ChessBase Magazine #225: The Receipt That Didn't Match the Invoice

I opened a file, and the name alone made me stop.

The file was labelled "Samarkand R8 (women): China lead with 15/16." The title is itself a claim, and not a small one. In the eighth round of a women's team event, China had taken 15 of a possible 16 points. In a team format that ratio is rare, because every round means four boards, four separate games, and a near-perfect score across all four means depth is working not just on board one but on board four. A score like that does not appear without rating depth. I do not publish such numbers without verification. So I opened the file.

Inside, nothing had anything to do with Samarkand. What I found was a product description for ChessBase Magazine #225. Analyses of selected games from Chess Festival Prague 2026 by Aravindh, Giri, Gurel and Navara. Twenty-seven miniatures. Ten opening articles. Three opening videos, presented by Werle, King and Ris. Alongside those, training modules spanning tactics, strategy and endgames. Four delivery routes: direct download, a PDF booklet, a postal booklet with a download key, and ChessBase Book for iPad, tablet and Mac.

The document's headline walks one way; the receipt inside walks the exact opposite way. Not a women's team event — a commercial publication. Not 15/16 — ten, twenty-seven, and three presenters. What is missing speaks loudest: no Samarkand, no China, no scorecard, no pairings.

This arithmetic of absence is not new to me. In May 2026 I published a four-thousand-word reconstruction of Spartak Moscow's title run. They scored 60 league goals; my shot-quality model said 47.3. I argued that the +12.7 overperformance was finishing luck, not structure. Two Russian Premier League analysts shared it the same day, and a television producer called it the spreadsheet that ended the eye test. My answer to every objection was one sentence: here is the code, here is the sample. I declined two panel invitations because I had not yet validated the model against a second season.

The first public xG model was not a prophecy. It was a receipt. From that week, every number I publish carries a method footnote — sample size, competition, season, model version. I abandoned match reports for model-first pieces that open with a data-source line rather than a scene. I have broken that rule exactly once, and I still apologise for it.

Context: what kind of object ChessBase Magazine actually is

ChessBase Magazine is a periodic training publication that bundles game-based analysis, opening articles, video and practice material into one package. Its raw material is live tournaments. Chess Festival Prague 2026 is the mine in this case: games are played there, then trimmed into training content later. The event ends, but its commercial life begins on the page.

This is where the first accounting should happen. In the chess content market, a single live game is sold at least twice: once to the organiser and broadcaster, and again to the reader and trainee. We can see the first sale — prize funds, broadcast rights. Nobody publishes the second. How many copies sell, how many readers finish, how many actually put a line into a repertoire — that denominator is absent, and that absence is precisely where my discomfort sits.

Four delivery routes are themselves a market segmentation. Direct download serves the digital-first buyer; the PDF booklet serves those who read on paper; the postal booklet with a download key serves those who want both; ChessBase Book for iPad, tablet and Mac serves the reader who does not want to open desktop software. One product, four buyer classes. Commercially smart. The question that follows is whether the quality of the receipt stays constant on every route.

Core: the economics of paid annotation, and its Bengali border

Opening theory and engine output are free today. The Lichess and Chess.com databases sit in every club player's pocket. YouTube offers thousands of free opening videos. Against that, what does a paid magazine actually sell? It does not sell analysis — analysis is no longer a product. It sells curation, editing, and time saved. What the buyer purchases is the removal of indecision: two structured hours in one place, instead of five opening videos scattered across ninety minutes.

The count of 27 miniatures matters here, because it is a measure of engagement, not of opening depth. Miniatures are usually selected for entertainment — so the reader gets the quick satisfaction of a win without the patience test of a long game. Then comes the denominator question: of those 27 games, how many genuinely contain a new opening idea, and how many are simply fast mistakes? The answer changes what the number 27 means.

The ratio of ten opening articles to three opening videos is also a signal. The video names — Werle, King, Ris — are labelled by ChessBase itself as opening-video presenters. In this context they are coaches, not competitors; and a coach and an analyst do different work — one teaches, the other testifies. Flatten the two into a single row and the product's promise stops being testable.

Now the Bengali border, because I keep that ledger year after year. Bangladesh has produced exactly five grandmasters. Niaz Murshed in 2026 — one year before Anand. Ziaur Rahman in 2026. Reefat Bin-Sattar in 2026. Abdullah Al Rakib in 2026. Enamul Hossain Razib in 2026. Then a long gap. Home norm tournaments effectively stopped after 2026, leaving a decade-long hole to 2026. The school pipeline went dark around 2026. The federation's annual budget is modest — a few lakh taka at most.

Against that denominator, another calculation belongs. When a federation runs on a few lakh taka a year, every decision to buy a foreign training package is really an answer to a different question: opportunity cost. A club player in Dhaka who buys this package could instead pay for several local coaching sessions, or a local tournament entry. I do not know how many copies this sells in Bangladesh. And that blank is my most valuable number — what I cannot claim without knowing is more reliable than what I could claim.

Contrarian: a name, a claim and a proof are not the same thing

"World-class players analyse their own games" — that sentence belongs to ChessBase. You cannot verify a product's quality from a product description, just as you cannot verify a person's character from their own autobiography. The standing of a contributor's name and the quality of an annotation are two separate variables; there is a relationship between them, not a causal chain.

There is a subtler trap. None of the opening-video presenters appear in this package as competing players, yet in a reader's mind those names land in the same compartment, and excess trust is born there. That is marketing, not deception; but as accounting it is name effect, not evidence.

The verification method for "new repertoire ideas" is straightforward, and ChessBase knows it. First check how often the line has been played in a database; then check the results; then check what the engine says; then check whether the line is still alive. Pass those four steps and the claim is theory. Fail them and the claim is advertising. The source gives no sample of that process, so I am placing the word "new" on the unverified list for now.

There is a fair explanation for the title-content mismatch: it is probably a scraping or metadata-translation error, not a fault inside ChessBase Magazine. But the user's loss is identical either way — someone looking for women's team results from Samarkand walks into a shop selling ten articles and 27 miniatures. A wrong title captures attention, and attention is the market's only real currency.

One more shadow enters this discussion, and I will not skip it. In the modern chess ecosystem, live data does not only travel to readers and trainees; some feeds move at the same speed to places where move-by-move information is converted into money. The player who buys a training package to improve is also having his own games transmitted quickly to betting operators' servers. The two purposes are not the same, and that fact appears in no training module's advertising.

Takeaway: what to watch in the next round

First, independent review. A product description is not the last word; the claims are only tested once a neutral reviewer publishes in detail.

Second, independent game sources from Chess Festival Prague 2026. Once specific games surface, the opening claims can be matched against a database.

Third, the actual results of the Samarkand women's event. If FIDE or a reliable chess outlet confirms a score like 15/16, the mismatch resolves; if not, it is a metadata crisis.

Fourth, the structure and compatibility of ChessBase Book. A change in licensing or device support across those four delivery routes changes the product's value.

Fifth, database verification of the opening trends. Whether what named players call a trend is actually rising in frequency at elite events, or is just a good story.

What this data cannot see

It cannot see whether a buyer who reads this package actually gains rating — there is no sample for that test here. It cannot see what alternatives that buyer had before purchase. It cannot see whether a miniature teaches a reader an opening or merely entertains them. And above all, it can say not one word about the Samarkand women's team event, because every number here belongs to a different subject.

So where is the next-round signal? I built the Russian Premier League model because the silence was measurable. The same applies here. Nobody is keeping the ledger in the chess content market: how many coaches buy which issue, how many readers change a repertoire after reading. Until that ledger exists, any claim — praise or blame alike — is an invoice without a receipt.