The Bayonet Attack and 9...a5: When the King's Indian Chooses Silence

The black pawn comes to rest on a5.

No sound of impact, no clock tick. In a small room on Nguyen Dinh Chieu Street in Saigon, where I have sat every Saturday evening for five years, that move usually passes in silence. People wait for 9...Nh5 — the knight's leap to the rim of the board, the move every database knows, every engine knows, and every player who has ever handled the black pieces in the King's Indian Defence has memorised since childhood. 9...a5 rarely gets mentioned. It is not glamorous. It does not produce games people retell on the way home. It simply pushes a pawn sideways, then sits and waits.

And yet, recently, while clearing out an old shelf of material, I came across a training DVD devoted entirely to that move. The cover promised no revolution. It said one simple thing: if you are tired of lines dissected to the last percentage point by computers, try 9...a5.

From that small detail, the story opens.

Context: a system that lives on patience

The King's Indian Defence was born from a counter-intuitive philosophy: Black deliberately concedes the centre, builds a bishop wall on g7, then slowly tightens with counterattacks on both wings. The familiar sequence runs 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5. White takes space, Black takes stance. That trade defines the whole character of the system: whoever understands the price of space wins.

The Mar del Plata Variation — named after the Argentine coastal city where it made its name in the 1950s — pushes the system to maximum tension. Both sides advance pawns to the fifth rank and the centre locks: 7.0-0 Nc6 8.d5 Ne7. White masses on the queenside, Black masses on the kingside, and the race to break through first becomes the soul of the whole variation. The names attached to it stretch across generations: David Bronstein, Efim Geller, Svetozar Gligoric, Miguel Najdorf, Bobby Fischer, and later Garry Kasparov, who turned the King's Indian into a personal weapon for nearly two decades at the top. The ECO codes for this branch sit in the E97–E99 range.

In the 1990s a new direction appeared that gave loyalists a headache: the Bayonet Attack, 9.b4. Instead of slow central play, White immediately pushes the b-pawn, grabbing queenside space, threatening b5, and clearing a path for the bishop to a3. The bayonet thrusts straight into Black's structure, forcing an immediate reaction instead of leisurely development.

The orthodox reply, accepted worldwide, is 9...Nh5. The knight jumps to the rim, prepares ...f5, and above all opens the position. After 9...Nh5 everything becomes concrete. White must calculate precisely, Black must remember lines, and the engine becomes the supreme arbiter again. That is why 9...Nh5 exists: it gives nobody time to think. In a four-hour classical game, being pushed into forced sequences from move twelve means spending your clock on what someone else has already worked out.

The other road: 9...a5

The DVD I mentioned at the start takes the opposite route. It proposes 9...a5 — a discreet move, long regarded as a fallback, occasionally appearing in the games of players who like surprises. Instead of bringing the knight to h5, Black pushes the a-pawn, attacking the b4-pawn White has just advanced, forcing an immediate decision: hold the pawn, trade it, or push on to b5.

The product's selling point lies in how it describes the character of the game. According to the promotional text, 9...Nh5 opens the position, forces the lines out, and makes computer analysis important again — whereas 9...a5 leads to sound positions with very few concrete lines. The product claims the chances are equal, and targets players who like to think for themselves rather than memorise.

This is a positional argument, not a claim of theoretical superiority. Black is not trying to prove that 9...a5 is stronger than 9...Nh5. The claim is simply that it is easier to handle for a particular group of players at a particular level. That framing is far more honest than the usual promise of a move that destroys everything.

The central question remains: is 9...a5 really as low-theory as advertised?

I reopened the database and spent several evenings reconstructing the positions on a board. In elite games, 9...a5 appears far less often than 9...Nh5. It is a known branch, not a new invention. That does not devalue it — plenty of effective weapons in chess existed for decades before anyone wrote a book about them. But it leaves an uncomfortable thought: if 9...a5 is so easy to play, why has it never become the main move?

The answer lies elsewhere, not in the quality of the move.

After 9...a5, White has several ways to proceed. He can keep the structure, push b5 to open lines, or accept a pawn trade and shift to a different type of position. What I found while reconstructing: Black does escape long forced sequences, but in exchange he enters a long-haul struggle where White's space advantage remains intact and every mistake is punished slowly but surely. Fewer lines to remember. But not one move fewer to understand.

That is the point advertising usually skips. Fewer concrete lines does not mean less work. It means the work shifts from memory to positional judgement. For a player with good strategic foundations, that is a bargain. For a player used to leaning on memorised lines, it can be a trap.

In amateur chess, people often confuse two things: a theoretical antidote and a practical antidote. A theoretical antidote is a branch that flips the engine's evaluation. A practical antidote only needs to leave the opponent disoriented for the first thirty minutes. 9...a5 belongs to the second category. It does not overturn the assessment. It only changes where the pressure sits — from Black's shoulders to White's.

A dated premise about computers

The argument that computers often cannot handle complex strategic positions sounds very familiar. It is a legacy of the pre-AlphaZero era. Since 2026, when engines moved to neural-network architectures, their ability to assess long-term positions changed fundamentally. Closed positions of the King's Indian type, once a blind spot for older engines, are now handled with high accuracy.

This does not destroy the product's value. It simply places it correctly: a human tool, packaged with a dated argument. Buyers can still learn a great deal from it. But anyone buying in the belief that they hold a territory machines cannot touch will be disappointed. In an era where any position can be evaluated by a free program in seconds, the value of a DVD lies in how it tells its story, not in possessing a truth machines missed.

A signal about source quality

One detail made me stop mid-way.

The document I came across carried a headline about a team event in Samarkand — a round in which all the top teams won — but the entire body of the text discussed the opening DVD I had just been analysing. The two pieces did not fit together.

Samarkand is a genuinely chess-active city in Uzbekistan, where team and open events are regularly held. A round report from there makes perfect sense. But it has nothing to do with 9...a5. The mismatch is a signal about source quality: someone spliced the opening of one article onto the body of another during aggregation. For readers, the consequence is concrete: every claim in the original document needs verification before citation.

What is telling is that the document names no author, includes no sample games, offers no engine evaluation, and provides no win-rate or accuracy figures. For an opening product, those are the minimum requirements for a buyer to make a decision.

A secondary branch and an absolute promise

Theoretically, 9...a5 remains a secondary branch. A well-prepared White player can still find play. Statements such as if Black knows what to do, things quickly become dangerous for White are the standard register of opening-product advertising. They are half true. The other half depends entirely on how well your opponent has prepared.

The biggest risk of an opening weapon is not that it is weak. It is that it gets reverse-engineered. In chess, whenever an unusual move rises to prominence, within months someone writes a counter. The Bayonet Attack went through exactly that cycle: it dominated for a while, was decoded, and became an ordinary part of theory. 9...a5 will likely follow the same path, if it ever becomes prominent enough to be noticed.

But there is a subtle difference. Branches that get decoded are usually branches with a direct threat — they force the opponent to find a defence, and in that process a counter appears. 9...a5 threatens nobody. It simply declares that Black is willing to play a slow game. That kind of move is harder to refute, because it offers no target to attack. It only changes the tempo.

The opening-theory market: where belief is sold alongside lines

Opening-theory publishing is one of the most stable revenue streams in professional chess. It is not glamorous like major tournaments. There are no grandstands, no television. It lives on a very human need: everyone wants a small edge before the game begins.

But that market is under unprecedented pressure. Free online databases, powerful engines, and community material make selling a handful of concrete lines difficult. Products can only compete on two things: narrative and curation. 9...a5 is sold not because it is stronger than 9...Nh5, but because it represents a type of player — someone who does not want to become a memorisation machine.

That is a clever pitch. It does not promise victory. It promises an identity.

And identity is hard to replace in a market where every move can be looked up for free. When the value of information hits the floor, the value of emotion becomes the commodity. Buyers are not paying to know 9...a5. They are paying to believe they belong to a group of players who do things differently.

What I see away from the board

I have organised a few small amateur tournaments in Saigon. In that role, one thing struck me: players drawn to closed systems like the King's Indian are usually not chasing quick wins. They are chasing autonomy. They want positions where the result depends on reading the board, not on who remembers more.

That is why 9...a5 has a place, even though it has never been the main move. It is not the tool of someone who wants to win through preparation. It is the tool of someone who wants to win through thinking.

The King's Indian was once the backbone of many world champions' arsenals. Kasparov used it as a statement during his years against Karpov. Fischer chose it at important moments. But its share of elite classical games has declined steadily in the engine era. The change comes from the pressure of precision: when both sides can prepare to move twenty-five, conceding space becomes a hard bet to justify.

So at the top level the system survives mainly as a surprise weapon. It has not disappeared. It has retreated into the moments when a player needs to break an opponent's preparation, needs a game where the opponent cannot arrive with ready-made answers.

Based on my experience following domestic and international chess events, the same pattern plays out in amateur tournaments. Closed systems are no longer chosen because they are strong. They are chosen because they are unfamiliar. And unfamiliarity, in a world where everyone studies the same body of material, has become a scarce resource.

What remains after a small move

In chess, people measure a move's value by its result. But some moves exist for another reason: they open a type of question the opponent is not prepared to answer. 9...a5 belongs to that group. It is not strong. It is not beautiful. It simply creates a situation where memory becomes useless and both sides must return to the most basic skill — reading the board.

An empty grandstand still echoes — you only need enough stillness to hear it. In quiet tournament halls, with no commentators, no spectators, only the clock and the sound of pieces meeting the board, moves like 9...a5 quietly do their work: keeping the memory of a system from being erased by the speed of machines.

The real death of a golden generation comes when we stop remembering them. That is true of people, and equally true of positions. The King's Indian has passed through many seasons of glory and oblivion. It does not need saving. It only needs remembering.

The product in my hand may not be the thing that changes how a generation plays. It may just be a small DVD, made for a very specific group, with a modest promise. But at a moment when everything at the top of chess is compressed into strings of numbers, the fact that someone still sits down to write about a pawn pushed sideways is itself a form of resistance.

Generations will pass, but the poem about them keeps being written. And sometimes that poem begins with a pawn resting on a5, waiting for someone patient enough to understand why it is there.