The 9...a5 Antidote to the Bayonet Attack: How the King's Indian Defence Split Into Two Schools

The 9...a5 Antidote to the Bayonet Attack: How the King's Indian Defence Split Into Two Schools

On a late August afternoon in Da Nang, my training room held only four people: two fifteen-year-old students, a wall-mounted board, and a laptop open to an analysis engine. One boy pushed his pawn to b4, planting it directly in front of Black's e5 pawn — the ninth move of the Bayonet Attack in the King's Indian Defence. He looked at me and said: “Coach, the book says Black must bring the knight to h5. But I can't remember the twenty moves that follow.” I told him to hold off on memorising, then opened a twenty-page document I had printed years earlier, one that contained a move very few Vietnamese players ever mention: 9...a5.

Some six thousand kilometres from Da Nang, a European schools chess championship had just closed in Tsaghkadzor, Armenia, with a prize ceremony for young champions and an accompanying cultural programme. The two events have nothing to do with each other. Yet both point to a question I have pursued across eighteen years in this profession: between memorising a sequence of moves and understanding a position, which one actually keeps a player at the board?

A System Born From Conceding the Centre

The King's Indian Defence belongs to the hypermodern school. Black does not seize the centre with pawns from the start; Black lets White build the d4-e4 pawn pair, then turns that very pair into a target. The bishop sits on the long diagonal at g7, far in the corner, waiting for the day it fires into d4. The pawns on e5 and c5 become two pincers squeezing the centre from both sides. This is chess built on faith in structure, not on occupying territory.

I remember the first time I saw the King's Indian on a real board rather than a screen. It was a youth event in central Vietnam, around 2026, when I was still moving between the roles of player and tournament organiser. An older player, twenty years my senior, sat at the board with the black pieces, and for the first forty moves he barely glanced at the queenside. He watched only the kingside, where the f5 and g5 pawns crept forward one square at a time. By move fifty, White's entire kingside had collapsed. I asked him his secret. He gave me a sentence I still use to teach: “I don't remember moves. I remember the position I want to have twenty moves from now.”

That sentence is the key to the whole story of 9...a5. The King's Indian is not an opening to memorise; it is an opening to understand. And precisely because of that, whenever a new line appears in it, the right question is not “is this line stronger than that one,” but “what does this line require the player to understand?”

The history of the King's Indian is bound to the Soviet chess school. From the 1950s, players such as Isaac Boleslavsky, David Bronstein and Efim Geller turned it from an opening considered risky into a strategic weapon. In the 1980s and 1990s, Garry Kasparov took it to the summit in world championship matches, and from there it became one of the most popular openings at every level. When an opening travels from suspicion to millions of players, it stops being an idea; it becomes an ecosystem. And every ecosystem produces branches.

Mar del Plata: A Two-Wing Race

To understand the Bayonet, you must pass through the Mar del Plata Variation. It takes its name from the tournament held in Mar del Plata, Argentina, in 2026, where the first games of this system were recorded and analysed. The basic move order: 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5 7.0-0 Nc6. Black has completed the frame: the e5 pawn controls d4, the g7 bishop presses along the long diagonal, and the c6 knight applies pressure to the d4 pawn.

What makes Mar del Plata distinctive is that both sides accept a race. White pushes the d5 pawn, locking the centre. Once the centre is locked, the two wings become independent battlefields. White masses pieces on the queenside, pushes b4 and c5, and attacks Black's pawn structure. Black masses pieces on the kingside, pushes f5 then f4, g5 then g4, h5, and finally launches a frontal assault on White's king. Whoever arrives first wins. Whoever is one tempo late loses a king.

I once analysed hundreds of Mar del Plata games during my years as an assistant analyst. What struck me was not the attacks but the rhythm. In Mar del Plata, both sides count. White counts how many moves are needed to open the b-file or c-file. Black counts how many moves are needed for the f4 pawn to touch g3. That rhythm is so unforgiving that a harmless waiting move can reverse the result. It is why Mar del Plata is considered one of the hardest lines in classical chess.

The Bayonet: The Ninth Move That Changed the Game

For decades, White in the Mar del Plata followed a familiar plan: push c5, open the c-file, and wage the queenside war. But in the late twentieth century a different direction appeared and quickly became fashionable: 9.b4. Players called it the Bayonet. The name says it all: instead of slowly opening the c-file, White drives a bayonet straight into Black's queenside flank.

The move 9.b4 does three things at once. It gains space on the queenside. It threatens the further push b5, chasing the e7 knight — the piece that had just left c6 to escape the d5 pawn — into an even worse square. And most importantly, it disrupts Black's familiar rhythm. Black wants f5 and g5, but now Black must first answer a question: where is that b4 pawn going?

Based on my experience following games from 2026 to the present, the Bayonet has a feature few mention: it shifts the game's focus from calculation to structure. In classical Mar del Plata, both sides hurl pre-calculated attacks at each other through move thirty. In the Bayonet, White often does not need to calculate that far. White needs only a better structure and a clearer plan. That is why the Bayonet became a favourite weapon for players seeking to escape the “computer war.”

9...Nh5: The Main Road and Its Price

Black's main reply to the Bayonet is 9...Nh5. The move brings the knight from f6 to h5 with a clear intention: to hop to f4, threaten the e2 pawn, and set up a kingside assault. Black will then push f5, g5, g4, and the battle unfolds exactly as in classical Mar del Plata. A typical main line runs 9...Nh5 10.Re1 Nf4 11.Bf1, after which Black continues f5, g5, and White must try to open the queenside before the kingside collapses.

The trouble with 9...Nh5 is that it opens the position. Once the knight has jumped to h5 and f4, the options narrow. White can calculate every move precisely, and at the highest level both sides can prepare at home through move thirty. Put another way, 9...Nh5 turns the game into a memorisation contest. For professionals, that is routine. For club players, it is a nightmare.

I have been on Black's side in many such games. Once, in a tournament in central Vietnam, I prepared through move twenty-five. My opponent deviated on move eighteen. I had to recalculate from scratch, and in ten minutes of thought I played two consecutive errors. I lost. After that game I understood something that became a teaching principle: home preparation cannot save you when the opponent leaves the book.

That very gap is the market for lines like 9...a5.

9...a5: A Strategic Choice

The move 9...a5 is a rare response. It directly challenges White's b4 pawn. Instead of storming the kingside as 9...Nh5 does, Black grabs queenside space first, slows White's plan to open the b-file, and keeps the pawn structure flexible. The idea behind it is clear: Black wants a strategic game in which plans matter more than precise move sequences.

Specifically, 9...a5 accomplishes several things. First, it attacks the b4 pawn, forcing White to decide: push b5 to keep space, or trade pawns to open a file. Both choices carry a cost. If White pushes b5, the queenside structure is fixed, and Black can drill into it with ...c5, or with ...b6 and ...Ba6. If White trades, the a-file opens, but Black can use that same file for pressure. Second, 9...a5 prepares ...c5, the second pincer into the centre. Third, and most subtly, 9...a5 keeps the kingside options intact. Black has not committed to ...Nh5, ...Nf4 or ...f5. Black can wait to see what White does before deciding.

In such a position, Black plays by reading and reacting. This is the style I call the chess of the knowledgeable, as opposed to the chess of the memoriser. The knowledgeable player does not need to know what move twenty-five is. That player needs only to know that once the structure settles, the kingside will decide matters and the b4 pawn will be a long-term target.

But here I must stop and speak plainly. Claims about the strength of 9...a5 usually come with a familiar argument: that computers cannot handle such complex strategic positions. The argument sounds seductive, and it has sold a great many instructional DVDs over two decades. But it weakens with every passing year.

What the Computer Does Not Say

I remember the COVID season of 2026, when Da Nang was locked down and every tournament was postponed. I sat in an empty room, no spectators, no noise, only a screen and a board. In the middle of COVID, sitting in an empty venue, I could hear the breathing of the tactical system. It was during that period that I began re-checking every old belief with data rather than memory.

And the data said something rather blunt. Modern chess engines — Stockfish, Leela Chess Zero — handle the strategic positions of the King's Indian far better than humans do. They do not get lost in closed, non-forcing positions. They evaluate the worth of a pawn structure, of a weak square, of a knight buried on e7. To say that “computers don't understand these positions” is a statement from a previous decade, not the present one.

This does not mean 9...a5 is useless. It means its value does not lie in beating computers. Its value lies in beating humans — specifically, players who have spent hundreds of hours memorising the main line 9...Nh5 and are surprised when an opponent branches elsewhere. This is practical value, not theoretical value. And the two must be distinguished clearly.

I must also say something about evidence. The materials promoting 9...a5 that I have read rarely include a single concrete number: no engine match rate, no average centipawn loss per move, no win-draw-loss rate from a database. It is all persuasive prose. For an analyst like me, that is a large gap. A move cannot be proven superior to another merely by calling it “more strategic.”

My verification method is simple, and it grew out of a mistake I made young. I once mislabelled a recording of SHB Da Nang, and from that day I knew that football does not forgive carelessness. In 2026, at twenty-five, I began work as an assistant analyst for the SHB Da Nang coaching staff. In the round-12 V-League match against Hanoi FC, I was assigned to log every set piece. I misidentified the position of a corner kick in the 63rd minute, and the head coach criticised me in front of the whole team. For a month afterwards, I re-watched the recordings of five rounds, divided the pitch into eight zones, and built my own notation system. My opponent reports then helped the team neutralise 70 percent of the dangerous set-piece situations in the return leg. A recording mislabelled early in life is the most expensive lesson: the eye always needs verification.

Applying that principle to 9...a5, I did three things. I searched the game database of strong players to see how often 9...a5 actually appears and with what results. I ran the engine at great depth to see how it evaluates the position after 9...a5 compared with 9...Nh5. And I put myself in both players' shoes to see which plan is easier to execute under real tournament conditions.

The result adjusted my view. 9...a5 is not a theoretical refutation. It is a practical tool. For a strong club player who rarely meets the Bayonet at a high level and has no time to memorise the twenty moves of the main line, 9...a5 is a sensible choice. For a professional facing carefully prepared opponents, 9...a5 can become a trap for the person playing it, if that person does not genuinely understand the plans behind it.

The Mislabelled Recording

One detail in this story made me pause longer than the technical analysis itself. When I went back to the original material, I found a strange mismatch: the article's title spoke of a European schools chess championship and a prize ceremony for young champions in Tsaghkadzor, Armenia, but the entire body concerned an opening-repertoire product, specifically the move 9...a5 against the Bayonet. Two entirely different subjects sat inside one document.

Having once mislabelled a recording, I recognised the sign at once. It was exactly the feeling of fast-forwarding a tape and finding that the team I was analysing was not the team on the pitch. Something had been put in the wrong place. And when a text contradicts itself from the title onward, every conclusion drawn from it must be lowered in confidence.

Notably, this kind of mismatch is not rare in chess content. A catchy headline about a youth event can be attached to a product introduction simply because both contain the keyword “chess.” For the reader, the consequence is concrete: they look for information about a European schools championship and receive an opening advertisement. For the writer, it is a stain on credibility. The empty pitch turns out to be the truest mirror of modern football, and the same holds for the board: with the noise of media removed, the real structure of a text is exposed.

This mismatch also reminds me why I abandoned the habit of writing by inspiration. In analysis, the worst thing is not being wrong; it is being right about something irrelevant. The reader loses time, and the writer loses credibility. Had I not checked the title against the body three times, I might have written an analysis of a European schools event based on the data of an opening product — a mistake I would live with for a long time.

An Old Format and a Question About the Future

There is one more aspect I want to raise, because it bears directly on how chess knowledge travels. Instructional DVDs were the gold standard of the industry for two decades. Today, however, opening knowledge flows through subscription platforms and interactive databases. A static DVD cannot be updated after publication. When a line is refuted by a major game, the DVD still sits there, teaching exactly what it always taught.

This is a structural disadvantage, not a content one. The content of a good opening product retains value. But the format determines its lifespan. A line like 9...a5, placed on an interactive platform, can be updated whenever a new high-level game appears. Placed on a DVD, it is frozen at the moment of publication.

For the Vietnamese market, this has practical meaning. Chess clubs in Hanoi, Ho Chi Minh City and Da Nang increasingly use software and online databases for training. My fifteen-year-old students already know how to search databases and analyse their own games. To that generation, an instructional DVD is an exotic object. They do not need someone to tell them that 9...a5 is good. They need someone to show them how to verify for themselves whether it is.

Lessons for Vietnamese Chess

I return to my student in Da Nang. He asked whether he should memorise the main line or learn 9...a5. I told him the question was not yet right. The right question is: what kind of player do you want to become? If you want to play professionally, you must know the main line 9...Nh5, because it is the common language of strong players. If you want to play at club level and score against players of your own strength, 9...a5 can be a good weapon — provided you understand the plans behind it, not merely the move.

That is why I believe Vietnamese chess must shift from teaching memorisation to teaching understanding. We have talented players who have risen to world class, such as Le Quang Liem and Nguyen Ngoc Truong Son, who proved that Vietnamese players can compete at the highest level. But behind them, the gap between a promising junior and an international player remains wide. Part of that gap lies in how we teach openings: teaching moves instead of plans.

A DVD teaching 9...a5 will teach a student to remember that after move nine, Black plays a5. A good coach will teach the student to understand that after move nine, Black is trying to slow White's queenside in order to gain time for the kingside. The difference between these two approaches lies not in the move but in what the student carries out of the training room.

Esports and football share one clock: whoever keeps the rhythm when pressure rises wins. Chess is the same. In a six-hour game, the first thing to collapse is not calculating ability but the ability to hold a planned structure while the clock runs. A player who memorises thirty moves will lose direction on move thirty-one. A player who understands the plan will know what to do when stepping outside the book.

So is 9...a5 the answer to that problem? I do not think so. It is only an example. It shows that even in an opening analysed to exhaustion like the King's Indian, branches remain where understanding matters more than memory. Its value does not lie in being stronger than 9...Nh5. Its value lies in forcing the player to think differently.

And that is perhaps what I want to say to those producing chess content in Vietnam. The market is flooded with opening introductions, promises of “secret weapons” and “the strongest refutation.” But readers do not need another promise. They need someone to show them how to verify for themselves. A headline about a European schools event attached to an opening advertisement helps no one. An honest analysis of where 9...a5 actually stands in the theoretical picture — with data, with caveats, with the uncertain parts laid bare — is what helps.

I once mislabelled a recording of SHB Da Nang, and from that day I knew that football does not forgive carelessness. Neither does the board. Every wrong move leaves a trace, and every trace can be read — if the reader takes the time to verify the source. That is the final lesson I draw from the story of 9...a5, and it is what I will keep teaching my students in Da Nang, in a room of only four people, where the noise of the outside world cannot drown out the breathing of a position being understood correctly.