Overloading and removing the defender

A game on lichess.org, 2014. Black to move, and Black is in check: the white queen on e4 hits the king on h7 along the diagonal. The black queen on c6 is also looking at the queen on e4, along a different diagonal, and the black rook on d8 looks down the file at the white rook on d1. Both white pieces are defended, and the same rook on e1 defends them both: the queen along the file, the rook on d1 along the rank.

One rook holds two pieces. As long as nothing happens, it copes. But if it has to capture something, the other piece is left undefended. Find how to answer the check so as to use this.

The defender did not leave of its own will here; nobody lured it with a sacrifice or checked it. It simply had to do two things at once, and in chess you get one move. That is overloading. But first let us look at an even simpler technique: you do not have to overload a defender, you can just take it.

Getting rid of the defender

In chapter 8 you learned to count attackers and defenders: if a piece is attacked twice and defended once, you can win it. If there are as many defenders as attackers, the piece holds. The techniques in this chapter change that count not on the square where the target stands, but somewhere else on the board. You do not add attackers; you take away defenders.

Removing the defender means simply taking the defender. Sometimes you give a piece of equal value for it, sometimes you sacrifice something more valuable: the target is worth it. What matters is that after the capture, whatever it was defending can no longer be saved.

Overloading means one piece has two jobs: it defends two squares, two pieces, or a piece and a square. As long as nobody presses on it, it copes. Make it carry out one duty, a capture or a block, and the other is left unattended.

The difference from deflection is subtle, and in real games the techniques often merge. With deflection the defender is lured away: it leaves, and everything it was holding falls. With overloading the defender leaves because it is doing its own job, just one of its two. And when you remove the defender, it does not leave at all; it is simply gone. The names are not there for arguments about classification, but to make you ask the right question when you look at a position: who defends, how many jobs do they have, and can they be removed.

Why are pieces left with only one defender in the first place? Because there are few pieces and many jobs. Every piece in a game defends something, and players economise: if the queen already defends a rook, why put a second piece next to it? While the opponent threatens nothing, the economy goes unnoticed. But every time you see an enemy piece with a single defender, you have a target, even if you cannot take it right now.

Removing the defender

The most common case looks like this. You attack an enemy piece, and exactly one piece defends it. If you take that defender, especially with check, the opponent has no time to save what has been left undefended. The check matters here: it stops the opponent from answering your threat with a threat of their own or moving the attacked piece away.

Bishop for knight is an even trade, and the whole gain is that afterwards the queen on d4 is undefended. It is almost always like this: taking the defender wins nothing by itself, and the win comes on the next move. So when you look at such an exchange, calculate one move further.

There are three ways to get rid of a defender, and you already know all three separately. Take it, as the bishop took the knight on c6. Trade it off: offer an exchange the opponent cannot refuse. Or chase it away: attack the defender with a cheaper piece, most often a pawn, so that it has to move. The last way is close to deflection from the previous chapter, and there is no point arguing about what to call it. It matters more to remember the general order of thought: first the target, then its defender, then a move against the defender.

You cannot always take the defender with check. Then you trade it off. Here an exchange works as a forced move: if you offer a queen trade, the opponent has to take, or they will be left without a queen. And if that queen was the only defender of something else, after the trade that something is left hanging.

Squares are defended as well as pieces. If the opponent's mating square is covered by one piece, remove it and the mate gets through. For that it is worth giving a rook or even the queen: mate pays any price.

Notice how the calculation works here. Black gave up a rook, and after Qxe1 White is a rook up. Had there been no mate, the sacrifice would have been a mistake. So before you remove a defender with a sacrifice, check that the blow after it really lands: mate, winning the queen, or at least more than you gave. The calculation is short, only two or three moves, but miss one of the opponent's replies and you are a piece down.

Most often this is how the defender of the back rank is removed. In chapter 5 you saw the back-rank mate: the king is shut in by its own pawns, and a rook or queen mates it on its first rank. The opponent knows this and keeps a rook on the back rank. If that rook can be taken, the mate comes back.

The rook on c1 was not overloaded here: it had one job, and it was simply taken. The price did not matter either: for the queen, Black got mate.

Overloading

You look for overloading differently. Here you do not need to take the defender: you need to find a piece with two jobs and press on one of them. You saw the clearest case at the start of the chapter: a rook defends two pieces, one along the file, the other along the rank. Trade off one of those pieces, the rook takes back, and the other is left bare.

Overloading is easy to confuse with a double attack, but you have to look at it from the other side. A double attack is your piece hitting two targets at once. Overloading is an enemy piece defending two targets at once. You can attack them with different pieces, as the queen and the rook did in the first puzzle of the chapter.

The same pattern turns up all the time, only the pieces change places. Look for it like this: find two enemy pieces under attack that are defended by the same piece. Then start with the capture that forces a reply, most often a queen trade or a check. That leaves the defender no time to choose.

The pieces most often overloaded are the queen and the rooks. They are long-range, and it is convenient to defend several pieces from afar with them. A knight or a bishop can be overloaded too, but less often: their reach is shorter, and two jobs pile up for them only when the opponent's pieces stand close together. If the opponent has several weakly defended pieces and one long-range piece defends them all, take a closer look: there is most likely a blow.

The king can be overloaded too. It defends the squares next to it and the pieces beside it, and in the endgame or after the queens come off the whole defence often rests on it. But the king, like any piece, moves once. If it has to take one thing, the other thing it was defending stays behind.

Overloading hits hardest when one of the defender's duties is to stop mate. A rook that stands on the back rank and also defends a pawn in front of the king is the classic example. Such a rook must not capture: if it does, the back rank opens.

Three ideas work at once in this puzzle. The rook on f8 is overloaded: it defends f7 and the back rank. The queen sacrifice decoys the rook to f7. And the bishop on c4 pins it there, so the rook cannot go back. That is what combinations usually look like: techniques do not come one at a time, they add up to a single blow.

A count with a catch

Overloading hides in the count of attackers and defenders. When you check an enemy piece, you count its defenders, and if there are enough, you move on. The mistake is counting the same defender for two pieces. In the puzzle above the rook on d8 honestly defended both the queen and the rook on e8. Taken one by one, both pieces were defended. Together they were not, because you cannot defend two squares with one move.

Count that position by the rules of chapter 8. The queen on d5: the white queen attacks it, the rook on d8 defends it. One against one, the queen holds. The rook on e8: the rook on e4 attacks it as soon as it is unpinned, and the same rook on d8 defends it. One against one again. Both counts add up, and Black seems to be fine. Now cross the rook on d8 off one of the lists, because it can only recapture once. At once you see that one of Black's pieces falls.

So it helps to ask the question differently. Not "is this piece defended?" but "what else is its defender busy with?" Go through the enemy pieces that hold the defence, and for each one list what it defends. If there are two items on the list, it is a candidate for overloading. If there is one, but that defender can be taken, traded off or chased away, it is a candidate for removing the defender. If there are two defenders, the techniques of this chapter usually do not work, and you have to look for the blow elsewhere.

Ask the same question about your own pieces. Overloading is not only a blow but a weakness. If your queen defends a rook and the rook holds the back rank, you are standing on one leg, and the opponent only needs to find the right exchange. A simple rule helps here: an important square or piece should have two defenders if one of them has other work. Better still, remove the cause itself: put the rook under protection, give the king luft, move a piece out of the line of fire in advance.

Finally, remember the move order. In the puzzle with the rook on d4, the trade …Qxf4+ won, while …Rxd6 lost, although both moves took something the overloaded rook was holding. The difference is that the queen trade with check forced White to reply, while taking the rook gave White time to take the queen on e5. Start with the move that leaves the opponent the fewest replies: a check, a queen capture, a threat of mate. Start with the wrong one, and the opponent will make their own strong move.

What's next

In every puzzle of this chapter the opponent's move was forced: they had to take, or lose more. But a forced move is not always as forced as it looks. Sometimes, instead of taking, you can first play another move, with check or a threat, and only then take. Such a move is called an in-between move. It wrecks calculations, and a missed in-between move is one of the main causes of blunders. That is the subject of the next chapter.