25 Years of the Endgame Academy: Karsten Müller and the Bet on a Chess Learner's Memory

On August 14, 2026, I opened ChessBase News while waiting for a transfer data sheet from Shenzhen. The headline hit immediately: a new endgame course by Karsten Müller, tied to a "25 years ago" milestone. I paused on that number 25 longer than usual. In data analysis, I am used to figures being cited to sell something, but this is a real milestone, and behind it sits a larger question: has chess education advanced far enough, or is it merely refurbishing what it already has? I do not buy courses for entertainment. Every time I pay to learn, I place one question on the scale: after six months, what percentage of it will I still remember. That is why I read this announcement with the eye of a process evaluator, not a consumer. Checkmate & Pawn Endgames is the first course in the new Endgame Academy series published by ChessBase. The content is led by Karsten Müller, a name European endgame researchers place among the most influential authors of the past two decades. The reviewer named in the announcement is Michael Kotyk, a young coach from Hamburg. The notable element is not that another endgame course exists, but how it is positioned: not video, but interactive exercises. I have followed the chess education market since 2026, from the era when VHS opening tapes still sold well. Twenty-six years later, the structure of the problem has barely changed: learners pay for time, and sellers must prove that time is not wasted. What has changed is the method of verification. Where we once trusted an author's reputation, we can now measure response time, correct-answer counts, and forgetting rates month by month. When data does not lie, we are the ones lying to ourselves. I once bought three different endgame courses and told myself I was learning. In reality, I was only watching. The gap between watching and learning is exactly the gap the interactive-exercise concept tries to close. Technically, Checkmate & Pawn Endgames is a sound pedagogical order: pawn endgames are the foundation of all endgame technique, and mating patterns are the most basic building blocks of calculation. Learners move from the most measurable material to the most abstract generalization. But correct sequencing does not equal sufficient depth. The announcement states no exercise count, no total runtime, and no progression structure. Those are three gaps anyone about to pay should question first. Michael Kotyk is introduced as a young coach examining the product under a magnifying glass. This is a notable communications detail: ChessBase is borrowing the image of a new-generation chess teacher to create a sense of independent verification. But the degree of the reviewer's independence is not specified. In my line of work, a self-introduced source needs another source for cross-checking. It took me three months to learn that a beautiful chart is worth less than a correct process. That principle applies here unchanged. Interactivity is not a feature; it is a promise about process. If exercises are well designed, learners must commit to a move before seeing the answer, must fail, must remember their own failures. If poorly designed, it is just a video with a button. Viewed more broadly, this is an ecosystem-consolidation move. ChessBase does not sell a single course; it sells access to a library embedded in its software. The longer users stay in that ecosystem, the higher their switching cost. I understand this logic well, having worked with the data of a Chinese club and seen how they built fan habits before building the team. A Chinese club taught me that data is not the destination, but a walking stick. An endgame course is the same. It cannot replace your calculation; it only keeps you from falling when you face a position you have never seen. The point I want to counter sits in the "25 years" framing itself. That number is used as a heritage badge. But heritage and innovation are two different things, and in this industry's history, products promoted by their age have often been refreshed in form more than in substance. I am not asserting that happened with the Endgame Academy. I am saying the available public evidence does not rule it out. Second: interactivity can be a trap. There is a psychological effect I have observed in myself and in many young students: the feeling of progress comes from completing exercises, not from understanding them. A learner solves a thousand pawn positions and still loses in a real endgame, because real play does not arrange positions for them. This is the gap between training data and match data, and a pretty interface cannot erase it. Third, and perhaps most important: a product promoted by its own vendor cannot be the only evaluation source. This announcement has no price, no exercise count, no feedback from independent users. Those three gaps together produce an undetermined evaluation state, not a negative conclusion. I want to be explicit to avoid being read as an attack. Karsten Müller's authority in endgames is real, built on decades of articles and specialist books. The problem is not the author. The problem is the reader, when we convert an author's reputation into confidence in a specific product we have never verified. In transfer analysis, I apply a principle called long-term valuation: do not judge a player over one season, but over three, including an injury season and a loss-of-form season. For a chess course, I propose the same frame. Do not ask "do I like this course", but "after two seasons, what percentage of it can I still use in real games". In the 2026 chess education market, there is a paradox nobody has named outright. Free platforms are increasingly strong on endgame exercises. Paid platforms are increasingly strong on structure. Learners have more options than ever, yet it is harder than ever to determine which actually works. After 2026, I no longer trust predictions. I trust only early-warning systems. For the Endgame Academy, my early-warning system has three signals to watch over the next quarter. First, whether independent reviews emerge from named coaches, not just from people inside the ChessBase network. Second, whether pricing and access model are published transparently enough to compare with alternatives. Third, whether subsequent courses in the series reveal a curriculum designed in advance, or merely products chained together by market demand. All three signals can become data within a quarter. That is short enough to wait, and long enough to judge. What I take away is not about any specific course, but about chess education shifting from competing on reputation to competing on training interface. Whoever wins that shift will not be decided by an author's name, but by what learners still remember after they stop practising. To me, a young coach in Hamburg invited to scrutinize the product is a more notable signal than the 25-year milestone. It shows the next generation of this industry is stepping up to the evaluation table, and they will have different standards. The question I leave behind: if every course has interactive exercises five years from now, what will remain to distinguish these products from one another?