Fritz 20 and the training revolution: when the engine is no longer only there to beat you

Fourteen seconds. That is how long a seventeen-year-old student of mine sat motionless in front of the board in my training room. He was not hesitating because he lacked ideas. He stopped because he had just realised he had repeated the same pattern three times in a single game — on move twelve, move twenty-one, and move thirty. The same kind of trade, the same flank, the same unconscious reflex. The person who pointed that out was not me. It was FRITZ 20. Fritz is not a new name in chess. ChessBase's engine line has existed for more than three decades and has been the central figure in moments that silenced the chess world. In 2026, in New York, Garry Kasparov drew Deep Fritz 4-4 in the Man vs Machine match. Three years later, in Bonn, Deep Fritz defeated Vladimir Kramnik 4-2. At the time, that result was read as the end of human ambition at the board. But FRITZ 20 does not retell that story. It tells a different one, and to my mind that is where the interesting part lies. Over twenty-three years of watching the chess industry, I have seen training change at least three times. The era of paper books, when everything depended on whether you could reach the right volume. The database era, when a vast archive of games opened to everyone. And the engine era, when a machine can assess any position to the last percentage point. FRITZ 20 belongs to the third wave, but it positions itself elsewhere: it does not sell power, it sells a method of training. Three roles are bundled into one name: a personal training coach, the toughest opponent, and the strongest ally. That framing admits something many young players still refuse to accept. Raw engine strength has long ceased to be a selling point. The race to calculate deeper ended years ago. What remains is the question of what the machine teaches you, and how. Watching students train over many months, I keep noticing one repeating pattern. Amateur players do not lack information. They lack the ability to see themselves. They open a database, watch hundreds of grandmaster games, memorise a few opening ideas, then walk into a real game with their old habits fully intact. What they need is not a bigger library, but a mirror sharp enough to reflect reflexes they cannot see on their own. In that sense, the architecture of FRITZ 20 should be read as a system, not a feature list. The opponent role solves the problem of missing quality sparring partners — something anyone who has trained in a small town understands. The ally role solves the problem of missing reference data. But the coach role is what makes the difference, because it does not simply hand over the right move. It points at the gap in your thinking. Not the move, but the gap before the move appears. What strikes me about how FRITZ 20 is positioned is the individualisation. Coaching does not produce identical players — it produces non-identical paths. A player with an attacking bent needs to be drilled in calculating lines, while a positionally strong player needs to be pushed into chaotic games to learn how to endure. A modern training engine has an advantage here: it does not impose one shared template, it can build a private environment for each person. That is exactly what a human coach wants to do, but rarely has the time for. For professional players, the value sits in preparation. Across a long tournament, knowing an opponent's opening tendencies and finding small deviations in their habits can decide an entire game. A machine that handles that volume of data saves a player dozens of hours of manual work — time better spent on something more important: rest, recovery, and keeping a clear head before the game. I have always thought about the board and the pitch in the same way. Every move, every pass rewrites the controlled area, and what actually decides the outcome happens in the empty space before the piece is touched. A good training engine should help you read the position that is about to appear, not just contemplate the one that has passed. That is what the Space Map means — not a still photograph of the current position, but a three-dimensional image of the possibilities waiting to be triggered. Here is the point I want to underline: every plan is a hypothesis until the piece is touched. The value of a training engine is not that it is right, but that it forces you to question what you believe is right. But this is the blind spot I suspect most users will walk straight into. Training against a strong engine does not automatically produce a strong player. I have seen plenty of students become dependent on the machine's evaluation — touching any position and turning analysis on before thinking for themselves. They become translators of results rather than decision-making players. At the board in a real game, there is no analysis button. There is only you, the clock, and time pressure — three variables an engine fundamentally cannot fully simulate. The Space Map never lies — it only exposes what we want to believe. The machine does not teach you how to endure a position worse than your opponent's, does not teach you how to use the last thirty seconds on the clock, and does not teach you how to read fatigue on the face of the person sitting opposite. Those are human variables, and that remains the part where a coach genuinely holds the role. In other words, FRITZ 20 is a directional tool, not a solution. A player who uses it as a training partner will improve. A player who uses it as a crutch will stand still, however strong the engine becomes. I once spent two hundred and fifty days coding three hundred and eighty matches to understand one simple thing: data only has value when the person reading it knows how to ask the right question. The same logic applies to every training engine. The question I leave behind is not how strong FRITZ 20 is. It is this: when the machine is good enough to point out every mistake you make, do you have the courage to sit with those repeating mistakes before pressing the analysis button. Fourteen seconds. It does not take longer.