9...a5 and the Bayonet: The Quiet Promise of Chess Opening Theory

The ninth move for Black. A pawn leaves a7, stops on a5, and the board goes quiet for a beat. No piece is captured. No fire flares in the centre. But the position has just turned into another corridor, where the two sides will walk together in the dark for the next twenty moves, neither one sure whether it is closing in on the target or drifting away from it. I sat with that move for a long while, in an apartment in Binh Duong, over a coffee brewed before dawn. For someone who has spent most of his career writing about chess, moves like 9...a5 are the hardest to write about: they do not create moments, they create a silence with weight. What brought me to this move was a long promotional text for an opening-training product — the kind of instructional disc that European chess-software publishers have been selling steadily for more than two decades. That product promised something that sounds very soothing: an antidote to the Bayonet Attack, the 9.b4 branch that has been keeping a good many Black players awake at night in the King's Indian Defence. And that antidote, according to the blurb, comes down to a single move. What made me stop was not the move. It was the way it was being sold. THE LAND OF THOSE WHO MOVE SLOWLY The King's Indian Defence, beginning 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7, is one of the boldest philosophical statements chess has ever produced. Black yields the centre to White, retreats behind his own wall of pieces, then waits for that centre to swell past its own size and crash back into itself. A hypermodern spirit, executed with an almost provocative patience. The system was not born in a laboratory. It was raised by the Eastern European players of the 1940s and 1950s, with Isaac Boleslavsky and David Bronstein laying the first bricks of what later became known as the closed Indian structures. By the middle of the century, the King's Indian had become a declarative choice: whoever plays it is not chasing an advantage on move ten, but a long battle. The Mar del Plata variation — named after the tournament held in Mar del Plata, Argentina, in 2026, where the system was tested before an international public — is the fiercest version of that whole family. There, Black does not merely concede the centre. Black prepares to slam it shut: knight to c6, pawn to e5, four pawns forming a wall across the board. White drives a pawn to d5, seals the centre and turns to the queenside. Black pushes ...f5 and builds a storm on the kingside. The two sides take different roads to the same destination, and that destination is mate. The system sits in the ECO range E97 to E99. For decades, the Mar del Plata was the private weapon of players who carried a certain wildness in them. Garry Kasparov was its greatest embodiment. He turned ...f5 into a declaration of war, and in some games he turned it into his own gravestone. Vladimir Kramnik, on the other side, helped turn this branch into a test of nerve endurance rather than a test of memory. Then, in the late 1990s, White found the needle. The ninth move b4 — the Bayonet Attack — jabs a pawn to b4, menacing Black's queenside structure before Black can close the position. The purpose of the blow is clear: drag the game into a branch White understands better, forcing Black to answer with a highly forcing sequence. The orthodox reply is 9...Nh5, bringing the knight to the rim, preparing ...f5, opening the game into a storm in which anyone who has forgotten their lines will fall. The Bayonet Attack thus became the weapon of people who hate being pushed. It also became the permanent headache of anyone loyal to the Black side of the King's Indian. The board does not hold only pieces; it also holds the calluses of a human life. Any opening branch that survives long enough carries the fingerprints of hundreds of people who tried it and failed with it. THE PROMISE OF THE NINTH MOVE That promotional text did not present 9...a5 as an invention. That is worth noting. It admitted this is a known branch, a secondary option, and placed its value elsewhere. What is being sold is not novelty, but strategy. The central argument rests on a simple contrast. 9...Nh5 opens the position, forces the lines, and makes computer analysis important again. 9...a5, by contrast, leads to sound positions with very few concrete lines. One side is a contest of memory. The other is a contest of judgement. To picture the position that ...a5 creates, look at the pawn structure. White has a pawn chain on the queenside anchored at b4. When Black pushes ...a5, that chain is threatened from the flank. White must choose: trade and open the a-file, or keep the tension and accept that the a-file will become a secondary front for many moves. Either way, Black gains the most valuable thing at this stage: a free tempo to arrange pieces according to his own plan rather than chasing White's. That is the entire value of ...a5 — not seizing an advantage, but seizing the choice of tempo. Read slowly, and the claim has three layers. The technical layer: ...a5 attacks the b4 pawn head-on, forcing White to decide whether to hold or trade, and in many cases steering the game into a closed structure where Black can develop at his own rhythm. The psychological layer: when the position no longer contains forced lines, the Black player no longer has to remember someone else's homework, but must think with his own head. The third layer is the sales layer, and that is the one worth discussing. The text targets a very specific buyer profile: the player who prefers thinking to memorising. This is identity positioning, not technical positioning. It does not say “this branch is better.” It says “you are the kind of person who plays this branch.” In an opening market that is already saturated — where almost every branch can be looked up free in online databases — the only thing left to sell is a feeling about the person you want to become. I have watched this market operate for more than twenty years. ChessBase, the German chess-software publisher, brought its Fritztrainer-style opening training line to market in the early 2000s. Years later, online learning platforms carried the model onto the web. These products survive not because of the lines — lines are available to everyone, and free databases have plenty — but because of the guiding structure, the narrative voice, the act of turning a dry thicket of variations into a journey with someone holding the lamp. One detail stands out. The entire argument rests on a human-versus-machine premise: positions computers handle poorly will be the human's habitat. It sounds seductive. But that premise has an age. THE BLIND SPOT OF A PROMISE The claim that computers often do not know how to handle complex strategic positions is the language of the era before 2026. In December 2026, AlphaZero — DeepMind's system that learned by playing itself — defeated Stockfish in a widely publicised exhibition match, and the way it handled closed positions and long-term pawn structures changed the general understanding of how deeply machines understand chess. The neural-network engine generations that followed handle Mar del Plata-style positions far better than the promotional text implies. The premise that “machines do not understand” is no longer a technical advantage. It is only a marketing tactic. The second blind spot lies in the promise itself. “Very few concrete lines” does not mean “very little calculation.” The Mar del Plata is by nature a sharp system, where one slow tempo can lead to a torn-open kingside. Switching from 9...Nh5 to 9...a5 changes the shape of the pressure; it does not remove it. A buyer may expect a strategic stroll and receive a game that still demands precise calculation at exactly the decisive moments. At club level, where opponents do not prepare deeply, that feeling may be enough to win. At higher levels, feeling does not replace data. The third blind spot is structural. Any branch advertised as an antidote signs its own sentence: once many people play it, White will analyse it, and the antidote becomes a target. Chess opening history is full of antidotes that once frightened people and now occupy a single row in a database. 9...a5 is a sound branch, known for a long time, and therefore prepared against for a long time. It is not a secret bomb. It is a door that has been open all along, only few choose to walk through it. There is one detail about the source itself that made me write more slowly. The document I read carried a headline about a team-event round in Samarkand, while the entire body discussed the opening product. The two have nothing to do with each other. That is the mark of a text that has passed through many hands with nobody checking it. To a working journalist, that detail matters more than the move itself: if a headline and a body can drift apart like that, the claims inside deserve cross-checking before they are believed. When there is no audience, chess returns to the sound of its own breathing. And when there is no data, opening theory returns to the voice of whoever is selling it. THREE THINGS TO WATCH I am not advising anyone to buy or not to buy. I have only three things to watch, and I will watch them exactly as I watch top-level games. The first is the appearance of 9...a5 in classical games at 2700 level and above. Based on my experience following top-level games, a branch only truly lives when it is played by people who have a reason to play it. Marketing can create a wave of purchases, but not a wave of games. If over the next two seasons ...a5 appears regularly in long-time-control classical games, then that promise has found a foothold. If not, it will stay exactly where it was: a fallback option. The second is the frequency of the Bayonet Attack itself. If 9.b4 remains popular in open and team events, the product's window stays open. If White shifts towards other approaches, the product quickly becomes a chapter in a history book. The third is the identity of the guide. No author's name, no record, no sample games with evaluations. A serious buyer has the right to demand all three before opening his wallet. In chess, credibility does not live in the introduction. It lives in the games a person has played and won. SIXTY DAYS AND THE EMPTY CHAIRS I remember the spring of 2026. When tournaments stopped one after another and nobody came to watch the boards, I shut the door of my apartment, did not answer the phone for two weeks, and every night reopened old games, rewinding and rewinding a single moment. Sixty days of listening to recordings, I imagined the roar of the empty chairs. It was in that stretch that I understood something about openings. People study openings to avoid thinking, yet the opening is precisely where a person is forced to think hardest — because on move ten, nothing on the board tells you that you are right. 9...a5 is an honest move in that sense. It promises nothing. It only changes the question. Age does not extinguish love for the board; it only changes its colour. At thirty-seven, I no longer sit analysing forcing lines to move thirty the way I did at twenty. I sit longer with quiet moves, with the silences that a promotional product cannot sell and a database cannot store. WHAT REMAINS The game ends, but the applause still echoes in my memory. In chess, that applause usually arrives late — after a branch has been stripped bare, after people have forgotten who tried it first. What remains after all of it is not an antidote. It is a habit: sitting with the position longer than is comfortable, enduring ambiguity, and accepting that no move buys peace. If an opening disc can teach that — rather than teach a sequence to memorise — it is worth more than its price. If it cannot, then in a few months, 9...a5 will lie somewhere in a database, as silent as it has always been, waiting for someone else to open it at five in the morning, alone, with a cup of coffee still hot.