ChessBase releases August 2026 training products: Petroff Defence and two positional-play courses

On September 13, 2026, ChessBase published its regular product bulletin under the headline "Review: All new releases in August!" The same day, a different label appeared in a transmission note: "Chess historians meet in Copenhagen." Two names for one document, and nowhere in the body of the text is there a single sentence about Copenhagen, about chess historians, or about any meeting in Denmark. A data-transmission error at the level of the label.

For someone who earns a living reading data tables, a fault in the label often tells more than the body of the article. I read the numbers three times before writing and once more afterwards. After four passes, what remains on the table is four indisputable facts: a database update for the Petroff Defence; two courses on positional play; a group of materials on calculation and strategic decision-making; and three author names, Bahne Fuhrmann, Andrew Martin and Kostya Kavutskiy.

That is why I chose to write about this bulletin, even though it contains no game at all. No results, no Elo ratings, no tournament records, no rules disputes. For someone used to reading score sheets and metrics, a text empty of competitive data is still a valuable data sample. It shows what is being sold, to whom, and at what rhythm.

ChessBase is not a newsroom

This publisher is a producer of software and training content, and the August bulletin sits inside a monthly release chain that is maintained with regularity. That rhythm is itself information. A digital-content business that lives on subscriptions cannot leave gaps in its catalogue, because a gap in the catalogue is where churn begins. In other words, the release calendar here serves retention rather than editorial need.

Within the August catalogue, the only opening-theory item refreshed is the Petroff Defence. That choice deserves explanation for readers who do not follow openings closely. The Petroff is a defence in which, after White plays e4, Black answers e5, White brings the knight to f3, and Black replies with Nf6. Black does not retreat into passive defence but strikes directly at the opponent's central pawn, accepting early trades in exchange for symmetry.

Based on my experience watching games at international events, the Petroff is one of the few options a player with the black pieces can choose when the goal is to cool the position from the very first moves. It does not create an advantage, but it removes the opponent's. For players entering a long tournament who need to conserve energy for decisive rounds, this is a weapon with its own value, unglamorous but extremely practical.

The problem is that Petroff theory has been excavated to a very deep level. Small differences in move order can lead to completely different structures, and modern analytical tools mean every branch is recorded, tagged and archived. A database update for this defence does not create new theory in the exploratory sense; it ensures that users do not walk into a game carrying lines that went stale two or three seasons ago.

The three authors of the release also deserve to be read as a strategic fact. Andrew Martin is a veteran voice of the publisher itself, tied to many opening books and courses. Kostya Kavutskiy is a US international master who is highly active in commentary and online teaching. Bahne Fuhrmann represents the group of authors specialising in skills training. Three names, three different audience files, and that is no accident.

What is actually being sold in August

The most notable feature of this release lies in the structure of the catalogue rather than in any single product. The material splits into two clear blocks. The first is opening theory, represented solely by the Petroff Defence update. The second is transferable skill, covering positional play, calculation and strategic decision-making. Two positional-play courses appearing in the same release cycle is a detail worth recording, because it shows the publisher is placing its weight on skills rather than theory.

This split reflects a paradox of the modern chess training market. Opening products have short life cycles, face constant update pressure, and are priced by their own rate of obsolescence. Skills products last far longer. A course on how to assess a pawn structure, how to calculate a sequence of variations, or how to decide in a balanced position will not lose its value after six months. A platform that owns both ends collects money from the theory race and from the mass skills market at the same time.

The three authors also operate along the standard curriculum model of the training-content market: theoretical explanation, annotated games, and practical exercises. This triad repeats so often that it has become a convention, and the repetition has a reason. Learners need a theoretical frame for orientation, real games to see the theory operate, and exercises to test themselves. Remove one of the three and the product will be judged shallow or impractical.

Another detail deserves to be placed on the scale: the material is segmented by different playing strengths. That is a market-strategy signal. The publisher is not aiming at a narrow elite group but stretching its products from beginners to club players. That strategy maximises the pool of potential buyers, but the trade-off is pressure to keep quality even across many levels, which is not easy when a release has to ship every month.

What is unfortunate is that the bulletin provides no verifiable quality metric. There is no engine-based assessment, no average centipawn loss, no sample content for readers to check for themselves. Every claim of value comes from the publisher itself. For readers used to reports with tables and sources, that is a large gap.

Here I have to say something my trade taught me long ago. There are players the world forgets, but data never forgets them. In this case, data remembers nothing either, simply because no data was offered. A product catalogue is not a metrics table. That is the distinction readers must draw for themselves, because otherwise a sales release will be read as a professional report.

The contrary view: when the seller grades his own work

This is the part I consider most important, and also the part most easily skipped when reading a product bulletin. The publisher and the reviewer are the same legal entity. Every image in the article is supplied and credited by that same entity. The tone of the text is supportive and promotional. I am not alleging wrongdoing here, because that is the nature of a product-announcement piece. But the reader needs to know that the signal value of the quality claims is very low.

I once trusted emotion, until a number knocked on my door at three in the morning. That number, in this case, is an absent number. And that absence is the most noteworthy thing of all.

The second issue is the two positional-play courses arriving in one cycle. There are two explanations. The first is depth: the publisher deliberately covers several facets of positional chess, from pawn structures to piece placement. The second is catalogue overlap, where the same subject is repackaged in two forms to fill the release calendar. If the second explanation is right, it signals a market that has reached saturation, where quantity substitutes for novelty.

Another possibility cannot be ruled out, and it deserves a place on the table: the theoretical caution of this cycle. Beyond one Petroff update, no new opening appears. Meanwhile most of the weight goes to calculation, strategic decision-making and positional play, that is, skills that do not depend on a specific theoretical line. There is a rather cold reading of this phenomenon. When analytical tools turn opening theory into a commodity, the competitive edge of a content publisher shifts to precisely the zone machines cannot reach. That zone is transmission. That zone is teaching people how to think.

Finally, back to the mismatched label. A labelling error in a product bulletin sounds minor. But place it beside the reality of this industry. This is a field accustomed to tracking Elo to the decimal, recording the value of every contract, and logging every move of every game. Yet the most fragile layer of data sits at the most elementary point: the name attached to a document. An archive never forgives a wrong label. It will repeat that wrong name every time someone searches for it.

One more thing must be said about the silence of the data. This bulletin mentions no dispute, no cheating, no financial crisis, no disagreement between federations. That does not mean such problems do not exist in the industry. It only means they fall outside the scope of a sales release. Reading absence as confirmation of safety is a methodological error I have seen far too many times.

Signals for the next cycle

There are four things I will be tracking in the coming months. The first is the topic mix of the next release cycle. If skills courses keep outweighing opening updates, that is evidence the platform is shifting away from the theory race. If the opposite happens, the opening market still has commercial room.

The second is the appearance of new Petroff Defence lines in databases and in live tournament play. An update only matters when it leads to a new move that actually appears on the board, not merely in a product listing.

The third is the degree of overlap with competitors' catalogues. When two different publishers announce the same topic, the same author, or the same timing, that is a sign of rising competitive pressure in an already crowded market.

The fourth is the integrity of the labelling layer. One mismatched headline can be an accident. Two is a process problem, and process problems always bleed into the content.

I light candles for data. But I always let the flame of emotion light the question. The real concern in a bulletin like this is not which products were sold. It is the question behind them. When every theoretical line has already been recorded and classified by machines, what a chess learner truly needs to buy is no longer information. It is a way of thinking. And no database can archive that on a student's behalf.