When is it Safe?

  • GM thamizhan
  • | Oct 17, 2011

Dear GM thamizhan,

I am busy going over some of my games, and I am mostly confused about the ideas behind pushing pawns on the castled king-side. 

I think in this first position the idea is a fairly simple tactic.  By playing h5 black is attacking white’s knight and the horse is unable to move without getting captured.  His own pawns are blocking his escape route, and being a minor piece down, somethingthing white can think about is to take advantage of black’s disrupted pawn structure to squeeze out a win.  Black on the other hand might be looking at a mating attack on the h2 square perhaps?



In the second position, I find the idea not so simple to understand.  The computer analysis suggested that I play h4, a move which I would never think of playing.  From what I can see, neither black's knight or queen is able to capture the pawn due to the rook on h1 and the knight on f3.  But what long-term advantage does a move such as this one give me?  Black has not yet castled, and there is no guarantee that he will on the king-side (or even castle at all).



I see a lot of master games where moves such as the above one is played.  My problem is that I am quite fearful of opening up pawns in front of my king once I have castled.  I am under the impression that those pawns should stay as close to the king as possible, protecting him from any threats that my occur.  I missed the obvious tactic in my first game because of this.  I overlooked the move based on my fears of pushing too many (if any) pawns close to my castled king.  Obviously my thinking process is faulty, and I need to rectify it Yell

Any advice (especially with examples) would prove most helpful.

- theunsjb


Dear theunsjb,

Excellent question, it is always nice to answer such specific questions. Before you read anything further here I would also suggest you check this previous article which is about attacking on the queen side. It should give you an idea about pushing pawns in front of the king.


As we have always said, if you understand a rule well enough, you will know exactly when you can break it. It is like asking someone to carry an umbrella to work during the rainy season. However, we all know that even during the rainy season, on a cloudless, bright and sunny day, you really do not need to carry that umbrella. In this case, every chess player in the beginning is taught to keep his king safe by keeping all the pawns in front of the king intact. However, it is important to understand there can arise a situation where that safety is not necessary and the pawns in front of your king might be put to better use when they move forward. In such situations, it would be alright to advance the pawns and leave your king a little exposed.

Take a look at the second position from your game for example. It is quite normal for white to consider the move h4 here. Here are some reasons why such a pawn push can be handy:

  • Exploit the position of the black knight on g6 (the h4 idea would not be as effective against a knight on f6 because that would weaken the g4 square)

  • White has a good amount of space advantage, and that implies that he is the one dictating the terms of this game. Black just cannot choose to launch an attack here because he will lack the piece coordination to make his attack successful with his cramped space.

  • White has absolute central control. When you control the center, you are in the driver's seat.

  • White has not castled yet, which also throws in an additional option for him to castle queen side if necessary.

And most importantly, white is trying to expand on his space by using such a move. If the knight on g6 has to run back, that will only make it more difficult for black to find squares to develop his other pieces. If you look at the queen side, the black bishop and the knight share only one developing square, that is d7. Hence, black will have to develop the knight first, then get it out of d7 and then develop the bishop to d7. Trying a fianchetto for the bishop with b6 and Bb7 is also possible, but it will weaken the c6 pawn. If black tries to castle queen side, it will take him a longer time and with white's space and central control, white will be able to break through the center and launch an attack on the open black king. To sum it up, black will most likely castle king side.

Having given all the pointers on why h4 can be a good move, you should also know it is perfectly normal to continue with any normal developing move such as Be3 or O-O instead of h4. Just because a computer is suggesting the move, it does not make it the best possible move. The computers are smart, but not smart enough to solve chess. Openings and early middle games in particular can be tricky for one to understand when to listen to a computer and when to shun it. In this case it is very much possible for white to just castle, centralize his rooks and plan a break through in the center.

In your first example, even though it looks like black is winning a piece straight away using the trap, there are some critical variations to consider before you decide to play h5.


I am sure you understand h5 is not an easy decision any more. The absence of the dark-squared bishop and the weak pawns on e6 and c6 are a problem for black to deal with. Given the situation, the complications arising out of h5 ought to be considered.

You had mentioned in your question about black attacking on h2 after winning the piece and opening up the 'h' file. Even if black manages to accomplish that, he will never be able to launch an attack on the white king as his own king is too weak. If someone is attacking, it will be white.

Here is an example similar to your second game when such a pawn move is considered good.


Without the h5 move white would have had a free hand in playing his central breaks such as e5 or f5, but now he cannot afford to make any hasty decisions. The black bishop along the long diagonal has become a dangerous threat just in one move. Not only did this move achieve a considerable amount of counter threats, it is also curtailing white's own attacking ideas.

If you think your king has an alternate safe position or that there is no real danger for your king, then it is acceptable to advance the pawns in front of your king. But even if you feel safe about your king, do not push them without a proper reason. A proper reason could be anything from chasing a knight from a key square or launching a mating attack; you will have to consider the possibility when you feel the necessity. One key point to keep in mind is not to allow any weakness in our camp when we try to push such pawns. If you have a weak square, it is a head ache to handle elsewhere in the board, if you have that in front of your king, it is going to be that much harder.


  • 5 years ago


    @ EngNgee On 16...Nc3 17.bc and then if 17...hg 18.ef when the pinned black pawn on e6 becomes a target for the white pieces. The King has a hard time stepping out of the pin because the white Queen would take on g4 and then use the h-file to kick the king back. So in that case black would try to hold on with something like 18...Bd7 19.Qg4 Rfe8 20.Rfe1 (white is already overpowering the e6 square) If black keeps trying to hold with 20...Re7 21.Re2 Rae8 22.Rae1 Kf7 Bringing a final defender to bear so now attacking the e6 pawn we have three white minor pieces, a white bishop and pawn. Black has defending the e6 pawn Three black minor pieces, a black bishop, and his king. So it's five to five an equal number of attackers to defenders which means that white needs to kick a black defender and thus he plays 23.Qg6+! Kf8 24.ef Be6 25.Be6 Re6 26.Qe8+!! Re8 27.Re8+ Kf7 When it is unclear if the white rooks could coordinate to effectively take on the Queen, so in conclusion perhaps 16...Nc3 was better unless I am missing something that could have overpowered the pinned e-pawn.

  • 5 years ago


    Please help me on this... In Game 1, why not 16. Nxc3 for Black?

  • 5 years ago


    Really a good one.

  • 5 years ago


    great article!

  • 5 years ago


    Very good article, thank you.

  • 5 years ago


    Thank you for the great article.

  • 5 years ago


    Nice article thanks.

  • 5 years ago


    often times you have to see if your opponents pieces are coordinated to make an attack on your king- it may appear at times that your king is not safe when it actually is, due to the lack of coordination in your opponents pieces.

  • 5 years ago


    great article :)

  • 5 years ago


    great article... much appreciated

  • 5 years ago


    That's a great piece, thank you. I also dither a lot over this issue and have no doubt missed many opportunities because of it.

    I generally tend to leave my King's Pawns in place unless I have reason to fear an attacking Knight or Bishop on the b or g file, or unless I'm threatened with a back rank mate. But then my Pawn play is pretty shoddy anyway and I have a great deal to learn.

    Your response to the first example (Black Pawn to h5 attacking White knight) is absolutely fascinating, I find it very difficult to look ahead in games; the number of variables usually just overwhelms me.

    Thanks for some very helpful pointers.

  • 5 years ago


    This topic has always been a difficulty for me.  Thank you.

  • 5 years ago


    Nice article, it helps me improve my game!

Back to Top

Post your reply: