From Tsaghkadzor to the Grand Final: The Move 9...a5 and the Queenside Corridor of the King's Indian Defense

From Tsaghkadzor to the Grand Final: The Move 9...a5 and the Queenside Corridor

Tsaghkadzor, Armenia, a late-July afternoon. In the hall of the European Schools Chess Championship, the clocks ticked as evenly as rain on a tin roof. I sat in the row against the east wall — the seat nobody wants, because from there you only see the backs of the players, never their faces. But from that hidden corner, I could see the whole battle on board four.

A thirteen-year-old boy in the Armenian team shirt held the black pieces. His opponent, a Polish player, had just pushed b4 — the ninth move. The Armenian boy's hand paused above the board. I knew that pause. It was not hesitation. It was the moment a player realises the opponent has led him into a room with no windows.

Then he pushed a5.

That move — 9...a5 — is why I was in Tsaghkadzor.

From the sideline, I see the whole game. I do not sit at the front board, where the national coaches crowd together taking notes. I sit at the edge, where you only see what the centre cannot be bothered to look at: hands, breathing, and a black pawn that has just left a7 to go looking for space.

Context: a schools event, and a room with no windows

The European Schools Chess Championship is a competition for school-age players, organised annually by the European Chess Union and rotated among member federations. Armenia — a country where chess is a compulsory school subject — is one of the familiar hosts of this circuit. Tsaghkadzor, a resort town in Kotayk Province at around 1,800 metres above sea level, has repeatedly welcomed youth and schools events. The evening cold keeps you alert, and for a chess writer, alertness is a working condition.

But the context of today's story is not the town. It is board four, on move nine.

To understand why 9...a5 was worth a trip, you have to understand where it sits on the map of opening theory. The King's Indian Defense, KID for short, is a classic hypermodern opening: Black concedes the centre to White, builds an indirect fortress with the bishop on g7, then counterattacks the centre with ...e5 and ...d6. The familiar start: 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5 7.0-0 Nc6. This is the Mar del Plata system — the sharpest line of the KID, where White attacks the queenside and Black attacks the kingside, and both sides know that whoever is one tempo late loses the whole game.

The Mar del Plata was once the weapon of champions. Bobby Fischer used it to reshape Western chess. Garry Kasparov turned it into an attacking machine throughout the 1980s and 1990s. Mikhail Tal, with a magician's eye, turned KID positions into games nobody dared blink through. In the recent generation, Levon Aronian — Armenia's own pride — has regularly carried the KID into major arenas, and at schools events like Tsaghkadzor, Armenian children grow up on exactly those games.

In the Mar del Plata, White's ninth move offers a famous choice: 9.b4, the Bayonet Attack. The idea is clear: White pushes b4 to expand queenside space, threaten b5 and c5, disrupt Black's ...c5 plan, and force Black to answer in a position where White has already seized the spatial initiative. The Bayonet is not a move to win material. It is a move to win the right to ask questions. After 9.b4, White is no longer playing chess — White is setting the exam.

And Black, for decades, has had a standard answer: 9...Nh5.

Core: why 9...Nh5 is the machine's road, and 9...a5 is the human's

To understand 9...a5, you must first understand why 9...Nh5 became the main move. The black knight jumps to h5 for two reasons: to prepare ...f5, opening the kingside and attacking; and to free the f4 square for other pieces while pressuring the centre. But the price is that the position opens up, and once it opens, the lines become concrete to the point of cruelty. After 9...Nh5, both sides enter a forest of moves where a half-tempo mistake is punished by a tactical blow. This is the terrain computers dominate, because computers are not afraid of memorisation. Computers remember everything.

9...a5 is different. This move does not demand memory. It demands understanding.

Look at the structure: after 9.b4, White's b4-pawn has advanced and become a target. Black's a5-pawn immediately strikes at the base of White's queenside structure. It prepares ...c5 faster, because Black no longer needs to spend time defending the a-pawn. It also prepares ...Be6, ...Nbd7, and ...Rc8 — constructive moves, not forcing moves. And most importantly: it does not open the kingside. Black keeps the g7 fortress intact, keeps a healthy pawn structure, and shifts the game into a war of plans rather than a war of variations.

The real value of 9...a5 does not lie in being objectively stronger than 9...Nh5 — it lies in changing the type of game from a memory contest into an understanding contest. This is the point most opening commentary skips, because it measures only by score and forgets to measure by the kind of skill the position demands.

In recent years I have noticed a repeating pattern at youth events. Young players, raised on databases and analysis engines, are often excellent in forcing positions. They know sharp lines to move twenty-five by heart. But when they enter a position where no move is forced, where every move is a strategic choice, they slow down. Not because they are weak. Because they have never been taught how to decide inside empty space. 9...a5 is exactly such a position.

On board four in Tsaghkadzor, the Armenian boy understood this. He did not try to attack at once. He pushed ...a5, then ...c5, then ...Be6, then ...Nbd7. None of those was a tactical blow. But after twelve moves he had a position where White's b4-pawn was a weakness and Black's pawn structure was a block of stone. He did not win with a punch. He won with a room.

Trade-offs: the price of choosing clarity

No move is free. 9...a5 carries three costs, and a serious writer must state all three.

The first cost is space. By not playing ...Nh5 and ...f5, Black gives up his fastest kingside attacking weapon. White gains time to consolidate the queenside and may steer into a position where the spatial advantage of the b4-pawn matters more than being a target.

The second cost is two-sidedness. 9...Nh5, however sharp, gives Black clear winning chances. 9...a5 more often leads to balanced positions, where Black can play safely but also finds it harder to win. For a young player who needs points, that is a meaningful trade. For a player who needs wins, it may be a concession.

The third cost is surprise. 9...a5 is a rarely played move. That means opponents are unprepared, but it also means the player using it has fewer top-level games to learn from. A rare move is a double-edged sword: it surprises the opponent, but it also walks its user into territory where he himself is blind.

People watch players run. I watch the whole formation shift. In chess it is the same: people watch the move. I watch the space that move opens or closes. 9...a5 is not a good move in the absolute sense. It is a good move in the relative sense — good for a player who understands plans, and poor for a player hunting variations.

Product and truth: when a DVD sells a belief

There is one more layer to this story, and that layer is what made me sit down and write.

On the international chess market there is a line of opening-training products on DVD — or, more recently, digital courses — in which an expert presents a complete opening system for one side. One such product covers the King's Indian Defense, and its headline is precisely the move 9...a5 I have just analysed. The product's marketing makes three claims: first, that 9...a5 avoids forced lines; second, that it leads to positions where strategy matters more than tactics; and third — the most notable claim — that computers often do not know how to handle such complex strategic positions.

The first claim is true. The second is true. The third deserves serious scrutiny.

At a time when chess engines were weaker than humans in closed positions, this claim had a basis. But in the current era, the strongest engines — Stockfish or Leela — evaluate the strategic positions of the KID reliably. They are no longer lost in rooms without windows. That does not make 9...a5 useless. It only makes the sales argument weaker over time.

And here is the most important point: that product offers not a single number. No engine match rate. No average centipawn loss. No database statistics. No sample size. A technical claim without technical data remains a claim. When someone tells me a move is better than the main move, I do not ask how strongly they believe it. I ask how many games they counted.

I have been on the other side of that test.

Seven months without football, seven months of asking why. In 2026, when the pandemic stopped every league in the world, I fell into a professional identity crisis. There were no matches to write about. So I did the only thing a data addict can do: I built my own database. Seven months, 214 goalless draws across five major European leagues from 2026 to 2026, classified into nine pressing models. When football returned in June, my first piece on how empty stadiums affected pressing tempo was emailed back by three Premier League clubs asking for more.

The lesson from those seven months is simple: a system does not lie, but it can only be heard when the data is thick enough. And the data here is not thick enough. 9...a5 may be right, may be wrong, may be a merely average move dressed in a good story. But a good story does not replace a database.

Contrarian: the error is in the headline, not the move

Now to the part I really want to say.

When I began researching this topic, I met a strange paradox. The source document I read carried a headline about an entirely different subject: it spoke of European schools chess champions, of a journey "from Tsaghkadzor to the grand final," of a cultural celebration. But the entire body content spoke of a King's Indian Defense opening product and the move 9...a5.

The two subjects are unrelated. One is a report on a youth event. The other is a product introduction. And they were stitched under the same roof.

I mention this not to catch a document out. I mention it because it is a lesson in how we read sports news. In an era when content is generated, copied, and restructured at the speed of light, a headline can say one thing and the body another — and the reader, who reads only the headline, carries a false conclusion into every later conversation. I have written about this in a football context for years. But it holds for chess exactly the same way.

A data mismatch at the source layer is not a small error. It is an error that multiplies. Because every analysis behind it is built on it, and every decision behind that rests on the analysis.

So when I stood in Tsaghkadzor and watched the Armenian boy push 9...a5, I did not see a product. I saw a person. And the interesting thing is this: the move a DVD is trying to sell to adults, a child from Armenia found by his own school instinct. He needed no marketing claim. He only needed to understand that the queenside is where the game is decided.

Depth: a schools event as an indicator of the future

This is where I must speak about what opening commentary never says.

A move does not exist in a vacuum. It exists inside a training ecosystem. And the training ecosystem is what decides, ten years later, which move becomes orthodox.

Armenia is an almost inimitable example. The country makes chess a compulsory subject in general education. The result is a generation of players with solid technical foundations, who are not afraid of closed positions, not afraid of wars of plans. Levon Aronian is the peak of that stream, but the stream has more than one peak. It has a basin.

And in Tsaghkadzor, I saw that basin.

What caught my attention was not one specific boy. It was the pattern. Young players from countries with a technical training tradition — Armenia, India, Uzbekistan, Iran — seemed more comfortable in strategic positions than players from countries where coaching culture revolves around analysis engines and memorising variations. It is a small signal, and I have no data to prove it. But it matches what I have observed for years.

In football, people often speak of youth coaching chasing results and neglecting technique. The push to physicalise the U18 age group is eroding the technical soil. In chess there is a similar but less discussed trend: coaching children for short-term results by cramming variations, instead of teaching how to understand positions. A child stuffed with three hundred lines wins this week's tournament and struggles for the next decade. A child taught to understand loses a few games this week and owns an entire career.

Patience is not stillness. Patience is waiting for the opponent's pressing rhythm. In chess, patience is waiting for the right moment to push a flank pawn. The Armenian boy on board four was patient in that way. He did not need a DVD. He needed an education, and he had one.

Industry: when the format dies before the content

The product I mentioned above is distributed on DVD — or, if newer, as a fixed downloadable course. Both formats are at the mature-to-declining stage of the chess-knowledge product life cycle. Today, opening knowledge flows through interactive subscription platforms, where content is updated continuously, where learners can ask and receive answers, and where a theoretical error is corrected within days rather than years.

A static DVD has an unfixable structural disadvantage: it cannot absorb post-publication engine corrections. When you buy a DVD, you buy a photograph of theory at the moment of filming. When you subscribe to a platform, you buy a living stream.

That does not make DVDs worthless. It makes their value bounded by their format. And in a market where buyers are younger, more subscription-native, and more demanding of interactivity, a static product fights a match whose rules do not favour it.

People remember a goal. I remember the space in the right corridor. In chess it is the same: people remember a good move. I remember the format through which that move was transmitted. And format, more often than people think, decides which content survives.

In Tsaghkadzor, I saw the future in raw form. The children there do not learn from DVDs. They learn from coaches, from online databases, and from sitting for hours at the board with a patient teacher. If you want to know where opening theory is heading in ten years, do not look at what adults are buying. Look at what children are playing.

Risk and blind spots: three things to watch

From everything I have observed and analysed, there are three signals I will track in the coming months.

First, the real-world score of 9...a5 in top-level games. If the move appears at the 2600+ level and scores positively, its credibility will rise sharply. If it appears only in amateur events, it will remain a personal taste.

Second, the frequency of the King's Indian Defense at the elite level. In the engine era, the KID has steadily lost ground in top arenas, as leading players increasingly prefer other defences. If that trend continues, the commercial ceiling of any KID product — however good — will be limited.

Third, and most important to me: whether young players like the Armenian boy on board four keep finding strategic moves by instinct. Because if they do, the whole "human versus machine" debate becomes obsolete. Not because the machine wins, but because humans have learned to play with the machine without losing their instinct.

What I got wrong, and what I still believe

I must confess something. There was a period in my career when I believed data could settle every question. I believed that if you had enough numbers, you had the answer. I was wrong.

In 2026, I wrote six pieces before a World Cup, predicting a team would exit in the quarter-finals because its defence was too thin. I had calculated carefully. I had the numbers. And I was wrong, because I underestimated bench depth and the coach's ability to read a game. The week after the final, I published a 4,800-word self-critique, analysing exactly what made my prediction collapse. Since then, I have added a section to the end of every deep analysis, called "What I got wrong."

So today, that section reads like this: I do not know whether 9...a5 is genuinely better than 9...Nh5. I have no data to assert it, and I will not pretend otherwise. What I know is that the move exists, that it has logic, that it suits a particular kind of player, and that it is being sold on a story stronger than its evidence.

But I also know this: in Tsaghkadzor, a thirteen-year-old boy played it without needing any story at all. He played it because it was right. And sometimes, that is the only test required.

I am sixty-nine. I still learn from the kids. Football never retires, and neither does chess.

Takeaway: a question for the next game

When the European schools championship closed in Tsaghkadzor, the organisers handed out medals and held a cultural ceremony — exactly as schools chess events do, because chess, in places like this, is not merely a competition but a part of identity. I left the hall with a notebook full of observations and one unanswered question.

That question is: can a move exist independently of the data about it?

Perhaps yes. Perhaps no. But what I know for certain is this — in ten years there will be a generation of young players who never needed a DVD to learn 9...a5. They will learn it from a teacher, from a living database, from a long afternoon at the board. And when they play it at some major event, someone will sit in the sideline corridor, noting every move, and wonder whether that move is truly strong — or merely a position in which, this time, the one who understood the plan beat the one who remembered the variations.

I will be one of those sitting there. And I will bring my spreadsheet.