9...a5 against the Bayonet Attack: a strategic move and the limits of an opening promise

On the wooden board by the window of a small room in Nha Trang, the b4 square carries a faint scratch. A fifteen-year-old student of mine made it during a rainy-season training session, right after he placed a white pawn there for the third time and asked: "Teacher, why do we keep pushing b4 and then lose it?"

I did not answer immediately. I rebuilt the Mar del Plata position of the King's Indian Defense, placed a black pawn on a5, and said: "Because while the white pawn sits on b4, it holds a piece of space. Anyone who questions that piece forces you to explain how much you actually understand about the game."

The 9...a5 I placed on the board that evening was not invented. It is a real move in opening theory, and it has recently been presented as an antidote to the Bayonet Attack in the King's Indian. The move itself is not new. The way it is being told is the part worth examining, and also the part most likely to mislead a student of the game.

Context: the board before move nine

The King's Indian Defense belongs to the hypermodern family. Black concedes the centre for a few moves, then strikes back with ...e5, ...c5 and a kingside campaign. In the Mar del Plata branch, Black develops the knight to c6, pushes ...e5, then retreats the knight to e7 and prepares ...f5. The name comes from a tournament held in Argentina in 2026, where the ...Nc6 and ...e5 structure first drew wide attention. By the standard encyclopaedia classification, the branch sits between E97 and E99.

The familiar starting position: 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5 7.0-0 Nc6 8.d5 Ne7. White plants the d5 pawn as a wedge; Black keeps the e5 pawn as the wall opposite. The centre locks, and both sides understand that the rest of the game will be fought on the wings.

Garry Kasparov used this system in world championship matches, turning it into a weapon for players who enjoy complex positions. In Vietnam, Le Quang Liem and Nguyen Ngoc Truong Son are the two names local fans follow most closely in the 1.d4 opening group, and Indian structures still appear in their international games.

Against that structure, White can choose 9.b4, the Bayonet Attack. The move looks like a blunt push, but it is really a measurement of space: the b4 pawn controls a5 and c5, clears the way for a bishop to reach a3, and prepares the b5 advance.

Black has two answers. 9...Nh5 is the main line: the knight hops to the rim, eyes f4, and the game becomes ...f5, ...f4, g5, g4, ...Ng6, ...Rf7, ...Bf8, h5 - an avalanche on the kingside, backed by an enormous body of theory that anyone who has played the King's Indian knows well. 9...a5 is the sideline: Black questions the b4 pawn immediately, before White can stabilise the queenside, and steers the game towards a more closed struggle.

The opening product promoting 9...a5 tells the story this way: the chances are balanced, there are very few concrete lines, computers often struggle with such complex strategic positions, and therefore a Black player can enter the Mar del Plata without memorising anything. It is an appealing promise. Like every opening promise, it deserves to be measured before it is believed.

The geometry of the b4 square

The b4 pawn does not occupy the centre. It occupies space. In a position where the centre is locked, space is the only currency still available for trade. A pawn on b4 controls a5 and c5, which means it takes away two squares the black knight and bishop want to use. It also clears the way for the c1 bishop to reach a3: from a3, that bishop both defends b4 and looks straight at d6 and e7 along the a3-f8 diagonal. A bishop on a3 pressures exactly the two pieces Black needs to build the kingside attack.

If White finds time to push b5, the queenside becomes a solid block of space. The c6 square leaves Black's control, and any plan involving ...c6 to pressure d5 becomes harder to execute.

It is not where a piece stands, but where a piece is about to go, that forms the real piece of space. That is why the b4 pawn should not be assessed by its current square, but by the set of squares it locks behind it.

Black asks back with 9...a5

9...a5 does one very specific thing: it forces White to decide about the b4 pawn right now, while White's development is still incomplete. White can trade with bxa5, accepting a stray pawn on a5 - at once a spatial asset and a long-term target. White can also keep the tension with Ba3, defending b4 and continuing the original programme.

Each choice leads to a different kind of game, and that is the real value of 9...a5: it does not force White into a chain of compulsory moves, it merely changes the subject. From a theoretical race on the kingside, the game shifts into a positional bargain on the queenside, where both sides must decide for themselves rather than replay memory.

For a club player, that is a genuine plus. For a 2600 player, it is a minus, because a non-forcing move also gives the opponent more ways to play. The same move, two different frames of reference.

Croatia's pressing problem in 2026 was never about speed; it was about how they redrew the map of the pitch. The same holds for 9...a5: it is not faster than 9...Nh5, it simply redraws the map of the board.

Why 9...a5 has a place in an overloaded database

Based on my experience tracking games over nearly two decades, opening theory has thickened so fast that memory has become a taxed resource. In football, the five-substitution rule deepens squads but turns the last twenty minutes into a war of attrition. In chess, the equivalent war of attrition is the sheer volume of opening lines: the more branches there are, the more likely a player runs out of energy around move twenty-five, and mistakes appear not from ignorance but from exhausted focus.

That is the economic reason these products exist. They do not sell the strongest move. They sell a lighter head. For an amateur, a system that lets you enter a game through understanding rather than memorisation has real value, even when the system is not theoretically superior.

In exchange, the buyer pays a different price. A less forcing system also means the opponent has more choices, and in a closed position the player who understands long-term plans beats the player who only remembers moves. A lighter head is not free.

Blind spot one: the computer story is dated

The claim that computers often struggle with complex strategic positions was once true. It belongs to the era before neural-network engines. Modern engines evaluate closed King's Indian structures far better than the older search-based generation, and the top players themselves changed how they read such positions after learning from machines.

When a dated technical argument is still used as a selling point, the reader should separate the technical claim from the marketing. The story remains useful to a buyer, but in a different way: it describes the fear of a club player facing a well-prepared opponent, not the current state of opening theory.

Blind spot two: the antidote and inflated expectations

9...a5 is a known sideline, not an invention. Any White player who prepares seriously for the Bayonet has a response, and a sideline that has been analysed thoroughly loses its surprise value over time. Its real worth sits at the amateur and semi-professional level, where opponents rarely have time to open a board and hunt for a solution.

Promotional language tends to inflate expectations with absolute sentences: if Black knows what to do, things may quickly become dangerous for White. Such sentences describe a direction, not a result. In chess, any claim without data should be read as a hypothesis awaiting verification, especially when the line being sold comes with no sample game at all.

Blind spot three: the noise between headline and body

One detail in the document I read is worth recording. The headline referred to round two of an open team event held in Samarkand, while the entire body discussed a King's Indian opening product. Two unrelated texts had been spliced into one package.

For a chess reader, this is a familiar reminder. Opening information travels through several layers: the author, the producer, the distribution platform, the reposter. Each layer can add noise, and the final layer is usually the least verified. Measure first, conclude later - a principle that applies to the board and to the page alike.

Blind spot four: the buyer is sold an identity

This product does not only sell moves. It sells the image of a player who thinks for himself rather than memorises, someone willing to enter complex positions armed with understanding. That is a highly effective way to position customers in the amateur chess market, and it is also why buyers struggle to evaluate the product objectively: they are purchasing a piece of their own identity.

Tactics only become complete when they are told in a language the players dare to believe. A promise aimed at identity is easier to believe than a dry analytical table, even when the table is more accurate.

What to verify in the next game

In my next game analysis, I will set up this exact position and check three things. First, whether White trades on a5 or keeps the tension with Ba3 - and what that choice says about the White player's preparation. Second, whether Black truly keeps the position closed or gets dragged into a kingside race at White's tempo. Third, whether this move appears in elite games or only lives inside commercially sold opening material.

For my student, the most important answer was about the piece of space. Losing the b4 pawn does not mean White has lost the position. It only means the bargain has begun, and whoever understands the map knows what is being traded for what.