, only WHOM you play!"

  • 5 years ago


    may be another approach is: a spiced up  me first approach as adviced by ( NM BMcC333)  during practice == lots of kamikaze blitz games with focus on checkmate only. i did have a similar problem(may be still have) but after doing the kamikaze exercise i don't fear no knight no more.

  • 5 years ago

    IM dpruess

    beautiful answer!! :-)

  • 5 years ago


    ps posporov051560: Love your solution.  Funny that 1.Ng5#  is also a victorious knight. ;)

  • 5 years ago


    The Janisch position is legal.  The trick is that the pawn structure for black implies that black must have made a minimum of 6 captures.  White has 10 pieces on the board, implying every capture by black was made by a pawn.


    It gets interesting when you wonder what the captures were exactly though... to achieve this pawn structure with only 6 captures, the black e3 pawn must have started on the queenside.  In fact, it implies that black has made 1 capture on the e file, and 4 captures on the queenside.  Since white is missing two kingside pawns, this might appear impossible, but, this can be done if white queens one of the kingside pawns, and then moves the promoted piece someplace where it can be captured by a black queenside pawn.


    Or maybe it's simpler to show "proof by example":


  • 5 years ago

    IM dpruess

    question: is the Janisch position posted by posporov a legal position? it took me half a minute to answer myself when i wondered this.

  • 5 years ago


    Well written.

  • 5 years ago


  • 5 years ago


    In his simplifications of positions, Capablanca favoured knights against bishops, vice versa todays accepted practice of trading knights.And he was master of endgames.

  • 5 years ago


    One useful bit of information regarding Knights may be helpful. Every time a KNight moves it lands on a different color square. Two pieces can only be forked by a KNight if they are on the same color - makes it a bit easier to search for possible forks and to potentially avoid them. Farnsworth's Predator at the Chessboard is great for understanding tactics of this sort. 

  • 5 years ago


    Remember that knight moves in L shape and it can jump to eight squares maximum. So just keep a watch on eight squares on whic opponent knight can go. Beware of forks, family forks etc. Most important, in knight endgames (when your opponent has got a knight, keep your important pieces at diagonally one square apart from the knight (like if your opponent's knight is at e4, then g6 and similar squares are for your king, queen or important pieces. Then opponent's knight will need a number of moves to put you in check. Caution is the only way to handle knights. As per Capablanka, queen and knight combination is much more dangerous than queen and bishop. Capablanka used knights in the best possible way. 

  • 5 years ago


    nice tempo

  • 5 years ago


    This is how Black wins, shoelessjoe25!

  • 5 years ago


    Hey, I asked this question to my friend at the club the other day!


     The answer is simple... pick Knight exercises and " a hundred!"


     Remember the motto: "Do a hundred!"


     Long term investment... you will always see them from there on... but keep doing a hundred... then do another hundred!


    PROBLEM SOLVED!! YOU DON'T HAVE TO THANK ME... Money mouthMoney mouth

  • 5 years ago


    @shoelessjoe25 Draw can't be achieved by repetition. If white tries to repeat by moving rook from b1 to c1 and back, etc., the king steadily moves from a3 to b3 to c3 to d2, eventually blocking c1 for the white rook. At every step, if the white rook moves out of the "area", i.e. to b4-8 or c4-8, the knight can block the rook out and you'll get a checkmate.

Back to Top

Post your reply: