9...a5 and the Promise of No Memorization: Inside a King's Indian Opening Product

In a press room in Shanghai, while the whole room was arguing about a transfer fee, I stayed behind with a laptop and an unfinished game. On move nine, Black did not play 9...Nh5. The hand pushed the a-pawn to a5, slowly, the way someone closes a window against a gust of wind. There was no applause. Nobody in that room knew a decision had just been made.

Seven years later that same move appeared on the cover of an opening-theory product: a digital training disc teaching how to meet the Bayonet Attack, sold as an antidote for players exhausted by memorization. A move that once drifted past in the silence of a press room is now boxed, labelled and shipped to club players who keep getting crushed on move fifteen.

I spent a few evenings with that product. What I found was a way of telling a story rather than an invention.

The King's Indian Defence opens with 1.d4 Nf6 2.c4 g6 and continues with ...Bg7. Black lets White take the centre first, then attacks that same centre — the approach analysts call hypermodern, though the essence is simpler: Black trades space for a durable structure and a clear plan.

The Mar del Plata is the sharpest branch of the system. After 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5 7.0-0 Nc6 8.d5 Ne7, White grabs space with d5 while Black mass-produces pieces on the kingside and prepares ...f5. In ECO terms the line sits around E97 to E99. It was a familiar weapon for Garry Kasparov at the top level in the 1980s and 1990s, and those games are why thousands of club players bought King's Indian books.

Then came the Bayonet Attack. White's 9.b4, pushing the b-pawn to b4 for queenside space and the threat of b5, became a popular answer to the Mar del Plata from the 1990s onwards. Black has one main reply: 9...Nh5, preparing ...f5, opening the position, and turning the game into a contest in which every move has been machine-checked to move twenty. That is exactly why the Bayonet became a nightmare for King's Indian lovers: you do not lose because you play badly, you lose because your opponent remembers more than you.

And this is the habitat of the opening-theory market — one of the most stable and least glamorous revenue streams in chess. Books, Fritztrainer-style discs, online courses, free-platform studies. They all sell the same thing: an answer to a move you are tired of meeting.

The product I am describing offers 9...a5 as an antidote to the Bayonet Attack. The strategic idea is easy to grasp: instead of opening the position with ...Nh5 and ...f5, Black hits the base of White's queenside with 9...a5, threatening to push a4 or open the a-file, forcing White to decide about the queenside pawn structure immediately. The position then closes, forcing variations are scarce, and both sides enter a positional struggle in which the black player is free to think for himself.

The product's real value is psychological, not theoretical. I want to be precise about this, because the promotional material confesses it: no engine evaluation, no win rate, no sample game, no named author. Only a promise about how few lines there are.

In my trade, when a product says "very few concrete lines", I write the phrase down. Here, the data does not exist. The buyer does not know whether "few" means five lines or fifty; does not know at what depth they were checked, with which software, at what date; and does not know whether they have already been analysed back by the other side.

The material frames things as human against machine: 9...Nh5 "opens the position, forces the lines, and makes computer analysis important again", while 9...a5 "leads to sound positions with very few concrete lines". Alongside that comes a notable claim: computers often do not know how to handle such complex strategic positions.

I read that sentence several times, and I think it is dated in an interesting way. More than fifteen years ago, when engines evaluated positions mainly by search and raw numbers, one could believe it. Today, neural-network engines handle long-term, closed, low-forcing positions far better than the sentence implies. Saying a machine does not understand a closed King's Indian is a marketing line from an earlier era, reused intact as a brand.

That does not make 9...a5 bad. It only makes the reason to buy blurrier.

There is one dry detail I want to preserve, because it belongs to the real world rather than the board: the original headline refers to a team round in Samarkand — round two, open section, all the top teams winning — while the entire body of the text is about an opening disc. Two unrelated documents were packed into one file. In journalism that happens daily, and it reminds me that the source I am reading never passed a single verification step.

Put differently: the antidote is being sold in a box whose label was stuck on the wrong side. I say this without sarcasm. I say it because it is the only instantly verifiable fact in the whole package.

There is one more layer, and this is the part I consider most honest in the entire campaign: the target reader is described as "players who like to think for themselves, not memorizers". That sentence sells a self-image. It tells the buyer: you are not lazy, you are not weak, you are simply a person with character. In a market where a product competes directly with free databases and free engines, the only thing that cannot be copied is the feeling of being yourself.

An opening disc sells you permission to believe you are still thinking freely.

And I will be honest: I have bought that feeling before.

Based on my experience watching games and preparation sessions, I have an odd habit. Every Wednesday evening I visit the junior training room, sit at the back table and record what they do. Not opening files. I record sound. Keyboard clicks. Someone inhaling when the engine returns an evaluation different from expectation. A fifteen-year-old asking the friend beside him: "If the machine says this move is best, why do I still feel uncomfortable?"

When a seventeen-year-old prodigy does not yet know he is about to become a legend, he does not argue about moves like 9...a5. He learns 9...Nh5, because that is what everyone around him plays.

That is why I do not believe this product is for the young players on their way to the top. It is for the rest of us.

The counterintuitive angle I bring from years spent outside every theory race is this: the real target of an antidote is never the opponent's position. It is the opponent's time.

When you play 9...a5 against someone who has prepared 9...Nh5 for three weeks, you are not trying to prove that 9...a5 is objectively stronger. You are liquidating an investment. You are taking from your opponent the one thing a database cannot return in time: familiarity. On the board, the first thing destroyed is not the pawn structure, it is memory.

But that is precisely why the strategy has a short shelf life. A secondary branch is only valuable until it is analysed back. In the opening world, the life cycle of an antidote runs from a few months to a few years, and it usually ends with a White-side answer printed in some theory magazine, decorated with a tiny exclamation mark.

There is another temptation King's Indian lovers know well: believing that if we know what we are doing, White will soon be in danger. That is an advertising sentence, and it cannot be falsified, because it depends entirely on the opponent's preparation level — a variable nobody can measure before the game. At club level it is frequently true. At international master level it is true far less often.

I still remember an afternoon at Hongkou stadium during the period when football had to be played in silence, when I brought a recorder and sat listening to studs scraping on artificial turf. An empty stadium still holds the applause of the people who are gone. So does a chessboard. Moves like 9...a5, 9...Nh5, and every unnamed sideline nobody bothered to name are still there, in the silence, waiting for someone patient enough to sit and listen again.

There is a dry reality the promotional material does not mention: a product like this usually costs about as much as dinner in Shanghai, while the same money, or less, already buys access to online databases holding millions of games and a free engine. The only difference is that someone curated and narrated it for you. That has genuine value, but it is the value of editing, not of truth.

And in this particular case, the editor is anonymous.

So what should be tracked? Three signals. First, whether 9...a5 appears in classical games above the 2700 mark — if it does, the product's promise is partly validated; if not, it remains a book line. Second, the frequency of the Bayonet Attack 9.b4 in professional opening trees — because demand for an antidote exists only as long as someone plays 9.b4. Third, the source itself: whether that Samarkand headline can be matched to a real tournament report, to show how carelessly this package was assembled.

If you are wondering whether such moves are still alive, the answer lies in the price list. A contract is only paper, while a dream is signed with sweat — and in this case, the dream is signed with the hours you spend reading a branch somebody else abandoned.

What is worth keeping from this story sits somewhere other than the board: a sideline, forgotten in analysis rooms, can still be woken up by a sufficiently attractive promise. The World Cup taught me that truth sometimes loses to a shot from distance, and chess is no different: good advertising can beat a correct evaluation. If you intend to buy that antidote, buy it as you would buy a quiet seat, not as you would buy a truth. After three decades of writing, I have realised I never wrote about sport — I wrote about how people find themselves a reason to keep sitting down at the board.