Not Samarkand, But ChessBase Magazine #225 — A Metadata Audit Ledger

I opened the file expecting exactly one thing: round eight of the women's team event in Samarkand, China leading 15/16. I scrolled to the end. Samarkand never appeared. Not China, not the women's teams, not the score. What appeared instead were ten information points, every one of them about ChessBase Magazine #225 and the Chess Festival Prague 2026.

So I ran the tally. Event description: zero. Standings: zero. Ratings: zero. Head-to-head: zero. In their place: 27 miniatures, 10 opening articles, three opening videos, tactics-strategy-endgame training material, and four delivery routes — direct download, PDF booklet, a booklet posted with a download key, and ChessBase Book for iPad, tablet and Mac.

That gap between title and body is the subject here. It is not a typo. It is a picture of how chess news is manufactured, who writes it, and who verifies it — or does not.

I digitized 41 national championships and found the 19-point ACPL gap staring back. In 2026, at 52, after twenty-six years building xG and PPDA models on football desks in Mumbai, I spent eleven months with a second-hand scanner pulling 1,240 games from 41 Bangladeshi National Championships, 2026 to 2026, working through Daily Star, Ittefaq and BSS clippings. Every game went through an engine. Niaz Murshed's 2026 games averaged 28 centipawns of loss; the 2026 title field averaged 47. I pitched the nineteen-point gap to three Dhaka sports desks. All three passed. The rejected pitch still sits at the front of the ledger, because an unpublished finding still has to be correct.

A year later, in September 2026, I covered the 43rd Chess Olympiad from a Mumbai desk. The wires carried results; I logged all 44 Bangladesh men's games board by board. In Batumi, 44 games turned into 13 winning positions and only 3 conversions — 23 percent. Nine of the team's eleven losses arrived after move 35, in time trouble. Conversion rate has been my signature metric ever since.

In 2026 the boards went into storage and Leonine Chess Club moved online. For six weeks I refused to count online games; the silence became a dataset. Then I built a time-per-move model on 412 of Leonine's COVID-era online games and found an error rate roughly 2.3 times that of over-the-board rapid play. I set that beside the federation's annual government grant — about Tk 11 lakh — and the online blitz prize money Ziaur Rahman was chasing.

I ran that same method against this Samarkand file. I cross-reference the tape, the box score, and the silence between them.

One information point names the analysts: Aravindh, Giri, Gurel, Navara. Another names three opening-video presenters: Werle, King, Ris. Another lists 27 miniatures. Another lists 10 opening articles. Further points describe the training material and the publisher's own quality claims. The last two cover distribution. Samarkand: zero. China: zero. Women's teams: zero. 15/16: zero.

This mismatch is not a results crisis; it is an information-flow crisis. After more than a decade of writing from data, I know a product brochure and a tournament report cannot live in the same file. One asserts. The other proves.

Look at the names too. Giri and Navara are elite annotators, which requires no explanation. The presence of Aravindh and Gurel is an editorial-market signal — a calculation aimed at Indian and Turkish chess audiences that are growing fast. That is a distribution record, not a rating statement. I can read it from India, where the content market expands quickly and the norm pipeline does not.

Werle, King and Ris are selected for opening videos, so their role is teaching rather than competition. Miniatures are short decisive games, usually chosen for entertainment and tactical instruction rather than deep theory. That is inference, since I hold no sample pages, but the packaging tells it plainly.

Chess Festival Prague 2026 is the raw material. Games are lifted from a festival, annotated, filmed, and repackaged into sellable training content — that event-to-content pipeline is the real business. Direct download, PDF booklet, posted booklet, ChessBase Book: four doors for three kinds of buyer — digital-first, tablet-led, and paper-only traditionalists. I belong to the last group.

What is absent must be recorded as absent. No games, no moves, no engine evaluations, no standings, no ratings. So ACPL cannot be computed and 'new repertoire ideas' cannot be tested. To call an opening idea new you need three things — a concrete move sequence, its absence from databases, and the engine's response. Promotional copy contains none of them.

The spreadsheet does not argue; it waits for the next correction. Here the correction belongs to us, not the publisher. When free engine output, YouTube opening videos and streaming deliver thousands of hours of chess daily, the case for paying for a grandmaster's handwritten annotation cannot rest on the sentence 'first-class training material.' Human annotation, sequenced curriculum and durable archive are the differences an engine cannot manufacture. Those must be proven, not announced.

Now the other side. Hunting for conspiracy here is foolish. The likeliest cause is scraping, translation or structured-metadata error — a title and a body joined from two different sources. The damage is not deception but confusion: a reader who opened the file for a women's round-eight score received a product page.

I stop there. I have entered national championships, Olympiad score-sheets and online silences into this ledger across five continents, and I keep one rule above all: I say what I do not know. I do not know China's score in Samarkand from this file. Pretending otherwise would put the credibility of the whole ledger at risk.

Second counterpoint, about chess media itself. We have grown used to printing a publisher's own copy as news. Verification then becomes volunteer work, and volunteers are not always available. If no independent review appears, 'world-class players analyse their games' remains a sentence rather than evidence. So: let every claim be published next to the name of the claimant.

What do I watch next? Verify the Samarkand women's event from FIDE or official result sources, separately from this file, because this file does not contain it. When specific Prague 2026 games surface in samples or databases, test the opening articles move by move. And until an independent review lands, treat 27 miniatures and 10 articles as a training bundle, not as theoretical progress.

One closing instruction rather than a summary: if anyone erases the wall between the Samarkand score and the ChessBase product, the entire chess-media ledger needs re-auditing. I digitized 41 national championships on the belief that old records do not lie. Truthfully, records say nothing at all — what we write is what the record says.