This is the revenge game, and it contains one of the strangest sights in Capablanca's work: the white king walking from c2 all the way to g4 while the queens are still on the board. It is not a stunt. Capablanca had first removed every source of counterplay, and once an opponent has no threats the king stops being a liability and becomes a third attacker. Take one idea into your games, an attack usually needs three pieces, and if you are short of one, look at the piece you have been protecting all game.
Roughly, one attacker removes or distracts a defender and two more deliver the blow. With only two you generate threats; with three you generate mate.
The king only marches after every black counter-chance has been shut down. Take away their play, then take liberties with your own position.
With queens on the board the king normally hides. When the opponent has no threats at all, it becomes the extra piece the attack was missing.
Watch the lesson, then play the game through on the board below. The board stays with you as you scroll. Make each move yourself while you read. You'll remember a plan you played with your own hands far longer than one you only watched.
Capablanca had lost to Chajes two years earlier with his own king caught in the open, and this game is the answer. He gets a small edge from the Rubinstein French, wins a pawn, and then does something most players never consider. Instead of hunting for a combination with the pieces he has, he spends a dozen moves making sure Black cannot do anything at all: the rooks are traded down, the queen is tied to defence, the pawn breaks are covered. Only then does the king begin its walk, Kc2 and Kd3, then Ke2, Kf1, Kg2, Kg3 and finally Kg4, arriving as a full attacking unit while the queens are still on. That is the practical meaning of needing three attackers. White's queen and rook were two; the king made three, and once it arrived the h5 break simply crashed through. In your games, do the prevention first and count your attackers honestly, and remember the king is a piece.
Chajes chose 4.Bd3 rather than the main lines, looking for an open game, and the play transposed into a Rubinstein French. Black's difficulty was ordinary and familiar, he never challenged the centre with a timely c5, which let White build a stable platform. There was also a tactical shot lurking, the natural-looking b6 would have run into Bxf6 and Qe4 with a double attack. Capablanca preferred the simple 10.Bxf6, and by the time the queens had settled he had the more comfortable position and, shortly after, an extra pawn.
White is a pawn to the good, the rooks have been reduced, and Black is doing nothing but waiting with the king shuffling between f8 and g8. There is no obvious break and no combination. This is where most players shuffle too. Stop and assess before you read on.
White is a pawn up with no clear break. Which piece can still be improved, and why is it safe to do it now?
Count White's attackers, then count Black's threats. If the answer to the second is zero, which piece has been sitting out the game? Think, then reveal.
The method is unusual and completely logical. First Capablanca prevents, he reduces the rooks, ties Black's queen to defensive duty, covers every square that might give counterplay, and only then treats his own king as available. Second he marches, the king goes from c2 to d3, then across to e2, f1, g2, g3 and g4, one careful step at a time, and at no point is there a check to disturb it, because Black has nothing to check with. Third he breaks, 39.h5 opens the king-side at the moment three white attackers are in place, and after 40...Qe7 41.Qf5 the position collapses, 42.Rd7 and 43.Kg4 finishing the job. It is worth stressing that the order cannot be changed. March the king first and it is a blunder; prevent first and it is technique. In your games, the sequence is prevention, then improvement, then the break.
The king has completed its journey to g3 and every white piece is on its best square. Black is still shuffling. The position is ready for the blow that a third attacker makes possible.
White to play. Three attackers are in place against the black king. Which pawn break opens the position?
39.h5, the break that the king march earned. After 39...gxh5 40.gxh5 the file is open and White has queen, rook and king all bearing on the black king, which is exactly the three attackers the position needed. Played twenty moves earlier the same break achieves nothing, because two attackers only produce threats that can be parried. This is the practical content of the rule: count how many of your pieces genuinely take part in the attack before you open lines, and if you are one short, find the piece that has been idle. Here it was the king, and it could only join because Chajes had been left with no counterplay whatsoever. In your games, prevention is what makes ambition safe.
Once the h-file opens the attack runs by itself. 41.Qf5 and 42.Rd7 tie Black completely, and when 42...Bxe5+ comes it is not a rescue but a last try, because 43.Kg4 simply steps forward. The white king is not fleeing the check, it is advancing with the attack, which tells you everything about who has the initiative.
The end comes at once, 44.Nxe5 and 45.Kf4, and Chajes resigns with his position falling apart. The game is worth revisiting whenever you have a small advantage and cannot find a plan, because the answer here was not a clever idea but a patient one, shut everything down, then bring the last piece. Carry it into your games, three attackers, prevention first, and the king counts.
The rule: one attacker to remove or distract a defender, two to finish. In this game: queen and rook were two, and the king made three. In your games: count real attackers before you open lines.
The rule: shut down the opponent's counterplay before you take any liberty with your own position. In this game: the king only walked once Black had nothing. In your games: Provention first, ambition second.
The rule: a king is a fighting piece whenever the opponent has no threats, regardless of the phase. In this game: it travelled from c2 to g4 unmolested. In your games: if they cannot check you, your king is available.
The rule: when you cannot find a plan, find your worst-placed piece and fix it. In this game: the idle piece happened to be the king. In your games: the plan is often just the next improvement.
The rule: a pawn break played before the attackers arrive achieves nothing. In this game: h5 was decisive on move 39 and would have been harmless on move 25. In your games: prepare completely, then open.
Because a defended king can usually meet two attackers. The rough division is that one piece removes, distracts or deflects a defender while the other two deliver the decisive blow. It is a guideline rather than a law, but it explains why so many promising attacks with two pieces fizzle out.
It would be in almost any position, which is why the prevention comes first. Capablanca spent a dozen moves ensuring Black had no checks, no breaks and no active squares. Once your opponent genuinely has nothing, the king is simply another piece, and here it was the piece the attack was missing.
It is the habit of asking what your opponent would like to do and taking it away before you proceed. In this game it meant trading the rooks that could become active, covering the squares Black's queen wanted, and denying every pawn break, all before any attacking move was made.
Because it opened a file at the moment three white pieces were aimed at the black king. Timing is the whole point, the same break with only the queen and rook in place would have been met and neutralised. Attack strength is about how many pieces arrive, not how forcing the move looks.
Created counterplay early, even at a cost. The losing pattern was passivity, waiting moves with the king while White improved freely. Once you allow an opponent to improve without any risk, you also allow ideas like a king march that would normally be impossible.
Six king-side attacking exercises, rising in difficulty, the operation this game was built toward. In each you have the winning side, so bring the attackers and open the king. Drag the pieces, the board plays the defence for you.
Most attacks fail for one reason, they begin with two pieces and hope. In a 2-hour Strategy Assessment I go through your own games, count the attackers in the positions where your initiative stalled, and hand you a plan built for your chess, not a hypothetical student's.
Book Your $130 Strategy Assessment 2 hours via Zoom · Worldwide · Limited spots