FRITZ 20: The Chess Engine That Dresses as Teacher, Opponent and Ally

The chess software FRITZ 20 has just been announced with a far-from-modest claim: it is not merely an analysis engine, but a training revolution for serious players — from those taking their first steps into structured training to those already competing at tournament level. For the Vietnamese chess community, where many young talents grow up alongside software, the real question is not how strong the machine is, but whether a stronger tool can produce a stronger player. At three in the morning in Shanghai, the room is lit only by the blue glow of a screen. A game hangs at move thirty-four. I press the analysis button and wait. Seconds later, an evaluation appears, cold and final: the move I believed was creative is flagged as a serious error, while the line I dismissed is the only path to holding the balance. I once trusted emotion, until a number knocked on my door at 3 a.m. It was not the first time, and certainly not the last, that a machine taught me a lesson in humility. When the engine steps out of the analysis corner For decades, chess engines were tools for professionals: powerful, dry and hard to approach. Amateurs opened them to check a single move, then closed them. That began to change when developers realised raw strength teaches nobody anything. A machine capable of beating a world champion can still be abandoned by a fourteen-year-old student, simply because the boy does not understand what it just said. FRITZ 20 is positioned to fill exactly that gap. Instead of only producing the best move, the software aims to accompany the learner: analysing personal games, pointing out repeated weaknesses, and suggesting training plans suited to each individual level. In other words, it tries to shift from the role of judging referee to that of patient teacher. Chess history has recorded milestones that show how long this journey has been. In 2026, in New York, IBM's Deep Blue defeated world champion Garry Kasparov 3.5-2.5 in a six-game rematch. Nearly a decade later, in Bonn in 2026, Deep Fritz — a direct descendant of the Fritz line developed by ChessBase — overcame Vladimir Kramnik 4-2. The striking detail is this: the 2026 match took place on a commercial computer, not a laboratory supercomputer. The machine had stepped out of the research lab, and from then on it was no longer the opponent of a handful of elite grandmasters, but the companion of an entire generation. The three roles of FRITZ 20 The FRITZ 20 pitch revolves around three images: personal trainer, toughest opponent, strongest ally. It sounds like marketing, but taken apart, each role reflects a different stage in a player's growth. The role of teacher comes first. An amateur often commits the same category of mistake hundreds of times without noticing: pushing pawns while forgetting the king, calculating deeply yet missing the middlegame, opening by the book yet losing the thread the moment theory ends. An engine only says this move is wrong. A teacher must say this move is wrong because of this, and FRITZ 20 is designed to do the second. That small difference in phrasing produces a large difference in how players absorb the lesson. The role of opponent appears once a player is strong enough. Beating a weak machine to win is a dangerous habit, because it lulls the human with a false sense of safety. When the board falls silent of spectators, the true value of a person begins to speak. The toughest opponent is not the machine that defeats us, but the machine that forces us to think on a deeper level than usual. The role of ally is the most subtle, and the most underrated. After every game, a player needs a friend who does not judge, who turns over every move and points out that at move twenty-two there was a better option. An analysis tool does not buy the past. It buys what the data has already forgiven. Vietnamese chess at a new threshold In Vietnam, chess is a sport with a respectable history that is often left behind by the media. People remember the medals, but few remember the training hours behind them. In 2026, at the World Blitz Championship in Khanty-Mansiysk, Le Quang Liem took the title — one of the rare milestones that brought Vietnamese chess broad recognition on the world map. That achievement did not come from magic. It came from thousands of hours at the board, and no small part of them spent at a screen. What is worth noting is that most of that training infrastructure is not inside federation training halls. It is inside small rooms, dormitories, private homes, where a high-school student turns on a laptop at eleven at night to review his own game. No coach sits beside him. Only a machine. That is why the debate around FRITZ 20 is not academic. For talents without access to international experts, software is the only teacher they have. If that tool is patient enough to teach rather than simply judge, its value goes far beyond a piece of technology. The counter-view: stronger engine, weaker players? There is a paradox the coaching world rarely states outright. The stronger the engine, the lazier players become. When every question has an answer within seconds, human calculation atrophies year by year. I have seen young players memorise hundreds of opening variations yet freeze completely when an opponent steps outside the book. They have data, but not understanding. They have answers, but not questions. I call that trap the engine eye. Players look at the board through the eyes of the machine that analysed for them, not through their own. They judge a move as good because the evaluation bar jumps, not because they understand why it is good. When they face a position that has never appeared in a database, they lose their bearings. This is the moment to be clear: the surprise is not in the strength of FRITZ 20, but in how it is used. A machine that only delivers answers produces dependent players. A machine that asks questions produces players who think for themselves. The difference between these two kinds of players shows most clearly in the endgame, when there is no theory left to cling to and only judgement remains. One point is easily overlooked when discussing chess data: volume of information does not equal quality of understanding. A player can store tens of thousands of games on a computer and still fail to improve, because improvement comes from asking the right question, not from saving the right file. The players forgotten by the media, those who quietly sit down again after every tournament, are usually the ones who understand this best. There are players the world forgets, but data never forgets them. The question is how data remembers them — as a dry memory, or as a living instruction. What lies ahead The true value of FRITZ 20 will not be decided by how many Elo points stronger it is than the previous version. It will be decided by how many players, after a year of use, dare to switch the software off and calculate with their own minds in a real game. For Vietnamese chess, this could be a chance to close the gap with the leading chess nations, provided the tool is used as a teacher rather than a crutch. I light a candle for data. But I always let the flame of emotion illuminate the question. Because after all, what a machine cannot teach is why a human still sits down at the board, every evening, when nobody is watching. The question it leaves behind is simple: if the machine already knows every move, then the remaining work — the most interesting part — belongs to whom?