Bayonet Attack and 9...a5: Which Escape Route for the King's Indian Defense?

I sat in front of the screen at two in the morning, reopened my game database and scrolled to move nine. White pushed the pawn to b4. On the board, it was just a small, modest, almost silent move. But to anyone who has ever played the King's Indian Defense, it is an alarm bell. The Bayonet Attack — White's dagger — has just been unsheathed. In more than thirty years of watching the chess world, I have learned one thing: small moves often carry large decisions. 9.b4 is not pretty. It is practical. And that very practicality turns it into a nightmare for generations of Black players. I replayed the SEA Games 29 footage to find what people had forgotten, and I do the same with the chessboard. This time, what was forgotten lies on a flank move. The King's Indian Defense was born in the 1920s as a manifesto of the hypermodern school. Black concedes central space to White, then strikes back at that very center from the kingside. A bold, somewhat reckless idea that demands deep calculation. The Mar del Plata Variation — named after the Argentine coastal city — is the main line of this defense. After 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 O-O 6.Be2 e5, White usually replies 7.O-O Nc6 8.d5 Ne7, closing the center. From here, the two sides split into two fronts: White attacks the queenside, Black attacks the kingside. The race begins. The golden age of this defense is tied to Garry Kasparov. He turned it into a personal weapon, forcing the entire chess world to re-examine how to respond. But by the 2010s, as chess engines became overwhelmingly strong, many lines of the King's Indian were decoded to the last centipawn. Top players gradually turned away. At the elite level, it no longer holds the status it had in Kasparov's era. At club level, however, the story is completely different. The King's Indian Defense remains a popular choice. It is dynamic, it attacks, it gives Black a chance to play for a win rather than play not to lose. In the Mar del Plata line, the Bayonet Attack with 9.b4 is one of the most uncomfortable challenges. White immediately opens the queenside, seizes space, and prepares a flood of pawns. Black must respond precisely. The orthodox response for decades has been 9...Nh5. The knight leaps to the rim, ready to jump to f4, controlling key dark squares. Theoretically, it is an acknowledged move, analyzed to the end. But it has one fatal weakness for club players: it leads to forced, concrete variations that engines have dissected move by move. No team of seconds supports you at the club. You must remember. And once memory collapses mid-game, you lose everything. That is the context in which a recent chess DVD proposes an alternative: 9...a5. Instead of moving the knight, Black pushes the rook pawn. A quiet, almost humble move. I scrolled through classic games in my database. 9...a5 appears sporadically, not prominently. But its idea is clear: Black pressures White's queenside pawn structure, prepares to open the a-file, and avoids the forced variations of the main line. This is not a new invention. Players have known about it for a long time. What is notable is the way it is packaged as an antidote to the Bayonet Attack. In terms of sophistication, 9...a5 is strategically oriented rather than tactical. Compared with 9...Nh5, it is less dominated by the computer. In terms of execution stability, it sits at a medium-high level for club players. Fewer concrete lines mean less to memorize. But it still demands deep understanding of the queenside structure and the art of piece regrouping. The DVD claims that 9...a5 leads to sound positions with very few concrete lines and that, objectively, chances are equal. I read those two sentences and paused. Where is the data? There is no engine match rate. No ACPL — average centipawn loss per move. No win rate, no draw rate. No sample games cited. No concrete number backs up the claim of equal chances. In my documentary profession, a claim without data is a claim left hanging. I trust footage more than testimony, because footage does not lie. Here, no footage is offered. This does not mean 9...a5 is wrong. Its strategic idea is reasonable. It genuinely avoids the forced nature of 9...Nh5, where engines handle things better than humans. But the equal-chances claim must be understood correctly: it is a practical balance for a prepared Black player, not a refutation of the Bayonet Attack. There is an important nuance here. When the DVD says engines often do not know how to handle these positions, that is a marketing-friendly phrasing. Modern chess engines do not fail to handle them — they evaluate positions differently from humans, often favoring moves that humans struggle to find under time pressure. 9...a5 may make the position harder for engines to read, but that does not mean it is objectively better. It only means: it is harder to analyze, and therefore harder to neutralize through pure computer preparation. I return to the structure of the position. In 9...a5, Black accepts a somewhat loose queenside pawn structure in exchange for flexibility. The a-file becomes a front. White may seize space, but Black gains time to regroup. No move is forced. That is precisely the appeal. The target audience of this product is clear: King's Indian players at club level who do not want to memorize the long variations of 9...Nh5. This is a real market. Not everyone has the time to sit in front of a computer for five hours a day. But there is something worth reflecting on regarding the value of surprise. If 9...a5 becomes popular among club players, White will adjust. They might switch to other anti-King's Indian systems — the Sämisch, the Fianchetto, the Averbakh — or study the resulting queenside structures more carefully. Then the preparation burden shifts back to Black. This is the rule of every opening weapon: its value is inversely proportional to its popularity. This product belongs to the traditional chess-education market, where practical opening DVDs target club players. Meanwhile, the industry has partly shifted toward interactive courses, subscription platforms, and streaming-based learning. A standalone DVD on a sideline has limited reach. But it still reflects a real trend: amateur players increasingly prioritize understanding over memorization, practical positions over optimal theory. There is an issue off the board that I cannot ignore. The original article's title is Samarkand R7 (open): Uzbekistan continue dominating run, extend lead — a report on round 7 of a tournament in Samarkand, where Uzbekistan continues its dominant streak. But the content below does not mention Samarkand, does not mention Uzbekistan, has no round 7, no results whatsoever. The entire content is about a chess DVD and the move 9...a5. This is a mismatch between title and content. For someone whose profession is verifying information, that is a red flag. It points to a problem with the source data. Perhaps a tournament report and a product advertisement were merged. Perhaps an automated process paired two different texts. The real concern is not the DVD itself — it can still be useful. The concern is the integrity of the information. If a headline about Uzbekistan was attached to content about an opening move, how much other information in that data stream might be similarly misaligned? I do not write by crowd emotion. I always ask: what is this data actually telling me in reverse? And here, the data says: before believing anything about Samarkand or Uzbekistan, verify the source. There is one more detail. 9...a5 is not a new invention. It is an old line that has been repackaged. That is not bad — many good ideas that fell out of fashion deserve revival. But we must distinguish between a practical line and one that is over-hyped. The difference lies in data, and the data is missing. Back to the board. 9...a5 may be a practical weapon for King's Indian players. But do not let a marketing line promising equal chances lull you to sleep. Balance on the board never comes by itself. It is built through understanding, through time, and through genuinely tested games. I have run hurdles my whole life, only I was never timed. And in chess it is the same — flank moves never tell their own story. You must rewind yourself, count yourself, verify yourself. If you do not, you are just playing by a headline that no one has checked.