The 9...a5 Move in the King's Indian Defense: The Battle Between Human Strategists and Memorizing Machines

I remember a late-July afternoon, sitting alone in a small chess club tucked behind the slope on Nguyen Thien Thuat Street in Nha Trang. The sea outside was flat and silent; only the ceiling fan and the ticking of a chess clock filled the room. On the board, two of my students were playing a King's Indian game. The boy with the Black pieces, barely seventeen, pushed his a-pawn to a5 on move nine. I froze mid-pour while filling my tea.

Not because the move was strange. That move has been in the books for a long time. I froze because of the way he did it — slowly, without hesitation, as if opening a door no one else in the room knew existed. He did not look at me for approval. He simply bowed his head and kept calculating.

A record is never old; it is only waiting for someone who knows how to listen.

That was the moment I understood something about Vietnamese chess today: a generation that learns the game through engines, yet still longs to understand positions with its own mind. And 9...a5 — that modest queenside pawn push — has become a symbol of a debate far larger than a single move on a board.

I have followed chess for more than forty years, from games played on old wooden boards in provincial clubs to international tournaments I commentated on for thirty-two years. In that time, I watched chess move from the era of printed books to the era of databases, then into an era where every amateur carries in their pocket an engine stronger than any world champion of the last century. The story I want to tell today begins with a small product — an instructional DVD — but ends with a larger question about how humans choose to face artificial intelligence.

Context: a family of openings called the King's Indian Defense

The King's Indian Defense is one of the oldest and most debated defensive systems in modern chess. Black lets White occupy the entire center, then attacks that center from a distance with minor pieces and pawn breaks. That is the hypermodern philosophy — let the opponent take space, then squeeze that space from within.

The familiar opening moves: 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5. White and Black together build a position where both sides know exactly what they will do. White attacks the queenside, Black attacks the kingside. Whoever arrives first wins. That is why the King's Indian Defense becomes fertile ground for players who love tactical violence — and equally for players who love deep positional understanding.

In Vietnam, this system has had a life of its own. I remember the 1990s, when national youth tournaments were still held in borrowed halls, and I often saw northern coaches teaching students the King's Indian because it trained long calculation. The children had to memorize a whole forest of variations, but the reward was an understanding of positional structure — something Vietnamese chess lacked at the time.

Within that family, there is a special branch named Mar del Plata — after the Argentine city where a 2026 tournament lifted it to prominence. This branch appears when Black plays 7...Nc6, accepting a parallel two-wing attacking race. White pushes pawns on the queenside, Black pushes pawns on the kingside, and both race to see who breaks the enemy king first. This is the kind of chess where one small mistake on move twenty can end the game on move thirty.

And then, from Mar del Plata, players discovered a powerful White attacking option called the Bayonet Attack — the pawn thrust 9.b4.

Why 9.b4 makes Black think

The Bayonet Attack is one of the most uncomfortable weapons White can unleash in the King's Indian Defense. After both sides have castled in the Mar del Plata move order, White pushes the b-pawn to b4, opening lines for the queenside pieces and preparing a major attacking front on the queenside.

What makes 9.b4 such a headache for Black is its dual nature. This pawn move both attacks and defends. It prepares White to push b4-b5, driving Black's knight from its central post, then to open the b-file so the rook and queen can pour down. At the same time, it keeps the queenside pawn structure solid, preventing Black from easily creating a favorable pawn exchange.

I have commentated many games where the Bayonet Attack appeared, and I remember the feeling of players holding the Black pieces when they saw 9.b4. They knew they had to react immediately, but every reaction carried a cost. If passive, White would push b5 and squeeze the queenside. If they counterattacked hastily, they might open doors for White's pieces.

For decades, Black's orthodox answer was 9...Nh5 — the knight leap to the board's edge, preparing to blast open the center with pawn breaks.

9...Nh5: the path of the machines

9...Nh5 is Black's main line against the Bayonet Attack. The knight jumps to h5, ready to leap onward to f4 to pressure White's center, while opening the way for the f7-f5 pawn break to shatter White's pawn structure.

The character of 9...Nh5 is that it opens the position. Both sides are forced to play concrete, precise moves, and any deviation is punished immediately. This is the kind of chess that engines excel at — because it demands extreme calculation, and an engine can evaluate tens of millions of positions per second.

For years, I watched young Vietnamese players pore over games by top grandmasters in the 9...Nh5 branch. They downloaded hundreds of games, memorized variation after variation, and prepared for a fierce theoretical battle. But the price of this path is high: it demands extraordinary memory and an enormous amount of time to keep updating, because every month the engines find a new variation.

I once told my students that the 9...Nh5 path is like competing in a stadium where the stands are packed with machines, each one knowing exactly what move you must play. The stadium is empty, yet the applause of 2026 still echoes in every old record — and in every opening variation, the memory of thousands of past games still waits to judge you.

Precisely because of that burden, part of the chess community began searching for an escape.

9...a5: the path of the positional thinkers

The product I am discussing is an instructional DVD that proposes a way to meet the Bayonet Attack with the move 9...a5. Instead of leaping the knight to h5 and opening the position, Black pushes the queenside pawn to a5.

This is a move with a clear strategic logic. Pushing a7-a5 attacks White's b4-pawn, forcing White to decide whether to hold or trade. At the same time, it opens the way for the queenside rook and prepares the pawn breaks c7-c5 or c7-c6 to pressure the center. Black does not try to seize the initiative at once; Black builds a solid, patient position and waits for White to make a mistake while trying to attack.

Most importantly, the product emphasizes the "strategy over tactics" character of 9...a5. This is a promise of a path with fewer forcing variations, less memorization, where the player needs to understand plans rather than recite sequences of moves. For club players — those without thousands of hours to update theory every week — that promise is very attractive.

I understand that appeal from my own experience. In recent years, coaching children in Nha Trang, I have realized that they can access any chess engine for free with just a phone. But the more they access, the more lost they become in an endless sea of theory. They need a path to understand, not just to memorize.

The product goes further with a more ambitious claim: that in the complex strategic positions arising from 9...a5, engines often do not know how to handle matters. This claim deserves careful analysis, and I will return to it later.

The product's final assessment is that the position is objectively equal, but practically dangerous for White if Black knows the plans.

Decoding the position: what actually happens after 9...a5

To understand why 9...a5 has appeal, one must look at the pawn structure and the plans each side pursues.

After Black plays 9...a5, White's b4-pawn is attacked. White has several choices. White can hold the pawn by pushing b4-b5, gaining space but exposing a weakness on the queenside. White can trade pawns, opening the a-file for Black's rook. Or White can ignore the threat and continue building a kingside attack — but then Black will trade pawns and secure a comfortable position.

The subtlety lies here: each of White's choices leads to a different type of position, and in each type, Black has a clear plan. That is the strength of an opening system built on positional understanding: it does not require you to remember every exact move, it requires you to understand what you will do in each structure.

In the first type, when White holds the pawn with b4-b5, Black can switch to a faster kingside attack, because White's queenside is locked and White cannot easily open a second front. Black pushes pawns on the kingside, opens files, and brings the heavy pieces into play.

In the second type, when White trades pawns, Black gains an open a-file and a solid foothold on the queenside. From there, Black can switch to a central plan with c7-c5, pressuring White's d4-pawn.

In the third type, when White ignores the threat and attacks the kingside, Black simply gains a space advantage and a better pawn structure, then transitions to the middlegame with a comfortable position.

These three types do not require the player to remember hundreds of variations. They require the player to understand three plans. For a busy club player, that is a very reasonable trade-off.

Why the move is not new at all

An interesting thing I noticed when studying 9...a5 in depth: it is not a new invention. Throughout the history of the King's Indian Defense, the a7-a5 pawn push has appeared many times as a space-gaining and tempo-gaining blow on the queenside. Top grandmasters have used it in elite games, not as a secret weapon, but as a reasonable fallback.

What this product does is not invent a move, but systematize it into a complete repertoire — a full set of plans for amateur players. That is genuinely valuable work: turning a move scattered across games into a path one can follow from beginning to end.

But this is also a point to note. When a move is "packaged" into a product, it is often presented with a higher degree of confidence than the evidence warrants. The product says 9...a5 is an antidote, but provides no engine data, no win rate, no database statistics to prove it is better than the main line 9...Nh5.

That is a large gap. An assertion without data remains an assertion.

When machines learned to understand positions

This is the point I want to dwell on longest, because it touches the heart of the story.

The product's central claim is that in the complex strategic positions arising from 9...a5, engines often do not know how to handle matters. This claim was once very common in the 2000s and early 2010s, when people believed engines were only good at tactics, only good at calculation, while deep strategic understanding remained the domain of humans.

But the world has changed. Modern chess engines — from Stockfish to Leela — are no longer just calculating machines. They have learned to evaluate strategic positions reliably, thanks to neural networks and millions of self-play games. In the King's Indian positions once thought to be "understood only by humans," modern engines offer sharp evaluations and reasonable plans no less than grandmasters.

This means the "engines don't understand" claim is steadily losing value. It may still hold partly for weak engines or for amateurs using engines casually. But against a well-prepared opponent — one who has analyzed the position with the strongest engine — that promise is no longer credible.

I once spoke with a young coach in Ho Chi Minh City who teaches his students to analyze every game with an engine. He told me something I have never forgotten: "The engine doesn't understand positions the way we do, but it plays positions better than we do." That is a subtle but important distinction. An engine not "feeling" the beauty of a position does not mean it cannot beat you in that position.

The empty stadium and the machines in the stands

Every stadium I have walked through has a secret door, and the entrance is always a person. In chess, that secret door is often a seemingly small move that opens an entire world of plans behind it.

9...a5 is such a door. It does not appeal for its appearance. It appeals for what it promises: a path out of endless memorization, a route for players who want to understand rather than recite.

But the irony is that the very promise depends on a wavering assumption: that humans understand positions better than machines. In today's chess world, when every opponent can access the same engine, the advantage of understanding is narrowed. The one who understands still has an edge over the one who merely memorizes, but that edge lasts only until the memorizer learns to use the engine to understand.

I remember a game I commentated a few years ago, when a young Vietnamese player held the Black pieces and lost to a foreign opponent in a King's Indian position. After the game, he told me he had prepared the theory carefully but could not understand why his opponent played moves the engine did not suggest. It turned out his opponent had analyzed the position with a stronger engine, at greater depth, and found a plan he had never seen. He lost not for lack of theory, but for lack of understanding supported by the engine.

Lessons from grassroots chess in Nha Trang

The story of 9...a5 reminds me of my own journey with grassroots chess in Nha Trang.

In 2026, I happened to watch a clip of a boy running 400 meters in 47.80 seconds at a grassroots meet. He was not selected for the provincial team for lack of training funds. I spent two months following him, filming on a phone, and the video I cut reached 200,000 views in a month, helping him secure a local sponsor.

That experience taught me that forgotten stories often lie in the humblest places. And 9...a5 is such a story in the chess world: a humble move, overshadowed by the main line 9...Nh5, yet holding a different philosophy of play.

When I teach children at the Nha Trang club, I always try to balance two paths. I teach them to analyze with engines, because that is an indispensable skill of this century. But I also teach them to understand positions with their own minds, because an engine can be turned off, but understanding cannot.

The seventeen-year-old student who pushed 9...a5 that afternoon is an example. He did not recall the move from a book. He understood why the move was reasonable in the position before him. And that is exactly what a strategy-based repertoire — like 9...a5 — tries to convey.

The story of a DVD in the streaming era

There is another aspect of the story I want to address, because it says much about the chess industry.

This product was released as a DVD — a physical, static format that cannot be updated. In an era when opening knowledge changes monthly, a DVD cannot incorporate the corrections engines find after the product's release. That is a structural disadvantage.

The chess industry has shifted strongly toward subscription and streaming platforms. Today, a young player can access interactive opening courses, continuously updated, with exercises and automatic feedback. A DVD — however good its content — remains a product of a bygone era.

This does not mean its content is worthless. A repertoire based on positional understanding retains long-term value, because strategic plans do not age as fast as tactical variations. But how it is conveyed — through a DVD — is becoming increasingly outdated.

I think about this when looking back on my documentary career. I started with VHS tapes, then moved to digital, then to online platforms. With each transition, I lost some of the old audience but gained a larger new one. The chess industry is going through a similar process.

The blind spot of the "engines don't understand" belief

Now is the time to speak plainly about what I consider the product's biggest weakness.

The product rests on a premise: that engines do not understand complex strategic positions, and therefore humans can gain an edge by playing such positions. This premise was once true, but it is steadily losing ground.

Modern engines do not merely calculate well; they evaluate strategic positions subtly. They understand the value of pawn structure, of active pieces, of space. In many King's Indian positions humans consider vague, the strongest engines offer clear and consistent evaluations.

This makes the promise of "playing positions engines don't understand" an increasingly hard promise to keep. An opponent prepared with an engine will not be surprised by strategic complexity; they will embrace it and find a way to exploit it.

The second blind spot is the complete absence of data. The product provides no engine match rate, no average centipawn loss, no win-loss statistics from a database. Meanwhile, the claim that 9...a5 is better than 9...Nh5 is a strong technical claim, demanding corresponding evidence.

The third blind spot is dependence on the author's reputation. Repertoire products often sell on the presenter's fame. If the author is a famous grandmaster specializing in the King's Indian, credibility would be far higher. If the author is anonymous or little known, credibility is lower. In this case, information about the author is not provided, and that is a significant gap.

A counterintuitive angle: the real value lies elsewhere

The most counterintuitive thing about this story is that the product's real value may not lie in the move 9...a5.

If we judge the product solely on the question "is 9...a5 theoretically better than 9...Nh5," the answer is: we cannot know, because there is no data. But if we judge it on the question "does it help club players understand positions better," the answer may be yes.

A strategy-based repertoire has its own educational value. It teaches players how to think about pawn structure, about plans, about transitions between phases. Those lessons can be applied to many other positions, even if the specific 9...a5 never appears on their board.

This is another way to view the product's value. Instead of treating it as a theoretical weapon to beat Bayonet specialists, we can treat it as an educational tool helping amateurs understand the King's Indian Defense more deeply.

This view also fits market reality. Most buyers of such a repertoire are not grandmasters, but club players wanting to improve their understanding. For them, value lies not in whether the move is engine-confirmed, but in whether it helps them play better.

When belief meets reality

I have spent many afternoons at the Nha Trang club observing how my students approach products like this.

They download them, skim them, then try to apply them in online games. Some succeed and feel more confident with the King's Indian. Some fail and return to the main line 9...Nh5. No result is universal, because chess does not work that way.

What I learned from observing this is: the value of a repertoire depends on who uses it. A strategy-based repertoire like 9...a5 can be a gift for a player who loves to understand, and a trap for a player who loves to memorize. If you buy it hoping for a magical sequence of moves, you will be disappointed. If you buy it hoping to understand positions more deeply, you may be satisfied.

I think this is true of every opening product, not just 9...a5. An opening never wins a game by itself; it only creates a playground. The player is the one who decides the outcome on that playground.

A question of format and the future of opening study

The story of this DVD also opens a larger question about how we will study chess in the future.

For decades, opening knowledge was transmitted through books, magazines, and multimedia products. Each format had its strengths: books for slow reflection, magazines for fast updates, and multimedia for visual illustration.

Today, interactive platforms are changing how openings are studied. They let learners move pieces on the board, receive instant feedback, and track their progress. They can also be continuously updated as theory changes.

A DVD lacks those capabilities. It is a product of an era when opening knowledge changed more slowly and when updating required buying a new product.

But this does not mean a DVD's content is worthless. Some positional lessons do not age over time. The philosophy of playing based on understanding rather than memorization retains its value. The issue is only that how it is conveyed needs updating.

I think about this when considering how I will convey chess knowledge to the next generation. I will use every available tool — engines, interactive platforms, video — but I will always emphasize understanding. Because an engine can be turned off, but a mind that knows how to think cannot.

Signals to watch

If I had to name the signals a chess observer should watch in this story, I would choose four.

The first is the practical performance of 9...a5 against the Bayonet Attack. If in coming years the move appears more often in elite games and scores well, its credibility will rise. If it remains a rare branch, it will remain a choice for club players.

The second is the frequency of the King's Indian Defense at the elite level. In the engine era, this system has at times declined in frequency as top players moved to other defenses. If that trend continues, the commercial ceiling of such a product will shrink.

The third is the development of engines in evaluating strategic positions. If engines continue to improve this ability, the "engines don't understand" premise will grow weaker, and the value of repertoires built on it will fall accordingly.

The fourth is the migration of opening study to interactive platforms. If this trend continues, static formats like DVDs will be used less and less, and their content will need restructuring for new formats.

Reflecting on humans and machines

At sixty-five, I have witnessed many transitions in the chess industry. I remember the era when a good player was judged by the ability to calculate in the head, when opening books were treated as treasures, and when analyzing a game required hours with a board and a notebook.

Today, everything can be accessed in seconds. A seventeen-year-old in Nha Trang can analyze his own game with an engine stronger than any world champion of the last century. That is a miraculous change, and also a challenge.

The challenge lies here: when every answer is available, understanding the question becomes more important than ever. An engine can tell you the best move, but it cannot tell you why that move is best. Understanding the reason remains human work.

The 9...a5 move in the King's Indian Defense is a small example of this. It is not the best move according to an engine. It is a strategically reasonable move, and its value depends on whether the player understands it.

Conclusion: a move that is never old

I return to that late-July afternoon, when the seventeen-year-old student pushed 9...a5 and I froze mid-pour while filling my tea.

After the game, I asked him why he chose that move. He answered that he was not sure it was best, but he understood it. He understood that it attacked the b4-pawn, that it opened the way for the rook, that it kept the queenside flexible. He chose it because he understood it, not because an engine told him to.

That is what I want to convey in this article. In a world where engines can answer every chess question, the value of understanding remains intact. A move is never old; it is only waiting for someone who knows how to listen.

9...a5 may not be the final answer to the Bayonet Attack. It may not be the best move according to any engine. But it is a reminder that chess remains a human game, and that understanding a position remains an irreplaceable skill.

The question I leave readers with is not whether 9...a5 is better than 9...Nh5. The question is: when engines can play better than us in every position, why do we still play chess? Perhaps the answer lies in the very moment my student bowed his head, kept calculating, and understood something an engine could not understand for him.

At sixty-five, I still believe every game is a human story. And the best stories — like the story of the modest 9...a5 — are still waiting to be told.