Most players treat an isolated queen's pawn as a problem to be solved. In this consultation game in Vienna in March 1914, with young Richard Reti at his side, Capablanca treats it as a platform. The pawn on d5 gives Black the only pawn in the center, a claim on e4, and a job for every piece. Reti was sure of the natural rook move at the key moment. Capablanca played a quiet bishop move Reti had not even considered, and by the story I tell in the video, Reti went away and studied Capablanca for years. The game also has an honest side: Capablanca's team walked into a line that throws away the win, and he published the refutation himself. Then comes a rook pawn thrown forward, a bishop that no pawn can touch, a king that marches to the center, and a breakthrough on the other wing. Take one idea from this game into your games: while the pieces are on, use the isolated pawn, do not nurse it.
With pieces on the board, an isolated queen's pawn is the only pawn in the center. Use the squares it controls and the lines it opens, so your pieces are too busy to defend it.
A pin on the king is absolute: the piece cannot move. A pin on anything else is relative: moving is legal, so check what it costs before you call the piece stuck.
Do not hit the point your opponent is defending best. Open a front elsewhere, stretch the defenders, then bring your king in and break through.
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.
After nine moves Black has the only pawn in the center. Yes, it is isolated, and in an endgame that can be a real cost. With a full board, look at what it buys: control of e4 and c4, open lines for the pieces, and targets on f2 and h2. White's job is to pressure the pawn until Black's pieces are tied to it. Black's job is to keep those pieces free. So when the pin on f6 becomes annoying, Capablanca does not settle for the natural developing move. He plays 14...Bd4, then gives up his bishop pair with 15...Bxc3 and walks straight out of the pin with 16...Ne4, because a pin on the queen is not a pin on the king. The trade he makes on move 19 is the whole doctrine in one line: 19...gxh4 20.fxe4 Bxe4 wrecks his own pawns, but no White pawn is left on the d-file or the f-file to challenge that bishop. Structure is what you pay; activity is what you buy. By the endgame, the bishop on e4 is the best piece on the board, supported by the very pawn everyone told you was a weakness.
This was a consultation game: Faehndrich and Kaufmann against Capablanca and Reti. The book's players block prints only Capablanca's name for Black, so the study record and my endgame lesson fill in Reti. White's 4.exd5 is, in Capablanca's words, "perfectly safe" but gives no advantage, and after 5...c5 Black accepts the isolated pawn to take the initiative. Then 10.Ne2 is the first slip: Capablanca says Qd2 or Re1 should have been played, and that the knight will have to go back to c3. It does, on move 12. Meanwhile 10...h6 and 11...Bg4 free the light-squared bishop, and 12...Nd4 and 13...Nxe2+ use the platform at once. The lesson starts after 14.Qxe2, the moment that so struck young Reti.
White's bishop on h4 pins the knight on f6 to the queen, and the d5 pawn needs watching. This is the moment to use the Goldilocks principle: find three candidates and rank them good, better, best. Reti wanted the natural rook move, 14...Re8, developing with a hit on the queen. There is also 14...g5, which keeps both bishops and breaks the pin by force, at a cost to the king. Compare your own candidates before you read on, and do not ask only which move is strongest. Ask which problem each one solves and which it creates.
Black to play. Reti wants 14...Re8, and the pawn push 14...g5 keeps the bishops. What is your plan for the c5 bishop and the pinned knight?
Three candidates, then good, better, best. Think about what you can damage in White's camp, not only what you can develop. Think, then reveal.
Be honest about 14...Bd4: the engine does not rate it above the others, it holds a level game, and its first choice is the very 14...g5 Capablanca turned down. His note says he thought about this for a long time. White answers 15.Qd3, so after 15...Bxc3 16.Qxc3 the queen recaptures and the pawns stay undoubled. That is the "failing that" half of his note: the pin is about to go. Now 16...Ne4 hits the queen, and here is your pin lesson. Only a pin on the king makes a move illegal. Against the queen you just check the price: 17.Bxd8 Nxc3, and if 18.bxc3 Rfxd8 White has doubled pawns in a level game, while 18.Bh4 Ne2+ 19.Kh1 Bxf3 20.gxf3 is worse still for him. After 17.Qd4 Capablanca plays 17...g5 after all, and I would have been happy with Qd6, but with fewer pieces on he judges the risk differently. Then 18.Ne5 Bf5 19.f3 gxh4 20.fxe4 Bxe4 leaves Black a pawn up with four isolated pawns and a monster bishop. His own long note on 19.Bg3 is a warning too: at the third move of his line, 20...Bxc2 allows 21.Rxf7! Rxf7 22.Nxf7 Kxf7 23.Qf2+, picking up the loose bishop in a level game. In your games, most calculation errors live in the first three moves, not the seventh.
White has just played 21.Rf2. Black is a pawn up on paper, but the extra pawn is the doubled h-pawn, and the black king is the weak spot. It will stay that way unless the queens come off or White's king becomes just as loose. Capablanca later wrote that this next move is why the game was won. It may look strange, but it does one job and does it fully.
Black to play. Which pawn move opens White's king and turns the bishop on e4 into an all-powerful piece?
21...h3. The doubled rook pawn has no other future, so it goes forward to break up White's king. Capablanca: "it is through the great force of this move that the game was finally won." Be careful how you read that. The move is the engine's top choice, but the position is still close to level, and it does not win by force: 22.g3 would have held. What it does is set a practical problem, because 22.gxh3 Kh7 opens the g-file for a black rook and swings the game heavily Black's way. White played 22.Re1 f5 and then went wrong with 23.gxh3, where 23.c4 or 23.g3 kept the fight. Now comes the sting. The right move was 23...Kh7, as the book itself says, getting the rook to the g-file, and 23...Qg5+ was just as strong. Capablanca chose 23...Qf6, and it should have cost him the win: 24.Rxe4! dxe4 25.Rg2+ Kh8 26.Rg6 Rg8 27.Nf7+ Kh7 28.Qxf6 Rxg6+ 29.Ng5+ Rxg5+ 30.Kf2. White has the queen against two rooks, and the engine calls it dead level. Capablanca published that line against himself: "Black cannot win at all." The allies played 24.Nf3, the queens came off, and Black still stood clearly better. That is the standard to take to your games: write down your own errors, and check the forcing replies before you relax.
The endgame is two rooks and a bishop against two rooks and a knight, and the bishop has targets everywhere. Going after White's king directly would meet too many defenders, so Capablanca creates weaknesses elsewhere first: 26...Rb6 provokes 27.b3, leaving the queenside pawns on the bishop's color, and 27...Rc8 keeps a white rook off c3. Then the king starts walking with 29...Kg6. The allies should have stopped it with the check 30.Rg3+, the engine's clear first choice, and they chose 30.c3. Worse followed: 32.h4+ Kf6 was the real slip, and the engine's edge for Black jumps from about one pawn to about five, because the check only helps the king toward e5. After 33.a4 comes 33...b5, a pawn offered to open lines for the rooks, a small minority attack. Honest note: the engine prefers the immediate 33...Ke5, and both moves win. In your games, once you have stretched the defenders, put the king in and then open the door.
After 34.axb5 Ra1+ 35.Rf1 Rxf1+ 36.Kxf1 Ke5, Black is a pawn down, and it does not matter: his king is a monster, while White's rook and king have little to do. 37.Nd4 f4 is the one winning move, with nothing else close, and it drives the rook to the sad square h3. Then comes 42...h5, which is very nearly zugzwang: the rook has no safe square, a knight move lets the black king in, and 43.Kf1 loses the rook to Bg2+, so 43.Kd1 or a pawn move like 43.b4 is all that holds things together. After 43.Kd1 Bf5 44.Nxf5 Kxf5 the blockader is gone, and 47...d4 leaves a forced mate on the board, so the allies resigned. The isolated pawn that started it all has become a passed pawn marching through the center. Capablanca's verdict on this ending: "one of my very best." You now know why. He played the whole game toward it.
The principle: with pieces on the board, an isolated queen's pawn gives central squares and open lines. In this game: the pawn on d5 backed the bishop and knight on e4. In your games: ask what the pawn lets your pieces do before you worry about defending it.
The principle: a pin on the king is absolute; every other pin is relative, so moving is legal. In this game: 16...Ne4 stepped out of the pin because taking the queen cost White his own. In your games: name the pin first, then price the move.
The principle: find three candidates and rank them before you commit. In this game: Reti saw one obvious move; Capablanca thought for a long time and chose the plan that kept his king safe. In your games: do not play the first natural move you see.
The principle: calculation errors usually come early in a line, when you miss a forcing reply. In this game: 24.Rxe4 would have saved White, and Capablanca missed it at the board. In your games: look for your opponent's checks and captures before you go deep.
The principle: one weakness can be defended, so open a second front and bring the king. In this game: queenside pressure came first, then ...Ke5 and ...f4. In your games: in the endgame, a centralized king can choose which side it attacks.
It is a pawn on the d-file with no friendly pawn on the c-file or e-file, so no pawn can ever defend it. That is a long-term cost, especially in the endgame. But with many pieces on the board, it is also the only pawn in the center, which gives you control of key squares like e4 and c4 and open lines for your pieces. In this game Black's d5 pawn was never the story; the pieces it supported were.
An absolute pin is against the king, so the pinned piece cannot legally move off the line. A relative pin is against anything else, usually the queen or a rook. The piece may move, and the only question is whether it is worth it. Black's knight on f6 was relatively pinned, and 16...Ne4 was possible because after 17.Bxd8 Nxc3 Black wins the queen back, damaging White's pawns or winning more if White goes wrong.
Not by force, and that is worth knowing. Capablanca credits it with winning the game, and the engine does make it the top choice, but the position stays close to level if White answers 22.g3. Its strength is practical: the doubled pawn is useless at home, on h3 it breaks up White's king, and the natural capture 22.gxh3 is a serious mistake because 22...Kh7 brings a rook to the g-file.
Because the result does not grade the move. After 24.Rxe4! dxe4 25.Rg2+ Kh8 26.Rg6, White reaches a queen against two rooks with the black king exposed, and the engine confirms it is dead level. The allies played 24.Nf3 instead, so Black got away with it, yet Capablanca still published the line against himself. Grade your own wins the same way: what was the best reply, and did your opponent find it?
Calling a weakness a weakness without asking what it does for you right now. Before you spend moves defending an isolated pawn, ask what it controls, which of your pieces it helps, and how long the pieces will stay on. Then play toward the ending you want, the way Capablanca did. His pawn was isolated from the opening and he was still pushing a d-pawn up the board when the game ended.
Six pin puzzles, because this video turns on knowing what a pin really is. A pin, absolute or relative, is part of each solution, and the difficulty rises as you go. Find the move, then play it out. Drag the pieces; the board answers for the defense.
Whether you treat a weakness as a burden or a platform says a lot about how you play. In a 2-hour Strategy Assessment I go through your games personally and build you a roadmap for your chess, not a textbook student's.
Book Your $130 Strategy Assessment 2 hours via Zoom · Worldwide · Limited spots