9...a5 Against the Bayonet: A Promotion, an Empty Column, and a Headline That Needs Checking

Last year, sitting at a desk in Mumbai, I built a habit: measuring the distance between a headline and its body. I have watched the game for 45 years, built xG and PPDA models on football desks for 26, and then turned back to chess score-sheets and archives. Still, the gap between a headline and a body stops me every time. A report recently came to my hands. Its headline says that in Kazakhstan, President Kassym-Jomart Tokayev and the so-called "FIDE President" Timur Turlov held a high-level meeting. Opening the piece, I found inside it the promotion of a chess DVD: against White's Bayonet Attack 9.b4 in the King's Indian Defense, Mar del Plata variation, Black's antidote 9...a5.

I drew two empty boxes at once. The first box for the headline's claim: zero sources, zero dates, zero verification. The second box for the technical claim: zero engine data, zero ACPL, zero sample size. When two empty boxes sit side by side, what happens is what happened in my ledger — the empty column itself is the primary evidence. The spreadsheet does not argue; it waits for the next correction. This piece is the accounting of that wait, because here the story left for the wrong address before it ever reached the board.

The King's Indian Defense cannot be explained in one line, but what is needed here is clear. After d6-I...Nf6, Black builds a hypermodern wall with ...g6, ...Bg7 and ...O-O; he lets White build a big centre and then strikes with ...e5 or ...c5. In the Mar del Plata branch, Black plays ...e5 to lock the centre, White pushes to c4, and attacks begin on opposite wings — White on the queenside, Black on the kingside. The Bayonet Attack 9.b4 is White's most direct decision: pushing the queenside pawn forward, he shakes the base of Black's ...e5, presses on c5, and pushes Black's natural plan toward defence.

From here the story narrows. Black's conventional answer is 9...Nh5 — the knight goes to h5 to secure ...f5, the game opens up, becomes concrete, becomes forced, fills with engine-memorised lines. The DVD says there is an alternative: 9...a5. The a-file pawn advances to question White's b4, closes the queenside, and pulls the game from tactical obligation into strategic struggle. The author claims there are very few concrete lines here, that engines often cannot handle such positions, that with perfect play White keeps equal chances, but a prepared Black becomes dangerous.

The technical object is therefore clear: against White's Bayonet, Black chooses a strategic struggle via 9...a5 instead of the forced 9...Nh5. There is no claim of a brand-new novelty, no grand theoretical announcement; there is a rearrangement of a known practical resource, an attempt to defeat White's memory bank. But exactly here my ledger's first warning lights up: the clearer the object, the more absent its proof.

I searched on five levels. No win-draw rates. No high-level sample games. No frequency of how often 9...a5 is played against 9.b4 in databases. No engine evaluation, no ACPL, no move-number distribution. The author only says engines cannot handle these positions. Modern engines do evaluate positions objectively; sometimes they fail to capture human practical difficulty — that argument is partly right. But not a single example is given for it. The claim is a marketing argument, and the evidence is absent — the distance between the two is the real event to me.

Here my own ledger comes in useful. In 2026, at 52, after 26 years building xG models on football desks, I spent eleven months digitising 1,240 games from 41 Bangladeshi National Championships (2026–2026) with a second-hand scanner, digging through Daily Star, Ittefaq and BSS clippings. I ran every game through an engine and logged average centipawn loss. Niaz Murshed's 2026 games averaged 28 ACPL; the 2026 champion field averaged 47. I sent that nineteen-point gap to three Dhaka sports desks. All three returned it. That rejected pitch still sits on the first page of my ledger, because an unpublished finding still has to be correct.

That habit is what now teaches me to ask: who is measuring? If this DVD claims 9...a5 is "sound" and "few concrete lines", in which column does that claim sit? My ledger has a column called "positions won, points banked". In September 2026 I covered the 43rd Chess Olympiad in Batumi from a Mumbai desk, logging all 44 Bangladesh games board by board while the wires printed only results. Filtering for positions where the engine gave +1.00 or better, I counted 13. Bangladesh converted 3 — 23 percent. I also found that 9 of 11 losses came after move 35, in time trouble.

That habit is what teaches me that in Batumi, Bangladesh usually did not lose on move ten; it lost after move thirty. The same question applies to the DVD: with 9...a5 Black may survive move ten, but will it survive move fifty? The strategic King's Indian structure is not actually safe. When Black later plays ...f7-f5 and attacks come from both wings, the board becomes tactical again. A "strategic" line does not mean a quiet line; it means the risk is delayed into time pressure, not erased. My ledger holds many precedents — structures that looked safe on paper collapsed in the last ten minutes on the clock.

Now to the commercial positioning, because an absence of data and a presence in the market, read together, form a definite picture. An antidote DVD on 9...a5 appears exactly when White's 9.b4 becomes a real practical problem for Black players. The target audience is probably club-level and advanced amateur King's Indian players — those who want a humanly playable repertoire without memorising engines. At elite level nobody builds a repertoire on the promise of "few concrete lines"; there, preparation means specific depth. So the DVD's value sits between two audiences: one wants less memory, the other wants less risk.

Here a column I named long ago comes to mind: "platform-adjusted". In 2026, as boards were packed away, Leonine Chess Club moved its events online. For six weeks I refused to treat any online game as data — mouse slips, no touch-move rule, unverifiable settings. Then I built a time-per-move model on 412 games from Leonine's COVID-era events and found the error rate roughly 2.3 times that of over-the-board rapid play. At the same time I was setting the BCF's annual government grant — about Tk 11 lakh — beside the online blitz prize money Ziaur Rahman was chasing. The numbers did not recognise each other.

For this DVD I ask the same thing: is this proof of preparation, or advertising for preparation? As the online boom is no substitute for Olympiad norms, so the advertising of a strategic line is no substitute for conversion on the board. Participation and conversion are two separate columns, and most promotion leaves the second one empty. If the DVD only says "play this, engines won't understand" but never measures the user's real results, then to me it is not game information but market information.

This is where the confusion is created. The piece points toward engine-neutrality, as if a strategic structure meant fewer errors. But safety and strategy are not the same. A line can be less concrete and still be dangerous in the wrong opponent's hands; and in a prepared opponent's hands, any "safe" line becomes a path to mere surrender. In my ledger of 1,240 games I have seen again and again: slow, strategic structures often win in the wrong place and lose in the right one, because humans forget the structure and trust the rhythm.

The headline side is not unimportant here. The so-called "FIDE President" title needs verification. By publicly available official records, the FIDE President is Arkady Dvorkovich; Timur Turlov is better known as an administrator and patron of chess. So there are two possibilities: either the title is wrong, or a different FIDE post has been mislabelled. Either way, this is a factual flag, not player data. When a headline puts a head of state and a chess DVD in the same sentence, the reader should trust the body, not the headline — and the body demands verification.

Now to the part where my strongest objection lies. The claim is that 9...a5 gives "sound" positions and "very few concrete lines". But there is no sample. In how many games has this line been played? At what level? In what percentage did Black stand equal or better? Without answers to these questions, the word "sound" is empty to me. To call a line sound I need three things: a sample, an engine evaluation, and an accounting of the human-versus-engine gap. The DVD has none of the three.

And here the difference between correlation and causation becomes urgent. If 9...a5 gives good results in some games, it does not follow that the line is the cause of the results. Those games may have been played against opponents who could not handle the pressure of 9.b4; in those games Black may have been stronger than White; the result may have come from middlegame skill, not the beauty of the opening. In my ledger this error returns again and again: the opening is credited for the middlegame's work.

Similarly, another trap in the DVD's claim is the phrase "engines won't understand". Modern engines evaluate every position; they are objective. The problem is that their evaluation does not capture all of human practical difficulty. But that is not the engine's incapacity; it is a gap in human preparation. The DVD uses this gap in its own favour, as if the incapacity were the engine's, while the responsibility lies with the preparer. To me this is the same old trick that conflates online ratings with Olympiad ratings.

One thing needs clearing up: denying this DVD's value is not my aim. The opposite. A human-playable repertoire is a real need for club-level players, and 9...a5 may be a reasonable answer to that need. My objection is not to the product but to the absence of proof. Any claim of a new antidote gains weight only when an open column sits beside it — sample, frequency, conversion rate.

A sentence returns again and again in my ledger: I digitised the past so the present could not lie to me. The same work is needed here. How fast is 9.b4 usage growing, in which tournaments, at which levels; and how the ratio of 9...Nh5 versus 9...a5 as Black's answer is changing over time — with those two rows of numbers, the story would change. Without them we get only one author's confidence, not a repertoire.

My own experience says where such gaps come from. During the 2026 hiatus I stopped counting online games for six weeks; by the end, that silence had itself become a dataset. Likewise, this DVD's missing data is itself data — it tells us the market builds the product before the proof. And that sequence is the warning sign for me.

Another angle deserves thought: who buys such an antidote? Those tired of engine-memorised preparation, those short on time, those who want to survive at the board rather than memorise. For them 9...a5 may truly be a relief — fewer lines, less memory, more planning. But relief and strength are not the same. If a line does not force White, it does not force Black either; rather it pushes both sides toward uncertainty. Uncertainty is an opportunity for the experienced Black and a trap for the inexperienced one. So it matters who the DVD's real user is — a club player or a tournament player.

There is one more layer I have noticed for a long time. South Asia's chess market has plenty of demand for training material but little culture of measurement. Before buying, we do not ask — in how many games has this line been tested? Who tested it? At what rating level? If the DVD answered these questions, it would become not a product but a research paper. For now it is only a promise.

I know this critique will sound harsh to many. But my 45-year habit is one: a number that does not exist cannot be named. 9...a5 may be sound, may be dangerous, may be both — but on the author's claim we cannot know. We only know a DVD is on the market, a headline speaks of a meeting in Kazakhstan, and between them a column lies empty.

One last thing. Over the past decade the chess-training market has changed, online platforms have changed, but the standard of proof has not. A line becomes an "antidote" only when the opponent's practical results turn against it — and proving that needs year-by-year frequency and conversion rates. In my ledger that work is still unfinished.

So the next time this line appears on a board at an elite tournament, I will count three things: how often 9...a5 is played, in how many of those Black survives past move 40, and in how many of those Black actually banks the point. When the numbers arrive, the story will write its own correction. The spreadsheet does not argue; it only waits — and the empty box beside this headline is still part of that wait.