King's Indian: Why 9...a5 Is Replacing 9...Nh5 Against the Bayonet Attack

Two weeks ago, in a group analysis session at the Nha Trang chess club, a seventeen-year-old student of mine set up the game he had just lost in thirty-four moves. He was Black, playing the King's Indian Defence, and he had followed the book precisely to move nine: 9...Nh5. His opponent pushed 10.dxe5 and then jumped in with 11.Nd5. For the next twelve moves he had to recall every single move exactly — no thinking allowed, only remembering. He misremembered on move seventeen. Six minutes were left on his clock, and the game ended there.

What stayed with me was not the wrong move. It was what he said as he cleared the board: “I knew them all, teacher. I just didn't understand what I was doing.”

This week a chess publisher released an opening-theory DVD proposing 9...a5 as the answer in place of 9...Nh5 against the Bayonet Attack, 9.b4. Their marketing is notable: “very few concrete lines,” “strategy instead of tactics,” “you don't need to memorise.” Reading it, I thought immediately of that student. What he lacked was not memory. What he lacked was a map.

Context: why the Bayonet became a headache

The King's Indian Defence in its Mar del Plata branch is one of the most ferocious systems in modern chess. After 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, the two sides enter a classic two-wing race: Black pushes ...f5, ...f4, then ...g5, ...Ng6, ...Rf7 and storms the kingside; White pushes c4-c5, b4-b5 and squeezes the queenside. Whoever arrives first wins. That is why the Mar del Plata has survived seven decades and still appears at the highest level.

The move 9.b4 — the Bayonet — was created to break Black's rhythm. Instead of preparing c4-c5 slowly, White grabs queenside space at once, threatens b4-b5 to drive the knight away, and, more importantly, delays ...f5 by exactly one tempo. In a race where every tempo is paid for in blood, one tempo is a great deal.

For years, Black's standard reply was 9...Nh5, heading for ...f5 through a series of forced sequences. Sharp, precise, and extremely easy to memorise — in the bad sense of the word. You do not need to understand the position; you only need to remember eighteen moves. Forget, and you die. Remember, and you can beat someone who understands the position better than you do.

That is the spatial piece the Bayonet creates: a narrow corridor where the better memoriser usually beats the deeper thinker.

9...a5: trading tactics for strategy

The DVD proposes 9...a5. At first glance it is a modest move to the point of irritation: a rook's pawn, no piece developed, no obvious threat, and one precious tempo spent in the ...f5 race.

But it strikes the sore spot. 9...a5 attacks the b4 pawn directly — the spine of the entire Bayonet idea. White must choose: push b5, lock the queenside and accept a long positional fight; or trade with bxa5, opening the a-file for Black's rook and dismantling with his own hands the anchor he has just built. Both choices lead to a position with far fewer forced lines than 9...Nh5.

Structurally, Black trades a development tempo for clarity on the queenside. Once that flank is settled, Black can commit fully to the kingside without worrying about b5-b6 or Ba3 hitting the d6 weakness. The race still happens, but it becomes less about who moves first and more about who understands more.

This is the most commendable thing about the product: it states plainly that the position is objectively equal. No promise of a forced win. No dream of White falling into a trap. It sells something harder to measure — practical difficulty for the opponent. As someone whose job is measurement, I respect that honesty; it substantially lowers the risk of overclaiming.

There is one detail I always check before trusting any opening recommendation: data. This material cites no figure at all — no practical score, no engine match rate, no reference games. That does not make the move wrong, but it turns advice into a hypothesis awaiting verification rather than a conclusion. In my line of work, a good hypothesis remains a hypothesis until the sample is large enough.

The contrarian view: three gaps in a good pitch

The first gap is in the slogan itself. The DVD asserts that computers often handle these complex strategic positions poorly. That was true around 2026. In the era of modern engines, Mar del Plata structures are evaluated at depths no human can reach. Read as a claim about machines, it has expired. Read as a claim about people — that amateur players lose their way in strategic mazes without a guide — it still stands. The boundary between those two readings is exactly where a buyer can be led astray.

The second gap is the clock. The entire appeal of 9...a5 is that you are allowed to think. But rapid, blitz, and especially Swiss-system games with short increments do not reward thinkers — they reward fast pattern recognition. A position demanding multi-move planning will punish you precisely when you need it most: when the clock is under ten minutes. In other words, the product promises less memorisation while quietly requiring a condition most club players do not have — time.

The third gap is packaging. The promotional title invokes Capablanca and a triumph in Buenos Aires 2026 — the Olympiad at which the Cuban legend took individual gold on board one in one of his last appearances at the international summit. A fine name, a fine memory. But it has nothing to do with the King's Indian Defence, and still less with the Mar del Plata. Capablanca in 2026 played the closed openings of his era, and never once in his life faced the Bayonet as we know it today. This is borrowed historical prestige used to sell contemporary technical content — an old advertising device, and a careful reader should separate the two layers.

One further point, purely technical: 9...a5 is not a new invention. It is a direction that already exists in practice, systematised. Its value is inversely proportional to how fast the fashion for the Bayonet itself turns over. When 9.b4 cools, so does the product.

At Vietnamese clubs, where opening preparation time is squeezed by dense tournament calendars and limited training conditions, the “few lines, easy to remember” pitch has very real appeal. I have watched hundreds of games by young domestic players and seen a recurring pattern: they win in the opening, draw in the middlegame, and lose at the moment they must make a decision that is not in the book. 9...a5, taught properly, attacks exactly that third weakness — but only if the teacher is willing to spend time on plans instead of lines.

So how should it be read

I still tell my students to look at it. Material that teaches you to form a plan rather than recall a sequence has genuine pedagogical value, even if you never play 9...a5 in your life. The lesson lies elsewhere: when a position has few forced lines, your mistakes surface more slowly and you have more time to correct them — but there is also nowhere left to hide behind “I forgot the theory.”

Tactics are only complete when told in a language the players dare to believe. And that belief, in chess as in football, does not come from a flashy headline about a legend in 2026. It comes from sitting at the board, understanding why the b4 pawn must be challenged, and deciding for yourself whether to push b5 or trade.

My seventeen-year-old student began relearning from move nine this week — not to memorise, but to redraw the map. The question I leave with him, and with anyone considering that DVD: if your opponent reads the same material tomorrow, what is left on the board?