ChessBase Magazine #225: 27 Miniatures, 10 Opening Articles, and the Limits of the Chess Content Industry

Three in the morning in Delhi, the ceiling fan turning like a broken clock, and in the inbox a subject line: "All new releases in September!". I opened it. Inside was the blurb for ChessBase Magazine issue 225 — seventeen information points, not a single game named, not a single move, not a single engine evaluation. A product advertisement, written exactly the way product advertisements are written. Journalism does not read like that.

What woke me up was not seventeen points or seventeen thousand. It was 27. Twenty-seven short games, bundled into a section called Special, with the words "highly entertaining" attached.

I have sat with short games for forty years, in several forms: at tournament tables in Hanoi in the eighties, on VTC broadcast tapes in the nineties, on a slow-motion screen in Delhi now. And I learned one thing about short games: they are the cheapest content unit in the entire chess industry, and also the easiest to sell.

The question I asked did not sound like a woman, but the answer they dodged did not sound like a man either. This blurb dodges one very specific question: among those 27 short games, is there a single game never previously published, or are they all games that have been sitting in the database anyone with a current ChessBase copy already downloaded months ago?

Mistakes are not frightening; bending down at two in the morning to pick up the tape is what makes a forecaster. I picked up this blurb and read it the way one reads minutes of a meeting.

My thesis fits in one sentence, and I know it will annoy people inside the industry: ChessBase Magazine 225 does not sell new theory. It sells the labour of filtering. The whole architecture — 10 opening articles, 27 miniatures, three video series, one tournament annotated — is the same act repeated in different formats: selecting from a sea of existing data, then selling the selection back. That is an honourable trade, even a hard one. It is not the trade of invention, and the buyer deserves to know which one he is buying.

Seventeen points and one hole

Before dissecting, let me reconstruct what the product says about itself. Issue 225 contains: annotation of the tournament in Prague, with four named contributors — Aravindh Chithambaram, Anish Giri, Gurel and David Navara; a Special section holding 27 short games; 10 articles on openings; opening videos presented by Jan Werle, Daniel King and Robert Ris; and training material in tactics, strategy and endgames.

Delivery runs on three parallel channels: a download with a PDF booklet; a printed booklet with a download key; and ChessBase Books for iPad, tablet and Mac.

That is the entire hard evidence. Seventeen points, all supplied by the seller. No game list. No scorelines. No average centipawn loss. No claim of a single move never played before.

For outsiders, the word miniature may mean nothing. To be clear: a miniature is a short decisive game, usually under 25 moves, decided by a concrete tactical error rather than accumulated positional pressure. Its training value lies in pattern recognition — a club player sees a motif, stores it, reuses it. Its editorial value lies somewhere else, and I will get to that.

A repertoire is the opening system a player chooses for the long term, with pruned branches and fallbacks and a trade-off between complexity and memorability. A novelty is a move never previously published in a recorded game, at a specific date, with engine assessment and usually a publication timestamp. These three concepts differ in kind, not in degree. A publication can be full of repertoire ideas and contain no novelty at all. Issue 225, judged on its own words, is the first kind.

Twenty-seven miniatures and the economics of annotation

Now to the part nobody in the industry likes said out loud.

Annotating a game is labour. A good annotator rebuilds the variation tree, checks with an engine, writes commentary at every turning point, and then faces the hardest question: what does the reader learn at move fourteen. A 60-move game can consume two serious working days. A 20-move game ending in a clean tactical stroke consumes a few hours.

Cost is not the whole story. A short game also has a property long games lack: it counts as a headline. Twenty-seven. That is a marketable number, and it is true.

A 20-move game is cheaper to annotate than a 60-move game, but it is richer in pattern value for the learner — and that trade-off, rather than editorial inspiration, decides the shape of the Special. I am not saying this to diminish the product. I am saying it so the buyer knows what kind of goods are in the box: a collection of tactical motifs, not an anthology of great games.

There is another possibility, and I will keep it open throughout: the 27 miniatures may be an editorial statement about elite chess — that games are being decided earlier, that time controls are compressed, that opening preparation has grown so deep that technical gaps are pushed into tactical error. That is a good argument. Without a game list, one cannot distinguish a good argument from a convenient production schedule. An advertisement, by its nature, has no obligation to help me tell the difference.

The language of caution

Reading those seventeen points closely, I found something genuinely admirable: the writer never breaks advertising law.

They use "latest trends" and "new repertoire ideas". They do not use "novelty". They do not say "never published before". They say "ideas". An idea can be recycled from a 2026 game, repackaged with a new diagram, and still honestly be an idea. A new move cannot be recycled.

I have read thousands of prospectuses in forty years, inside chess and outside it. The worst ones promise too much. This one promises exactly enough, which signals either a good legal department or an editor with a conscience — possibly both.

But that caution leaves a hole the reader must fill. If the 10 opening articles contain no novelty, what do they contain? They contain what a club player needs more than novelty: a map of the main branches, warnings about traps that have gone stale, and a few suggestions safe enough to play on Saturday night without memorising thirty moves.

I once mispronounced a player's name, and then read the position correctly when the slow-motion replay appeared in front of me. The lesson was not pronunciation. It was the difference between the name and the position. Here too: "new repertoire ideas" is the name; what the buyer actually receives is an edited map. Whether the map is new requires opening the book. I have not been allowed to open it.

Three names, one stable

Werle. King. Ris.

Those three names sit together for a reason. Jan Werle, Daniel King and Robert Ris are long-associated with ChessBase's recurring video series; King is a veteran English grandmaster known for a calm voice and one of the kindest explanatory instincts in the business. In television we call a fixed roster of presenters a stable.

A stable has two faces, and both are real.

The first is trust. A subscriber does not buy a stranger. He buys a voice his ear already knows, a teaching rhythm his brain has synchronised with. For instructional products, familiarity has direct economic value: it lowers the cognitive cost of learning and lowers churn.

The second is renewal risk. A stable tends to say stable things. New angles usually arrive with new people, and new people are not in the stable. When a publication keeps the same three names across many issues, the buyer gains continuity and loses the possibility of surprise.

I have spent enough hours in studios to know this is nobody's personal failing. It is the arithmetic of a release calendar. One issue every two months, three video series, a roster already validated. Nobody dares introduce a stranger while the audience is paying to hear a familiar voice.

Six months of delay, and why it exists

This detail matters more than it looks: the tournament coverage is of Prague, and it sits inside a September issue.

The gap between the event and the reader's hands is roughly half a year. During those months, every one of those games has been in online databases, analysed on YouTube, scored publicly by engines.

A buyer seeking news will be disappointed. A buyer seeking instruction finds half a year not a defect but a precondition. The annotator needs time to check with the engine, to find the places where human judgement diverges from machine judgement, and to write a sentence no automated tool can generate.

The delay is evidence of the business model: this product prices slowness, and slowness only has value if the reader is studying rather than following.

That also explains why such a product survives in an age when everything is free. Data is free. Selection and explanation are not. People pay for someone who sat with 27 short games for six months and said: this one matters, that one does not.

Three delivery channels and one paper booklet

This is the structurally interesting part, and the least noticed.

Three formats for one body of content is an expensive engineering decision. Each channel carries formatting, testing and support costs. Why all three?

Because buyers are not homogeneous. One group reads on paper, marks with a pencil, folds corners. Another works directly in the database, dragging variations, running engines. A third — larger than most people think — props a tablet on the desk and clicks through variations while reading.

Keeping a PDF booklet inside a product that is already a database tells us something about the customer: the database-native user is still not the whole market, and the seller knows it. That is a concession to reading-native users, and I — a reading-native user, sixty years old, still fond of paper — will not complain. But it also signals that the industry's digital transition is paused halfway, not finished.

The data vacuum, and why it worries me

Now the direct criticism.

A modern chess training publication, if it wants serious assessment, must supply a minimum: a game list, average move counts, engine error metrics per game, and for the opening section, first-publication dates for the main lines.

The blurb for issue 225 supplies none of these. No list, no metrics, no dates.

I understand why. It is an advertisement, not an academic dossier. The consequence is that any judgement about technical quality, including mine, stands on a false leg. I can discuss architecture, business model, editorial strategy. I cannot tell you whether the seventh opening article is any good. An honest writer states that boundary instead of filling it with adjectives.

In my trade this is the most common fatal error: evaluating a product by the feeling of the product. I have made that error. In 2026 I mispronounced a player's name three times in the first half of a World Cup opener, and the price was not embarrassment — it was two weeks of re-watching tapes from all 32 squads. Since then I hold one rule: no footage, no technical verdict.

Here I have no footage. So I judge only what is visible.

The Indian name on the annotation list

One detail I cannot skip as someone covering chess for the Indian market: Aravindh Chithambaram is among the four annotators.

For Indian readers this is the largest value anchor in the whole package, bigger than 27 miniatures, bigger than three video series. An Indian player annotating a European event, in English, inside a globally sold product — that is a statement about standing, not a credit line.

It is also a test. What does a reader gain when an active player annotates? The view from inside the playing hall — the feel of tempo, clock pressure, the options discarded in the final three minutes. What is lost is the library view of the professional annotator with six months to cross-check.

David Navara, at the other corner, is the domestic anchor of a Czech event. Anish Giri explains the subtlest things in modern opening play better than almost anyone. Gurel brings a different tempo. Four readers of one game — a good editorial decision, and the only part of the package I can praise without opening it.

Where I might be wrong

Time to cross-examine myself, more seriously than people usually bother to do with an advertisement.

First: the 27 miniatures may genuinely be a statement about contemporary elite chess. Watching hundreds of open and invitational games over a decade, I have a feeling — a feeling, not data — that big games are being decided earlier. If so, the Special is journalism, not accounting, and I should withdraw my cost argument.

Second: a stable may be the pedagogically correct choice. In teaching, continuity of voice outweighs novelty of presenter. If I were selling a course rather than a periodical, I would keep the same three names too.

Third: the PDF booklet may not be a concession to older readers but a tool for coaches — people who print, hand out, underline, photograph and send to a group chat. If so, keeping the PDF targets the education market, and it is far wiser than my diagnosis.

Fourth, and the most uncomfortable: I am writing a long critique of a leaflet. A seventeen-point leaflet. If so, the reader is paying for an argument he did not order.

I keep the fourth, because it is the most likely of the four to be right. And I keep it the way I always keep things: say it out loud first, then keep writing.

Silent applause can kill emotion, but it cannot kill strategy. An advertisement without an audience has nothing to say. An advertisement with 27 short games has a structure worth reading.

A falsifiable prediction

I will not close with a summary. I will close with a prediction anyone can check, and I will check it myself.

If ChessBase keeps the architecture of issue 225 for subsequent issues, I predict that by roughly issue 230 one of two things will happen. Either the opening section will carry engine-verified labels on main lines with first-publication dates — because competitive pressure from opening-course platforms forces the inclusion of something rivals lack. Or the miniatures Special will be institutionalised as a numbered annual fixture, because it is cheap and it sells, and anything cheap that sells stops being occasional.

The test is simple: open the next issue, count how many opening articles carry a date for their main line, and see whether the Special returns.

Ten years in the middle of an argument, and the fire lit by a question in 2026 still burns. That question was about a forward line that refused to push up. This year's question is about a content industry that sells selection and calls it something else. The buyer has a right to know what he is paying for, and the writer has an obligation to name the difference between the map and the territory — even when the mapmaker signs the cheque.