9...a5 against the Bayonet Attack: Reading an 'antidote' sold on dated theory

This summer I went back to an old game on my shelf. It came inside a chess opening DVD. The pitch was simple: the Bayonet Attack has become a nightmare for King's Indian players. Here is the antidote. The move is called 9...a5. What caught my attention was not the move itself. 9...a5 is not new. It has existed in theory for years, a known secondary branch. The novelty was the packaging. The product promised a transformation: from the concrete forcing lines after 9...Nh5 into a strategic, low-memorization struggle that leans less on the computer. The title attached to the product pointed to something else entirely — a round-two report from a team event in Samarkand. Two unrelated texts spliced together. That was the first signal, and the most important one. It took me an evening to untangle. Then I realised the real subject was not a publisher's technical accident. It was how the opening-theory industry actually operates. To understand why 9...a5 is sold as a rescue line, you need the broader frame. The King's Indian Defense starts with 1.d4 Nf6 2.c4 g6 and ...Bg7. It is the classic hypermodern opening. Black concedes the centre early, using the bishop on g7 and the files to counterattack later. In the Mar del Plata variation, Black develops the knight to c6, pushes the e5-pawn, and launches a kingside attack. White, meanwhile, floods the queenside. Both sides race on opposite wings. It is one of the most chaotic structures in top-level chess. ECO roughly E97 to E99. Historically, the King's Indian was the opening of big personalities. It is bound to David Bronstein, Efim Geller, and above all Garry Kasparov in his peak years. Kasparov used it as a psychological weapon — not only to win, but to make opponents uneasy from the first move. Its frequent appearance in world-championship matches left a deep mark on generations of players. Even in its golden age, the King's Indian was never a safe choice. It was always the opening of those who accept high risk in exchange for surprise. In the Stockfish and neural-network era, the King's Indian has broadly lost ground at the very top. The reason is difficulty. It demands deep calculation, precise defence, and a certain psychological sacrifice. With the Grünfeld, the Nimzo-Indian, or the Italian Game, players can look for safety. With the King's Indian, you take a risk every time you set up the pieces. The Bayonet Attack — 9.b4 — is a branch of the Mar del Plata. White does not hurry the kingside attack. White pushes the b-pawn to b4, expands on the queenside, and pressures Black's pawn structure. It is the fashionable, irritating answer. Black's main reply is 9...Nh5 — putting the knight on the rim and opening a volatile battle. That is where the product appears. It proposes another fork: 9...a5. Look at the technical logic. After 9...Nh5, the game enters a forcing sequence. Black pushes f5 and threatens the kingside. White, with the knight on h5, either supports the advance or accepts concrete calculation. Both sides must rely on memory and machine help. There is no room for hesitation. That is why the line is tied to deep preparation. After 9...a5, the picture is different. Black does not immediately open the kingside. Black challenges White's b4-pawn, pushing a5 to pressure the structure. The game shifts into a manoeuvring phase. The two sides have not yet clashed. White must choose how to protect the space or accept trades. The tempo is slower. The moves are more strategic than tactical. That is the appeal the product exploits. For players who dislike memorising long variations, 9...a5 promises an escape. For those who like to think independently, the move is attractive. What is notable about the Bayonet Attack is that it does not demand much calculation from White. 9.b4 is a simple strategic move: expand space, pressure the structure, prepare the next plans. White does not need to recall a long forcing chain. White only needs to understand the idea behind the move. That is why the Bayonet became popular. It is learner-friendly. It lets White play strategic chess without superhuman memory. Black, by contrast, faces a hard problem. Answering with 9...Nh5 requires knowing specific lines. Answering with 9...a5 requires having a clear strategic vision. Neither is easy. Now check the data. In all the material I had, not a single concrete game was cited. No ACPL figures. No win rate. No draw rate. No database frequency. No author name. That is a serious gap for a pure technical analysis. But this is a commercial product. And in the commercial market, those numbers are not always needed to sell. I have followed opening-theory products for thirteen years. Their packaging follows a common pattern. Identify a fashionable opponent variation. Offer an answer, usually a known but less popular branch. Emphasise the psychological benefits: less memorisation, less computer dependence, a fit for independent thinkers. The 9...a5 product sits exactly in that mould. The important point: 9...a5 is a legitimate move. It is not a strange invention. It is not a novelty — a move never seen in a database. It is a known secondary branch that can give Black a balanced position. But it is not an absolute weapon. The difference between 9...Nh5 and 9...a5 can be described this way. 9...Nh5 is a head-on confrontation. 9...a5 is a war of attrition. In a head-on confrontation the two sides contest calculation. In a war of attrition they contest patience. Both demand high skill. Neither is easier than the other. What the product advertises — ease — is an over-promise. Here I want to pause. The product claims that after 9...Nh5 the computer matters again, while after 9...a5 computers often do not know how to handle these complex strategic positions. That argument bears the mark of the pre-AlphaZero era. Before 2026, this could be true. Engines then relied on tree search and static evaluation. They often undervalued positions requiring material sacrifice for long-term advantage. That is why many King's Indian variations were once human weapons against machines. After AlphaZero and the shift to neural-network engines, that gap has narrowed considerably. Modern engines handle long-term plans far better than the product describes. The claim that computers cannot understand it no longer holds as presented. It is marketing language surviving from a past era. I recognise the pattern in many other products. Opening adverts usually rest on one of three pillars. Novelty nobody knows about. Counter-intuitive surprise. Simplification of the process. The third pillar is the most common, and the most error-prone. Because no King's Indian Mar del Plata variation truly eliminates calculation. Sharp positions stay sharp. The computer still has to be consulted. Memory still has to be stored. What 9...a5 achieves is to move attention from one hotspot to another. It does not remove pressure. It only shifts pressure from the kingside to the queenside, from concrete calculation to strategic assessment. For players from club level to international master, 9...a5 can be a useful tool in the box. For top grandmasters, it is only a backup, a way to diversify, an option for days when they do not want to face the main line. The product will help most those trying to diversify their opening repertoire. It will help least those who need a weapon to beat strong opponents in a major event. This is also where the title–body mismatch becomes troubling. When a text contradicts itself from headline to body, the reader has no basis to trust any other claim. Source transparency is the foundation of all opening analysis. Without it, every promise floats. I wonder: how many players bought this product hoping for a clean solution to the Bayonet Attack? How many will be neutralised by a well-prepared opponent at their next tournament? The strength of a move does not lie on a theory chart, but in the heart of the person brave enough to play it at the board. The opening-theory industry is one of the most stable segments of professional chess. It is not as loud as tournaments. It is not as visible as title matches. But it exists, and it feeds a large section of grandmasters who specialise in commentary. Its business model is simple. Find a popular variation. Analyse deeply. Package it as a product. Sell it to those who need it. These products compete directly with free databases and strong engines. Their survival rests on three things: author reputation, organised structure, and a promise of saved time. Miss all three and the product sinks. In the case of 9...a5, I have no information on the author. I have no information on the content structure. I have only a time-saving promise, and that promise is backed by a computers-cannot-understand argument — an outdated one. This makes me wonder about the advertising standards of this market. When is a product allowed to call itself an antidote? When does a promise become an unverifiable claim? The answer lies in the incentive structure. A product sells when the buyer believes. The buyer believes when the advert convinces. The advert convinces when it touches a specific fear. The King's Indian player's biggest fear is being forced into a variation they cannot control. The Bayonet Attack touches that fear. 9...a5 promises to soothe it. A closed loop. The problem is that the loop can be broken by board reality. When a White player prepares thoroughly, they may find a way against 9...a5. When that happens, the buyer will be disappointed. By then, the revenue is booked. This is the nature of the theory-product market. It sells hope. Not results. Not all products are like this. Some authors genuinely deliver value. Some products genuinely save hundreds of hours of research. But those usually come with concrete evidence. Sample games. Engine evaluation. Win rates. A reputable author name. Without evidence, a product is only a promise. The 9...a5 product lacks evidence. It has a promise. That is the whole story. In the opening-theory industry, one rarely mentioned fact is the participation of women. Opening products are mainly written and marketed by men. This reflects a broader reality of the global chess world. That absence is not about ability — top female grandmasters are increasingly dominant in youth events. It is about incentives and access. It is a topic I will return to in another piece. The King's Indian Mar del Plata will survive. Not because of any single product, but because the system itself has its own life. It belongs to those who accept risk in exchange for surprise. It belongs to those who believe chess is more than an engine evaluation chart. The Bayonet Attack will remain White's choice. It applies pressure and forces Black to prepare. 9...Nh5 will remain the main move because it imposes tempo. 9...a5 will survive as an alternative, used by a small group. The value of a move is not on an advertising page, but in the fate of those brave enough to believe in it at the board. In that world, the only border is the border of skill, of patience, and of the ability to accept defeat in order to learn. I will return to that game. The move 9...a5. I will test it in a few practice games. Maybe it suits me. Maybe not. But I will not come to it because of a mismatched headline.