New Chess Training Package: From Grischuk's 9.Nd5!! to the Endgame Fortress

In the Sicilian Kan game where Alexander Grischuk played White, the ninth move was a knight throwing itself onto d5 only to be captured. White accepted the material loss and did not calculate a way to win the piece back. What it bought was a bind in the centre, a Black bishop entombed behind a pawn chain, and control of the game's tempo. The full sequence is recorded: 1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.c4 Nf6 6.Nc3 Qc7 7.Be2 b6 8.0-0 Bb7 9.Nd5!!.

That move sits in a chapter titled "Grischuk's Stroke of Genius" inside a chess training package recently released as a download, compiled by Petra Papp. The package contains six further chapters, running from opening to endgame, plus a 52-page PDF file. To a beginner, 9.Nd5!! looks like an error. To anyone who has sat long enough at the board, it is a statement about structure: I do not need this piece, I need this square.

Context: one content package, seven authors, one familiar format

This is a product from the expert-training line, not a game or a tournament. There are no results, no Elo ratings, no standings. What is being sold is instruction, and what underwrites the instruction is the names of those who wrote it: Christian Braun, Dorian Rogozenco, Robert Ris, Oliver Reeh, Sergey Grigoriants, Mihail Marin and Petra Papp. This packaging model — many topics in a single purchase, with a printable PDF — is not new. It has been the standard model of the chess training-content market for more than two decades.

Based on my experience tracking the chess training market over many years, I see this model under twin pressure. On one side sits the vast free content library on online platforms. On the other sits increasingly cheap computer tools that turn variation lookup into a matter of seconds. When opening knowledge is commoditised, the only layer still sellable is the explanatory layer. At that point, author names become the asset, not the move sequences.

From the sideline, I see the whole game. People usually read the opening section first and skip the endgame section, yet the very structure of this content package shows its makers understand where long-term value is created.

The core: decoding seven chapters

The first chapter, "Give me the f-file!", compiled by Christian Braun, handles the Anti-Berlin line of the Ruy Lopez. The sequence: 1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3, after which Braun chooses 4...Bc5 5.Be3!?. The idea is to avoid the notorious Berlin endgame by accepting a slower middlegame structure, then opening the f-file with a pawn lever and a rook lift. I have analysed a great many Berlin endgames in their dry final phase, and this approach solves a simple psychological problem: many players do not want to enter an ending they do not understand.

The second chapter moves entirely to the endgame. Dorian Rogozenco presents "The Fortress", the fortress setup in opposite-coloured bishops with two connected passed pawns. This is one of those positions where the weaker side can hold the draw if the pieces are placed correctly. The chapter's focus is not attack but defence: where the defending pieces should stand, and in what order they should be deployed. For club players, this kind of knowledge is worth more than any combination.

The third chapter, the Barendregt Variation, compiled by Robert Ris, follows a King's Indian and Benoni structure: 1.d4 g6 2.c4 Bg7 3.Nc3 d6 4.e4 c5 5.d5 Bxc3+ 6.bxc3 e5!?. The result is a shattered White pawn structure against a Black central pawn mass. This is a classic strategic imbalance: White has the bishop pair and open files, Black has a solid centre. One technical note is required — in the original text, the move order in this chapter contains a numbering error, so the exact sequence should be checked against an authoritative source before use.

The fourth chapter returns to the French Defence. Sergey Grigoriants presents the Alekhine–Chatard Attack: 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 Be7 5.e5 Nfd7 6.h4!?. This is the textbook positional pawn sacrifice, striking at the kingside before Black can castle. The package's description — broadly, that the idea remains as dangerous as it was a century ago — is rhetorical framing, not a measurement.

The fifth chapter, "The Isolated c-Pawn", compiled by Mihail Marin, uses two Garry Kasparov – Anatoly Karpov games as teaching material on the fragmented structure. Choosing those two games is a deliberate pedagogical decision: the learner is led to the high strategic layer rather than sent hunting for novelties. Marin is well known for his opening-historical research, so the choice is consistent with his strength.

The sixth chapter, "Attack with Qh5!", compiled by Oliver Reeh, is a motif-based format: the learner is given a recurring attacking idea across many openings, with 32 example games and exercises. This is the structurally best content type for practice, because it turns a specific move into a recognisable pattern. People watch the pieces move. I watch how a motif repeats across different positions.

The seventh chapter closes the circle with precisely that 9.Nd5!! move by Grischuk in the Sicilian Kan.

The counter-intuitive angle: missing verification data

One thing the product introduction does not say needs to be said plainly. There is no metric to measure the technical quality of the variations being taught. There is no engine match rate, no average centipawn loss, no win-draw-loss rate from a database for any of the sequences. Words like "magnificent" or "as dangerous as a century ago" are assertions, not evidence.

This does not devalue the product. It simply places it in its correct category: a teaching product sold on author authority, not on performance data. And because opening content is published publicly, the product also cannot advertise "secret preparation". Its value lies in curation and pedagogy.

The first piece I wrote on the gegenpressing system drew heavy criticism. Data never takes offence. I spent six weeks reviewing footage, logged 412 situations, and published a correction. That habit taught me one thing: every technical claim needs independent verification, even when the person making it is a famous grandmaster.

With this content package, the simplest check is to cross-reference two concrete ideas — 5.Be3!? in the Anti-Berlin and 9.Nd5!! in the Sicilian Kan — against a game database. If the frequency of these two ideas rises at major events over the next twelve months, the product's forward-looking value will have been confirmed.

The system does not lie, but it can only be heard when the data is thick enough. Seven months without football, seven months of endlessly asking why — and those seven months taught me that the data gap is always where a writer is most likely to fool himself.

There is one more point. Using Kasparov – Karpov as teaching material is a prestige anchor, not a competitive update. Those two games have long sat in the common treasury. The added value lies in the framing, not in new information.

What to watch

Three signals are worth observing in the coming months. First, the release cadence of this product line — a regular cycle confirms a stable editorial model. Second, how often the two ideas above appear in opening databases. Third, the stability of the author roster — if the same names return in subsequent issues, that is the mark of a maintained expert bench rather than an ad hoc gathering.

I am sixty-nine. I still learn from the young ones. Football does not retire, and neither does chess — a knight sacrificed on d5 can still teach me something new about how a structure operates on its own.