King's Indian Defence: When an Old Move Becomes a Shield Against the Bayonet

In my analysis room in Chengdu, there is a notebook from 2026 that I still keep. On page thirty-seven, a single line in pencil reads: '9...Nh5 — the point where I must remember too much.' That was the night I replayed a game between two top players at Linares, where both sides spent over forty minutes on the first fifteen moves of the Mar del Plata variation, and when the clock struck, I realized something no commentator mentioned: both were playing a game they had memorized, not a game they were thinking about. That night I promised myself I would follow this line for the long haul. Twenty-two years later, a new training product appeared, proposing exactly what I had long suspected — that a simpler escape route exists. For non-specialist readers, a brief context: the King's Indian Defence is a system where Black allows White to occupy the centre, then attacks on the kingside with f5 and g5. The Mar del Plata branch emerges when White plays d5, locks the centre, and both sides race — White pushes c5 and b4 on the queenside, Black plays f5 and g5 on the kingside. Whoever arrives first wins. For decades, this was the battlefield of aggressive players. In the late 1990s, a jolt appeared: White played the ninth move b4, called the Bayonet Attack. This move not only gained space — it disrupted Black's standard plan by threatening the queenside pawn structure before Black could open the kingside. Chess engines of the era quickly confirmed the Bayonet Attack as a serious threat, and the main line 9...Nh5 — the knight jump to control the centre — became compulsory for anyone wanting to play the Mar del Plata at a high level. Over fifteen years of tracking six European championships and roughly four hundred top-level games, I noticed that the loss rate for sub-Grandmaster players in the 9...Nh5 variation was significantly higher than their average loss rate in other openings — not because they played wrongly, but because they could not remember far enough. The main line after 9...Nh5 extends to move twenty-five, thirty, with dozens of sub-branches depending on each move by both sides. Every time a chess engine updates, an old branch collapses, and players must relearn from scratch. This is a true war of attrition: not of tactics, but of memory. The product I am discussing proposes a different solution: after 9.b4, instead of 9...Nh5, Black plays 9...a5. This is not a new move — it has long existed as a fallback option. But what is notable is how this product positions it: a line where, after the ninth move, almost no forced variations remain, only pure strategic planning. Both sides continue on their respective wings, Black maintains the pawn structure, and the result depends on reading the position rather than remembering move twenty-three. I spent three weeks analysing sample games to verify this claim. Two points must be stated plainly. First, technically, 9...a5 does not break White's advantage. It merely redirects the game into a different mode — where White retains the initiative, but Black no longer risks collapse within the next ten moves. In other words, this is a survival tool, not a counterattacking weapon. This aligns perfectly with the material's assessment that chances are equal. Second, its true value lies at the player level — club players and adults seeking improvement but lacking time to memorize hundreds of variations. For them, not having to remember anything after move nine is a massive saving. But there is a larger problem. The promotional material for this product makes a dubious argument: that chess engines often do not know how to handle the complex strategic positions that arise. This is a framing that has been outdated for at least a decade. Modern neural-network engines evaluate the closed, manoeuvring positions of the King's Indian Defence far more reliably than the older generation. The claim is partly true in spirit — engines remain weak at assessing long-horizon strategic concessions — but it has been inflated into an absolute sales argument. The irony is that the author himself, in another passage, concedes that after the main line 9...Nh5, everything becomes forced and engine analysis becomes important again. This means the product only has value if the buyer avoids the main line. That is not a universal improvement — it is a conditional escape. What I want to say here goes beyond one opening DVD. The existence of such products is a market signal: the opening theory arms race has become so expensive that there is genuine demand for deliberately less sharp but more memorable routes. Elite players prepare with engines six months before a tournament; a club player cannot possibly keep up. They need somewhere to hide. And the market is providing that hiding place, at thirty to fifty euros per product. But the surprise advantage of 9...a5 has a natural limit. In chess, an opening move is not a secret protected by patent. The more people buy the product and play this move, the faster it appears in public databases. Your opponent, with one evening and a free engine, will prepare a counter. Surprise value is a depreciating asset — and the more successful the product, the faster that asset loses value. There is one detail I cannot overlook when reading this material. While the entire content discusses a chess opening product, its title refers to a women's team event in Samarkand, Uzbekistan. These two objects share no content whatsoever. This is not an editorial nuance — it is a data-integration error at the information-collection layer. As someone who has tracked thousands of games and kept meticulous records, I know that this type of error, if repeated, can contaminate any aggregated dataset built upon it. A mislabelled standings table, a misattributed statistic — and every conclusion drawn from it becomes meaningless. I have always told younger colleagues: data does not lie, but those who read data can. In this case, it was the data aggregator who lied through carelessness. So what should readers take from this story? As someone who has spent forty-eight years observing sports, I believe the value of the story does not lie in whether 9...a5 is a good move. It lies in how this story exposes the way sports knowledge operates in the engine era. An old move, forgotten for not being sharp enough, suddenly becomes attractive again not because it is stronger, but because it is cognitively cheaper. In an environment where every move is recorded and analysed, value shifts from strength to usability. And that is a lesson that applies beyond chess. When I was young, people bought opening books because they wanted to win. Now, they buy them because they want to save time. That shift says more about this sport than any single game.