Chess: 9...a5 Against the Bayonet Attack — Mar del Plata's 'Antidote' and the Trap of an Old Promise

There is a habit I have kept through 32 years in this trade: whenever someone says 'the computer does not understand this position,' I open a database and cross-check before I believe it. Last night, reading the blurb for a chess DVD teaching the 9...a5 line against the Bayonet Attack in the King's Indian Defense, Mar del Plata variation, I did exactly that. The wording was so familiar it made me stop: 'computers often do not know how to handle the arising complex strategic positions.' I rewound the tape — yes, rewound, an old documentary habit — but this time I rewound two decades of engine history. I trust footage more than testimony, because footage does not know how to lie. And I am grateful for my ISTJ wiring: it forces me to pause in front of an ad line that sounds too smooth. In chess, a promise without data is just an unverified move.

Context: a branch undervalued in the engine era

The King's Indian Defense is the classic hypermodern system: 1.d4 Nf6 2.c4 g6, with Black conceding the centre early and then attacking it with ...Bg7 and ...e5. Its logic runs against intuition: instead of taking space, Black invites White to build a big centre and then turns that centre into a target. The Mar del Plata variation — with Black's ...Nc6 and ...e5 — is the sharpest version of the whole system, where the two sides attack opposite wings: White piles up on the queenside, Black blows open the kingside. The ECO codes for this branch fall in the E97–E99 range.

The Bayonet Attack is the move 9.b4, a pawn thrust gaining queenside space that White uses to smother Black's counterplay before it forms. It is not a beautiful move. It is a practical one, and precisely because it is practical it has been a headache for Black players for decades.

For much of that time, the orthodox answer to 9.b4 was 9...Nh5 — a move that opens the position, forces concrete lines, and compels both sides to remember exact moves. That is why 9...Nh5 became the main line in every theory manual, and also why it became a burden: any Black player must memorise hundreds of moves, and forgetting one branch loses the whole game.

The DVD I read offers another fork: 9...a5. This is not an invention. It is a known sideline, overshadowed by the pull of 9...Nh5. What the product sells is not a new move but a change in the character of the game: from an arena of memory to an arena of strategic thought. 'Very few concrete lines' — that is the core promise, and also the point I want to dissect.

Here I should be clear about how I follow chess. I do not sit at the board. I sit in the edit room, frame-scrubbing, counting moves, noting the moments the media skips. That experience tells me that theory products like this live or die not on whether the move is objectively strong, but on whether it can withstand an opponent's preparation. And that criterion can only be answered by data, not by enthusiasm.

Core: the promise and the gaps

I built a small comparison table, exactly as an ISTJ chess player does when evaluating a line.

On sophistication: 9...a5 sits in the 'orthodox but secondary' group. It is a recognised antidote, not a discovery. The material itself admits this, placing 9...Nh5 as the main move and 9...a5 as the alternative.

On empirical evidence: none. No ACPL figure, no win rate, no sample games are cited. The product promises 'very few concrete lines' but supplies no statistic to prove that fewer lines means easier play. That is the first thing I underlined in red.

On objectivity: the material states that chances are equal. That is a positional claim, not a claim of theoretical superiority. In other words, the product's value is practical and psychological rather than theoretical. It does not say Black is better. It says Black can play without memorising. That is a completely different sales pitch.

The most interesting part — and the point I want to dwell on longest — is how the product frames a human-versus-engine contrast. It writes: 9...Nh5 opens the position, forces the lines, and makes computer analysis important again; whereas 9...a5 leads to sound positions with very few concrete lines, positions that computers often do not know how to handle.

I read that sentence and saw a timestamp. That is the language of the pre-AlphaZero era. In today's neural-network engine age, engines handle long-term King's Indian strategic positions far better than that sentence implies. Saying a computer does not understand a KID strategic position is speaking the dialect of a decade gone by. I do not blame the copywriter. I simply log the timestamp, because that is a documentarian's job.

All of this leads to a provisional conclusion: the competitive value of 9...a5 is real but modest. It is a legitimate secondary branch that can reach equality, but it does not 'solve' the Bayonet in the sense the advertising suggests. And here I must say it plainly: the real value of a theory product lies in helping club players decide faster at the board, not in a promise that it neutralises a popular line at the very top.

Contrarian angle: when I have to audit myself

But wait. If I stopped there, I would fall into exactly the trap I warn others about: defaulting to the opposite position in order to look sharp. So I re-checked my own argument against the same data standard.

First, what evidence do I have that 9...a5 does not work? None. I only have evidence that it is not new, and that its claims lack figures. Those are two very different things. An old move can still be strong. A claim without data can still be true.

Second, what evidence do I have that neural-network engines 'understand' KID positions better? Some — but it is an inference from the general trend of the chess world, not from a specific measurement in this line. In other words, I am also arguing from belief, only my belief is newer. That is not a rigorous argument. It is a grounded assumption, and I should call it by its right name.

Third, and most importantly: I found a source problem that anyone citing this material must handle first. The headline refers to a team event in Samarkand, round two, with all top teams winning, while the entire body is an opening-DVD blurb. The two are unrelated. That is a hole in information integrity. If I put it into a script without a warning flag, I would be aiding a propagated error. For someone with 32 years in this trade, a title that does not match its body is not a small matter; it signals that the whole information package needs re-verification from scratch.

I remember a time I was dismissed. I rewound the SEA Games 29 footage to find what people had missed — 47 strides per minute in Nguyen Thi Oanh's final lap, a figure a director waved away because he assumed women know nothing about running. I did not argue. I just opened the data appendix. Here too. I do not argue with the DVD. I ask for one thing: show me a real game, with engine evaluations attached. If the product cannot, its promise remains unverified.

And there is something worth noting about the market. The product positions itself for players who like to think for themselves, not memorisers. That is identity-based selling, aimed at the self-image of a buyer group. Such marketing works, but it proves nothing about the quality of the move. I am all too familiar with this reasoning in sport: turning a personal taste into a technical advantage.

An empty stadium and the sound of spikes

You may wonder why a chess writer is dissecting a DVD here. Because for me, opening theory is also a track. Every opening course is a starting line; the audience only sees the runner reaching the finish, while a documentarian like me lives at every start. I have hurdled all my life, just never been timed. And that experience taught me that crowds always prefer the story of an upset, of a miraculous antidote, over the story of a secondary branch played carefully across many years.

During the pandemic, when stadiums stood empty, I recorded the sound of the Ha Nam women's football team's spikes on the old gymnasium floor and found a rhythm like a battle drum. An empty stadium, and I turned the sound of spikes into the heartbeat of the film. Nobody noticed it until it reached the screen and people heard it. 9...a5 is the same kind of thing: a small sound inside a loud symphony, not the lead trumpet, but possibly the thing that keeps the time.

That makes me ask: if an opening line produces no moment for the ages, is it still worth writing about? My answer is yes. What is worth writing is not the moment but the way a position forces people to think. And a secondary branch that teaches people to think is a secondary branch worth recording.

Takeaway: what to watch

I am not sure 9...a5 will appear more often at major events in the coming years. What I am sure of is how we should track it: open the database, count its appearances in high-rated games, and see whether White finds a counter after one theory cycle. If it holds, it will prove itself. If not, it will quietly return to its old place — a respectable secondary branch.

For me, the point is not whether a move 'beats' a computer. The point is what a DVD dares to promise, and whether the buyer has the tools to verify that promise. I trust footage more than testimony. I trust a real game more than an ad line. And I believe Vietnamese chess fans deserve analysis that can be opened up for verification, rather than pretty promises that leave no trace. If you have ever bought an opening course because you believed a line of advertising, try once asking the seller to show you the data. That is the first move, and the hardest one.