The move 9...a5, the Bayonet Variation and a Samarkand headline: two chess stories spliced into one report

Opening: three in the morning, sound off

At three in the morning I mute the commentary and let the board speak for itself. On screen is the position after White's eighth move in a King's Indian Defence, Mar del Plata branch. White has the pawn chain on d5 and e4. Black has e5 and d6. The two wings are almost entirely separate, and between them sits an empty space that both sides know will never be filled by a pawn again. I have watched this game four times tonight. No commentator, no crowd noise, just the click of the mouse between moves.

Then the morning inbox brings a document that makes me sit up straight.

Its headline is about a team event in Samarkand, round two, the open section, and how the strongest teams all won despite some near-draws. But the entire body of the text - fourteen information points, not one line missing - is about an opening DVD promoting the move 9...a5 against the Bayonet Variation of the King's Indian Defence.

Two different texts, sharing one headline.

I am not writing this to catch out an editorial error. I am writing because that mismatched headline accidentally reveals a far more interesting story: the story of a line everyone assumed had retreated into the past, still quietly alive in training rooms, camps, and DVD products sold to club players.

They say girls know nothing about tactics. So I write for them to read.

Context: the Mar del Plata board and why it still keeps people awake

The King's Indian Defence begins with four moves: 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6. Black has not occupied a single central square. Black lets White build a massive pawn centre, and only then comes back to attack that very centre. This is the hypermodern spirit of the Indian school: let the opponent take space, then turn that space into a burden.

Then comes 5.Nf3 0-0 6.Be2 e5 7.0-0 Nc6. Both sides lock the centre with calculated pawn pushes: 8.d5 creates tension, and 8...Ne7 forces White to decide immediately. Black has built the structure. White has space, a theoretically open d-file, and a favourable queenside. Black has a solid structure, the g7 and potential c6 bishops, and a kingside to attack on.

This is the Mar del Plata, named after the Argentine coastal city where it was honed in tournaments through the 1950s. The ECO codes usually used to classify this branch fall between E97 and E99 - that classification is mine, not present in the source text, and I say so plainly so readers do not mistake it for verified source data.

What separates the Mar del Plata from the rest of the King's Indian is speed. No move here is a waiting move. Both sides are racing on opposite wings, and the loser is usually the one a tempo behind. White pushes queenside pawns, opens the c-file, attacks d6. Black pushes kingside pawns, opens the f-file, attacks e4 and d5, then advances to f4, g5, g4.

The game therefore always has the shape of a race. Not a pure sprint, but a race of timing. Someone has to choose the right moment to stop defending and start attacking. Choose one tempo wrong, and your wing collapses first.

Based on my experience following King's Indian games at major events, I notice something rarely discussed: most club players do not lose in this line because they calculate badly. They lose because they do not know which phase of the race they are in. They push the right pawn in the right direction at the wrong moment.

The Bayonet: the move 9.b4 and a forgotten revolution

If the Mar del Plata is a race, then 9.b4 is White declaring he will run first.

The move 9.b4 is named the Bayonet Variation. Instead of opening the c-file slowly, White pushes b4 at once, threatening b5 to drive off the c6 knight - the piece holding Black's centre - and, more importantly, preparing a queenside pawn attack Black cannot ignore.

This variation is not an engine-era invention. It appeared fairly early in King's Indian history, but only became a genuine nightmare for Black when Vladimir Kramnik brought it into his arsenal in the 1990s. Kramnik - regarded as a strategist rather than an attacker - chose an aggressive pawn push like 9.b4. That is the most interesting paradox of the line.

When Kramnik and Garry Kasparov met at their peak, the Bayonet became one of the most important weapons in King's Indian games. Kasparov, who played the King's Indian more fluently than anyone of his era, was forced to keep searching for antidotes.

The main antidote, for decades, was 9...Nh5.

This move brings the knight from e7 to h5, prepares ...f5 and then ...f4 and, more importantly, places the knight on a square White struggles to challenge. White must reckon with Black pushing kingside pawns faster than White attacks on the queenside. After 9...Nh5 the position becomes sharp, concrete and highly forcing. Each side has few choices. Every move can be checked against variations.

This is the key point readers need to grasp: 9...Nh5 is not a defensive move. It is a move that opens the game into concrete forced lines, where whoever remembers more and calculates more precisely wins.

And precisely for that reason, in the engine era the Bayonet became the line club players fear at every level. When every forced line can be looked up by your opponent, a forcing branch means White can prepare to move thirty.

That is why 9...a5 suddenly becomes attractive.

Core analysis: the geometry of the two replies

I want to draw two diagrams on the board so readers can see the difference clearly. No notation needed - just the shape.

Diagram one - after 9...Nh5.

The black knight stands on h5. The f-file will open. A black pawn on f5 then f4 splits White's kingside. The g6 pawn advances to g5, g4. White must react on the queenside with c5, b5, Nb3, Ba3 and precise moves. The centre is closed, but the diagonals around it open up. The feature of this diagram is that everything is linear. Every pawn push by one side has a corresponding answer from the other. It is a dialogue, not a free conversation.

Diagram two - after 9...a5.

The black pawn on a5 attacks White's b4 pawn head-on. White must decide: push b5 to keep the structure, trade on a5, or ignore it and keep attacking the queenside. All three choices lead to different positions, but they share one thing: they are all slower, less forcing, less bound by forced lines. When White pushes b5, Black gets a square for a knight on c5 or e8, and the queenside pawn structure becomes more fixed. The race continues, but the speed drops. Both sides must play positional chess.

That is the entire value of the DVD product I received: changing the character of the game from forcing to strategic, from memorisation to independent thinking.

I must state this plainly: 9...a5 is not a new move. It is a well-known secondary branch, present in databases, analysed in chess magazines. The product does not claim to invent anything. It claims a more pragmatic approach.

The gap between that claim and reality lies here: a forcing branch can be checked by machine. A strategic branch cannot, or at least not easily. And that verification gap is exactly the space the product occupies.

In other words, the value of 9...a5 is not that it is stronger than 9...Nh5. Its value is that it is harder to refute with a single concrete line.

In the original analysis I have in hand, the "engine match rate" and "average centipawn loss" fields are both marked as insufficient information. No sample games are cited. No engine evaluation. No win rate, draw rate, or time-control data.

With a sports article, that is the first thing I check: is evidence offered anywhere, or only promises?

Here, they offer promises.

The engine era and the shrinking of the King's Indian

To understand why such a product still finds buyers, you need to understand why the King's Indian has shrunk at the elite level.

Over the past two decades, the King's Indian's share of elite classical games has fallen noticeably. Top players moved to openings with lower theoretical risk: the Grünfeld, the Nimzo-Indian, and further down the Italian or Spanish systems.

The reason is not that the King's Indian got weaker. The reason is risk structure. In an opening where Black concedes the centre and accepts a breached kingside if he miscalculates, one small error can lose instantly. When every line has been dissected by engines, elite players understand they cannot "play on inspiration" in this branch.

But here is the point I want my female readers to remember: an opening pushed out of the top tier does not die. It changes tier.

The King's Indian thrives at club level, in junior events, in training camps, where calculation is not tested to move thirty and where understanding structure matters more than remembering variations.

At the elite level, it survives as a surprise weapon. Specialists in this line - including young Uzbek players, Nordic players raised in the Larsen tradition, and Russian players who always keep it in stock - still deploy it when they need a game off script. Teimour Radjabov, Hikaru Nakamura and a newer generation have all used it as a weapon at a specific moment.

That is the real market for the DVD: not super grandmasters, but club players who want a weapon.

Samarkand and Uzbek chess: the story the headline abandoned

One thing I regret when reading that mismatched report: Samarkand genuinely has a chess story worth writing.

Uzbekistan won the 2026 Chess Olympiad team event in Chennai, India. The line-up of Nodirbek Abdusattorov, Javokhir Sindarov, Nodirbek Yakubboev, Shamsiddin Vokhidov and Jakhongir Vakhidov beat teams rated higher to take gold - one of the most shocking results in this century of team chess.

Before that, in 2026, Abdusattorov - then 17 - won the World Rapid Championship in Warsaw, Poland, becoming the youngest rapid world champion in history. That is a verifiable fact with real meaning: a Central Asian nation once inside the Soviet chess system has been reborn as an independent chess power.

Samarkand, the historic Silk Road city, hosted the World Rapid and Blitz Championships in December 2026. That event put Uzbek chess on the international media map at a new scale.

A report about Samarkand should tell that story. A report about an opening DVD should tell the story of 9...a5. The two texts sharing a headline is not a media catastrophe. But it is a reminder: in an age of automated news aggregation, source quality becomes the first line of defence, and the easiest one to skip.

I once followed Uzbek junior events via online streams. What I remember is not the wins. What I remember is how the children sat still at the board, not looking at the audience, not looking at the camera. There was a focus there I recognised, because I once sat like that.

The product: reading advertising like an analyst

Now let us read the product like an analyst.

The product positions 9...a5 as a practical antidote to the Bayonet. It promises to shift the game from concrete lines to a strategic struggle with less memorisation. It claims both sides have equal chances, and claims that if Black knows what he is doing, things can quickly become dangerous for White.

These three claims have three different levels of verifiability.

The first - that 9...a5 is a legitimate line - is fully verifiable and fully true. The move has been in chess theory for a long time.

The second - that the game shifts to a strategic, low-memorisation struggle - is partly verifiable. It is true that 9...a5 is less forcing than 9...Nh5. But "less forcing" does not mean "not forcing". The Mar del Plata is still the Mar del Plata. White still attacks the queenside. Black still has to push kingside pawns. The structure is still narrow, and an error still costs.

The third - that things can quickly become dangerous for White - cannot be verified from the text. It is a directional claim, true in many positions and false in many others, depending entirely on White's level of preparation.

I do not object to selling an opening product. These products are a healthy part of the chess ecosystem. They give club players a study path instead of drowning them in thousands of database lines.

What I object to is using absolute language for claims with no data behind them.

And there is one more claim in the text that made me pause longer than any other: the suggestion that computers often do not know how to handle these complex strategic positions.

That is the language of a bygone era.

The contrarian angle: engines understood the Mar del Plata long ago

I want to use this section to state plainly something many in chess know but rarely write.

The claim that "computers do not understand complex strategic positions" is a legacy of the pre-AlphaZero era. In that era, search-and-evaluation engines genuinely struggled with positions whose value lay in long-term structure rather than immediate material.

But that era is over.

Neural-network engines today - since reinforcement learning and neural networks were applied to chess - handle long-term strategic positions far better than the previous generation. They do not need humans to teach them structure. They discover evaluations humans took decades to accept.

So when an opening product still uses "computers do not understand this line" as a selling point, it is speaking to an old memory, not to current reality.

This matters for two reasons.

First, it underestimates the buyer. Today's club player has access to tools stronger than any grandmaster of the 1990s. They do not need reassurance from an expired argument.

Second, it blurs the product's real value. The real value of 9...a5 is not that computers fail to understand it. The real value is that it forces the opponent to think with his own head instead of reciting a memorised line.

That is a completely reasonable value. It does not need a false argument to sell.

There is one more blind spot worth naming. Even if 9...a5 succeeds practically, it does not create a permanent wall. In opening theory, every antidote breeds a counter-antidote. Once a sideline becomes popular enough to prepare for, White finds a plan, and the cycle repeats.

Tactics never lie; only the people reading them get it wrong.

The blind spot of source integrity

There is a problem I have not yet mentioned, and it is more serious than the technical claims.

The headline refers to a team event in Samarkand, round two, open section. The body refers to an opening DVD. No tournament, round, team, pairing or result appears anywhere in the body.

For anyone doing editorial work, this is a red flag at the highest level. It means the text was spliced somewhere - possibly in automated aggregation, possibly in archiving, possibly in format conversion. Anyone citing this text risks citing the wrong source.

I raise this not to make a fuss. I raise it because in a context where sports news is increasingly machine-aggregated, readers deserve to know when their source is unreliable.

A good sports article answers three questions: what happened, who is involved, and where is the evidence. This text answers the third with a question mark.

What is telling is that even ignoring the headline, the DVD content still lacks basics. No author name. No title. No competitive record in this line. No sample games. No engine evaluation.

An opening product with no named author is a product that cannot be assessed. The buyer does not know who they are learning from. That matters, because in chess the analyst's credibility matters nearly as much as the analysis - not because credibility makes a move correct, but because it tells the student how much verification is needed.

What to watch

If I had to set a watchlist for this line over the coming months, I would list four signals.

First: the frequency of the Bayonet in elite classical games. If 9.b4 stays popular, demand for replies stays, and the opening-product market stays alive.

Second: the appearance of 9...a5 in a classical game with a rating above 2600. One such game would not prove the line is stronger. It would only prove it deserves preparation. For an opening product, that is enough.

Third: White's responses. When someone re-analyses 9...a5 with a modern engine and publishes a plan for White, the next cycle begins.

Fourth: the quality of the source itself. Whether the Samarkand headline is corrected, or the text continues to exist in its spliced form. Either way, it is an indicator of the information infrastructure of whoever published it.

A note on gender and on reading

I want to say one thing, once, and then stop.

Over the years I have written about chess, I have received more comments about my gender than about my content. People ask whether I really play. People ask whether I really understand the King's Indian. People ask whether I am just copying from books.

Judit Polgar beat every leading player of her generation, and people still called her an exception. Hou Yifan stood among the world's best, and people still called her a special case. The language of "exception" and "special case" is a way of not changing the original assumption.

For me, reading a board has no gender. Seeing a bishop triangle on the queenside has no gender. Knowing that a b4 pawn forces Black to answer within three moves has no gender.

The only thing with gender is who is allowed to say what they see.

I write this for anyone who has been told they have no right to an opinion about a position. Not so they believe me. So they have one example of how to test a claim.

Sometimes, to understand a game, you have to stand in silence longer than others can bear.

Reference data table

| Item | Finding | Status | |---|---|---| | Opening | King's Indian Defence, Mar del Plata, Bayonet Variation | Confirmed | | White's main move | 9.b4 | Confirmed | | Black's main reply | 9...Nh5 | Confirmed | | Promoted move | 9...a5 | Known sideline, not a novelty | | Estimated ECO | E97 to E99 | Analyst's classification, not in source | | Engine match rate | None | Insufficient data | | Average centipawn loss | None | Insufficient data | | Sample games | None | Insufficient data | | Product author | Not named | Cannot be assessed | | Source headline | Samarkand, round two, open | Does not match body | | Source body | Opening DVD product | Independent of headline | | Overall risk level | Medium | Due to unsupported claims and source error |

Glossary of terms used

King's Indian Defence is a hypermodern opening in which Black concedes the centre to White early and then counterattacks that centre with pawn pushes and pressure on d4.

Mar del Plata variation is the sharpest King's Indian branch, with ...Nc6, ...e5, ...Ne7 and a wing-versus-wing confrontation: White attacks the queenside, Black attacks the kingside.

Bayonet Variation is the move 9.b4 by White, gaining queenside space and threatening to drive the black knight from the centre.

9...Nh5 and 9...a5 are Black's two replies to the Bayonet. The first is forcing and constitutes the main line. The second is strategic and constitutes a sideline.

Average centipawn loss is a metric measuring the average deviation of moves from the engine's best choice. It does not appear in the source text.

Neural-network engine is the generation of chess engines using reinforcement learning and neural networks, handling long-term strategic positions far better than earlier search-based engines.

Fritztrainer-style opening product is a DVD or training file format common in the German-language chess market, usually containing video analysis and an accompanying database file.

Sources and verification notes

This article draws on an internal analysis package dated 13 August 2026, plus public chess facts about the 2026 team Olympiad in Chennai and the 2026 World Rapid and Blitz Championships in Samarkand.

Source verification status: headline and body do not match. The body is identified as promotional text for an opening product. The headline is identified as unverified descriptive metadata.

The article makes no purchase recommendation, offers no result prediction, and uses no internal data that cannot be externally verified.

Closing: what to carry away

If there is one thing I want readers to carry from this piece, it is how to tell a verifiable claim from one that is merely stated loudly.

The move 9...a5 exists. It is legitimate. It can be used. There is nothing wrong with learning it. But its value does not lie in being advertised as an antidote. Its value lies in opening a different kind of game, where the player has to think for himself, and where the skill of reading a position matters more than memory.

That is a very good value. It does not need exaggeration.

And if you have been told you have no right to read a King's Indian position, try doing exactly one thing: set the board back to move nine, pick a side, and ask yourself where the next move goes. You will find you read more than others assumed.

The board is wider when there is no noise. I see it more clearly than ever.