Castling

WGM Natalia_Pogonina

Castling is a move performed simultaneously by a king and a rook with the aims of ensuring the king’s safety and connecting the rooks. There are three ways you can castle: long, short and 'artificial.' The latter is when your king and rook eventually get placed the short or long castle way, but it is done manually, move by move. Normally that’s not what you want to do, although sometimes you can’t castle for some reason, but still have to secure the king. In this article I won’t be reminding you about the rules of castling, as this concept is rather basic and should be familiar to you already. (But if you don't know, here's a video explaining it).
Many people castle automatically without giving it any real thought. Not a good idea. Also, castling on the wrong side can be a critical mistake. These two issues will be addressed in more detail.
Castle long or short?
Usually when preparing a certain opening line you learn the typical plans, including a choice of where to castle. However, in real life we occasionally end up in relatively unfamiliar positions when we have to make up our mind ourselves. For example, if the opponent surprises us in the opening. When deciding, consider the following factors:

King’s safety. Will your king be feeling safe on that flank? If your pawn structure is weakened and/or the enemy pieces are all targeting that side of the board, castling is possible only in case of emergency and after careful and precise calculation.
Your plan. When anticipating an active play on one side of the board, you often might want to evacuate your king to the other side. This is especially true of pawn storms: you may not want to place your king behind the pawns you are planning to advance.
Coordination of pieces. When castling, you have to evaluate where your pieces will be placed optimally.
Potential endgame. The king is an active member of the team in the endgame. Therefore, if the queens are off early in the game, it often makes sense to activate the king and keep it in the center as opposed to castling. However, you should remember that an attack on the king is possible even without queens, so sometimes good old castling is still the best choice.

To castle or not to castle?
In the article “Brave kings” I have already mentioned a case when the king can be efficient in the centre. There is one more important technique. When the rook stays on h1(h8), supporting the h-pawn and participating in the attack, the king goes to f1(f8). This is typical of, for instance, one of the lines in the French defense. The king can either remain on f1(f8), or hide on g2(g7) after g3(g6).
Interestingly enough, this latest castling was performed on move 48 in two games: Neshewat – Garrison, Detroit 1994 and Somogyi – Black, New York 2002.
Summarizing, we should be careful when following the book advice “castle early”. Choose wisely whether to castle or not, and which side to prefer.

Needing two wins with Black out of two games, I was grim and focused. Photo from the official website of the EIWCC-2012.
In the game against WGM Jovana Vojinovich from the recent European Championship both of our kings had a chance to castle to any side of the board. They were like two cowboys staring tensely at each other before making a decisive shot. White didn’t want to castle kingside in order to avoid being attacked. I also didn’t feel like castling kingside, being somewhat mistakenly concerned about my opponent’s potential attack. Even White’s weakened pawn structure wasn’t really an argument against castling queenside, or performing it artificially, due to the nature of the position – it was closed. I decided to postpone castling to stay flexible and have both the option of castling queenside and placing the king on f8. Eventually, my king ended up feeling safe on f7, while the White king got slaughtered in the center.

Vojinovic, J. (2356) vs. Pogonina, N. (2449)