9...a5 Against the Bayonet Attack: A Promise of Equality and a Data Gap

White pushes the b-pawn to b4 on move nine. No applause, no commentator raising his voice, just a pawn advancing two squares toward the queenside. Yet for three decades that quiet move — the Bayonet Attack in the King's Indian Defense — has forced thousands of Black players to memorize variation chains like multiplication tables, positions in which engines calculate a dozen moves ahead of humans. A recent opening DVD claims to have the answer: 9...a5, a gentle queenside push leading to "sound positions with very few concrete lines," and "objectively, chances are equal." Three advertising sentences, not a single number attached. People call that a shock; I call it data that has not been read. The King's Indian Defense is the classic hypermodern system: Black deliberately concedes the center to White and then strikes back at it. The familiar sequence 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 builds a bishop on the long diagonal, and the Mar del Plata Variation arises when Black plays ...e5 and ...Nc6, accepting a closed center so that both sides can attack on opposite wings. This system was once a primary weapon in world championship matches during the Garry Kasparov era. Its share at elite level has declined since the 1990s, but at club level the King's Indian Defense retains considerable vitality, largely because its attacking plans are clear and easy to navigate for amateur players. The Bayonet Attack is White's ninth-move answer: 9.b4. White grabs queenside space before Black can organize counterplay, while opening the way for the b4-pawn to advance to b5 and drive away the knight. Black's mainline reply is 9...Nh5, sending the knight to the rim and preparing ...f5 and a kingside attack. Theoretically this is one of the most heavily analyzed branches of modern chess, to the point that an amateur cannot enter it confidently with only a few free evenings each week. That is why the DVD exists. It does not sell an invention; it sells an escape route. And escape routes always have their own market: club players, tournament regulars who know King's Indian structures but no longer have time to memorize several hundred more moves of the 9...Nh5 branch. The move the DVD proposes, 9...a5, has its own logic, and this is where the data starts speaking. Instead of sending the knight to the rim, Black pushes the rook pawn, preparing to open lines for a rook or bishop while pressuring the pawn structure White has just expanded with b4. The strategic idea is clear: Black does not fight head-on in a closed center but shifts to a battle for squares, where pieces are repositioned rather than thrown into forced variations. The appeal lies in the fact that 9...a5 does not require precise move-by-move recall. It requires structural understanding. That kind of promise is easy to believe, because anyone who has struggled with 9...Nh5 wants an escape. But this is where I stop and check the source, because in my profession a claim without supporting data is only an untested hypothesis. Numbers are asceticism: you must give up convenience before you can see the truth. In the promotional material for this DVD there is not a single figure. No ACPL, the average centipawn loss per move. No engine match rate against the top choice. No win rate, no draw rate, no master-game sample. All that is present are a few qualitative statements, written to be memorable but hard to verify. I am not saying the move is wrong. I am saying it has not been proven. Based on my experience following games and opening databases, a sideline such as 9...a5 tends to operate through a familiar mechanism. Black accepts a theoretically looser position in exchange for a lighter memorization burden and better odds against an unprepared opponent. At club level that trade is usually favorable, because the opponent cannot recall thirty moves of the mainline either. At elite level the trade is usually unfavorable, because the opponent's analytical team has a multi-million-game database and can find an exploitation route in a single afternoon. The distance between those two environments is the entire market for the product. There is one further detail the reader needs to know, and it concerns information quality rather than chess. The material I received for analysis carried a headline about an open tournament in Samarkand, reporting that Uzbek players continued their winning run in round seven. The content beneath that headline discussed an opening DVD. Not one information point mentioned Samarkand, Uzbekistan, any specific game or any specific player. A tournament headline and an opening-product body, stitched together by a source-processing error. The three-source discipline I have followed throughout my career means something simple: when three sources point to the same number, check whether all three are genuinely independent. Here I do not even have a second source. The headline and the body are two different texts, and one of them has been filed in the wrong place. A reader looking for round-seven results from an event in Samarkand will find nothing in this material. A reader wanting to learn 9...a5 should not believe the product has been validated by an international tournament. Back to the board. The claim that "objectively, chances are equal" is the claim most easily misread. In opening language, equality describes a technical assessment, not a level of comfort. Black may reach a balanced position and still labor for the next forty moves, because equality means only that neither side has a forcing advantage — not that the game plays itself. This is where opening advertising often slips. A buyer reads the word equal and understands the word safe, then sits down in a position never played before against an opponent who has prepared it. The disappointment does not come from the move; it comes from the expectation. There is also a cyclical paradox: the more widely a sideline is advertised, the less surprise value it retains. If 9...a5 becomes common at club level, White will begin preparing for it, and the burden shifts back to Black. No escape route in the opening comes free. The opening market is a place where people pay for unverified variables. That does not make the market worthless; it simply imposes clearer requirements on the buyer. A good opening product must state three things: how many games it rests on, what Black's win-and-draw rate is within that sample, and how the strongest opponent has refuted it. Without those three, the remainder is belief. What is worth keeping from this story is a genuine strategic idea. 9...a5 is a reasonable option for King's Indian players who want to avoid the forcing branch, provided they accept learning structures rather than memorizing chains. Alongside it comes a verification method: any opening claim should be checked against game databases and engine evaluation before entering a preparation file. And finally a warning about sourcing: when a document's headline and body concern two different events, the problem is not on the board. Over the coming months I will track how often 9...a5 appears in King's Indian games between roughly 1500 and 2300 rating on online platforms. In parallel, White's score against the move, and whether new antidotes emerge. The most anticipated signal is a top player adopting 9...a5 in a classical game, because a single such game would be enough to move this sideline out of advertising and into theory. For now, the ninth move simply sits there, waiting for someone handling the black pieces patient enough to read it through data rather than belief.