FRITZ 20 and the Quiet Death of a Generation of Chess Engines

There was a rainy afternoon in Saigon, and I was sitting in a small chess café on Nguyen Dinh Chieu Street, where an old computer in the corner still glowed with its bluish screen. On it was a piece of software that the café owner – a former national-level player from the 1990s – kept like a treasure: Fritz. He told me that back then, every evening, the young players of the club would gather around that machine, holding their breath as it answered a move. None of them suspected that twenty years later, the name Fritz would survive only as a faint line in a few people's memories. This morning, ChessBase – the company behind Fritz – announced the release of FRITZ 20. In its press release, they call the product "a revolution in training," a "personal chess trainer," meant for "both those taking their first steps and ambitious players already at tournament level." Those words sound very familiar. And I, as someone who records things, asked myself: is there anything here worth remembering, or is it only the echo of something that stopped burning long ago? When a machine that once made the entire chess world tremble is introduced again under four words – "a training revolution" – I understand that I am witnessing something larger than a software upgrade. It is the moment when a golden generation quietly leaves the front page. The real death of a golden generation is when we stop remembering them. Fritz was not born in silence. In 2026, when the personal-computer wave was still young, two German programmers, Frans Morsch and Mathias Feist, created an engine bearing this name and handed it to ChessBase for distribution. For more than a decade, Fritz was the name mentioned whenever people spoke of chess and machines. It did not merely calculate; it was present. In 2026, one version of Fritz caused an uproar when it defeated the legend Garry Kasparov in a man-versus-machine match. Four years later, in 2026, Deep Fritz defeated Vladimir Kramnik – then the reigning world champion – by a score of 4-2. That was the moment the chess engine stepped out of the laboratory and became part of a sports story. Fritz was no longer software; it was a character, with fans and haters and nights when an entire chess club in Saigon stayed up to watch it play move by move. I remember the feeling of that café owner when he recounted the Deep Fritz – Kramnik match. He said he did not watch it to know who won. He watched it because, for the first time in his life, he saw a machine play chess like a human being who gets tired, hesitates, and chooses a weaker move in exchange for a better position later. That is something no scoreboard ever records. Then the market changed. In 2026, Stockfish was born as an open-source project. It needed no advertising, no boxed disc, no license. It only needed to be downloaded. Within a few years, Stockfish became the standard of computational strength, far surpassing every commercial engine of its time and – more importantly – free to everyone. In 2026, Leela Chess Zero appeared, carrying a neural network and a completely different approach, proving that artificial intelligence could learn chess in a way closer to humans than to rigid evaluation tables. Alongside those two milestones came the rise of integrated platforms. Lichess offered free analysis to anyone with an internet connection. Chess.com gathered the entire experience – play, study, analysis, training – into a single account, with increasingly sophisticated artificial-intelligence features. Once top-level chess analysis became a free service, the question facing a company like ChessBase was no longer how to be stronger, but how to be remembered. And FRITZ 20 was born in that context. In ChessBase's press release, I counted seven information points. Not one of them mentions computational strength. Not one ACPL figure, not one match rate against the leading engine, not one comparison with Stockfish or Leela Chess Zero. All that is offered are adjectives: "efficient," "intelligent," "individualized," "groundbreaking." This is the core of the matter: a product sold through emotion when it has nothing left to prove through data. I spent many evenings rereading old ChessBase press releases and comparing them with what the market has done over the past fifteen years. What I realized was not whether FRITZ 20 is weak or strong. It was that whoever wrote that release seemed to know that engine strength is no longer their battlefield. When you cannot win on the purely technical front, you move to the front of meaning. You call your product a "revolution," and you hope the memory of the name Fritz will do the rest. But memory is a difficult friend. It remembers Fritz of 2026, not Fritz of an undated update. In the entire press release I read, there is not a single line stating a release date, no author's name, no editorial context. For someone in my line of work, that absence is itself information. It suggests this may be evergreen advertising content, written to exist in search engines rather than to mark a moment in chess history. Look at the contradiction within the pitch itself. FRITZ 20 is said to be for "those taking their first steps" and simultaneously for "players already at tournament level." In any industry, a product that serves everyone usually serves no one. An engine for beginners needs a simple interface, clear explanations, exercises at the right level. A tool for professionals needs analytical depth, a vast opening database, customization. Merging these two audiences into one sentence suggests the product is trying to cover a market too broad, and that is usually a sign of unclear positioning. In the chess world, we are used to products that choose one direction and go all the way. Stockfish chooses pure strength, and it dominates there. Chessable chooses structured learning, and it leads there. Chess.com chooses an integrated experience, and it owns that. A product claiming to be everything is often a product that has not found its place in that picture. I do not say this to deny the value of FRITZ 20. I say it because I believe the truth, however quiet, deserves to be recorded. ChessBase was once a pioneer. It brought chess engines to millions of players and organized man-versus-machine matches that changed how we view the limits of human intelligence. But the times have changed. When a free tool can analyze a game better than any paid software, the value of paid software no longer lies in strength. It lies in experience. And experience is hard to prove through a seven-point press release. What is worth noting is a paradox in the chess world that few are willing to face directly. We do not lack strong engines. We lack people who understand what those strong engines are saying. A young player in Hanoi or Can Tho can download Stockfish in thirty seconds and hold the strongest tool in history. But that player still needs someone – or something – to show them where to look among thousands of variations. If FRITZ 20 truly solves that problem, it is worth its price. But ChessBase's release says nothing about how it does so. It merely says that it does. In that context, the fate of FRITZ 20 becomes a small story about what is happening to an entire industry. Chess-software developers are watching their value erode through the very success of their field. The more people play chess, the more good free tools appear. The more good free tools appear, the harder it is to sell a box of software. And when you cannot sell on features, you sell on story. I have followed Vietnamese chess for many years, and I see a smaller version of this story there. Tournaments that once drew large crowds are now quieter. Players once celebrated in the papers now appear only in a few short news lines. No one did anything wrong. Attention simply shifted elsewhere – to sports with more visual appeal, to platforms with more engaging algorithms. The empty stands still have an echo, if only we are quiet enough to hear it. There is one thing in ChessBase's release that made me pause longer than anything else: how they used the word "training." In German and in English, this word carries a very specific tradition – it evokes the image of a patient teacher, a long process, lessons without grades. But in an advertising release, that word is dragged into another space, where it must compete with promises of rapid progress. When "training" becomes a feature, it loses most of its meaning. People learning chess do not need another feature. They need time, and a companion. That is why I was not surprised that a product like FRITZ 20 leans so heavily on language. It is trying to sell something that cannot be bought with a click. And when a product tries to sell something that cannot be bought, it tends to attract the wrong buyers. Ordinary users may be drawn in by the word "revolution" and purchase it, only to find that what they receive is not much different from what they already had for free. That is the real risk of this kind of long-lived advertising content: it is not factually wrong, but it is also not honest in meaning. Back to the café owner in Saigon. When I asked whether he planned to buy FRITZ 20, he smiled. He said his old computer still ran Fritz, and though it was slow, though it was weaker than every modern tool, he kept it because it was tied to a memory. He told me: "What I remember is not how well it calculated. What I remember is that night I saw it hesitate." Hesitation. That is the detail no advertising release ever records. Over the past fifteen years, chess engines have become so perfect that they no longer hesitate. They calculate the best move in an instant and never waver. But it is precisely that perfection that makes them feel alien to humans. Chess players do not hesitate because the machine is weak; they hesitate because they are human. And a good training tool, in theory, should help a person live with their hesitation, not replace it with the machine's perfection. If FRITZ 20 understood this, it would have a story to tell. But ChessBase's release does not tell that story. It tells another story, about progress, efficiency, individualization – words that are technically correct but emotionally empty. This is the core I want to keep after reading that document: a generation of chess tools is gradually losing its own language. When it no longer knows how to speak of hesitation, it starts to speak of revolution. This leads me to a counterintuitive thought. We usually assume that in the race of chess engines, the strongest wins. But the history of Fritz shows the opposite. Fritz did not win because it was the strongest engine at every moment. It won because it was the engine best remembered. It did not merely calculate; it told a story. When it defeated Kasparov, it did not prove that machines were better than humans. It proved that a machine could also be a character, could have a moment, could keep an entire chess club in Saigon awake all night. The blind spot of collective memory is that we remember the result but forget the path. We remember that the machine won, but we forget that for a long time we were not sure whether it would. That uncertainty is what makes matches memorable. In a world where Stockfish can beat anyone with near-absolute certainty, uncertainty disappears, and with it much of the appeal of the story. That may be the deepest reason why people today discuss chess less, even though the tools are many times better. But there is one thing even Stockfish cannot take away: memory. The memory of nights spent awake beside an old computer. The memory of a café owner who still keeps fifteen-year-old software because it is tied to a feeling. The memory of the moment a machine hesitated. These memories appear in no press release and cannot be created by an update. From the perspective of sports business, the story of FRITZ 20 is an example of how brands lose their connection to local communities. ChessBase once had a community – loyal users bound to the name Fritz as to a part of their youth. But in its effort to reach a broader global market, the company seems to have lost the language with which to speak to that very community. Its advertising messages aim at figures, at reach, at conversion funnels. They do not aim at the café owner on Nguyen Dinh Chieu Street, who still keeps an old computer for reasons no data table records. And here the story becomes larger than a software product. It is the story of every sport chasing visually attractive disciplines, leaving behind empty stands and those who quietly remain seated. Chess is no exception. Young players raised in the age of Stockfish can play at a higher level than any previous generation, but they also have fewer chances to feel uncertainty – the very heart of any sport. When everything becomes calculable, the joy of not knowing disappears. The echo of the empty stands resounds within the very name FRITZ 20. It is a product made by a company that once filled the stands and now tries to call back those who have left. In its release, I do not hear confidence. I hear nostalgia – nostalgia for a time when the name Fritz alone was enough to sell, without words like "revolution." For readers interested in chess, what I want to emphasize is not whether FRITZ 20 is good or bad. I have not tried it, and I will not conclude. What I want to emphasize is how we should read releases like this. When a company talks about its product without offering a single verifiable figure, that is information. When a company targets both beginners and experts, that is information. When a release has no date, that is information. The gaps in a story often tell more than what is written. In my years following chess, I have learned that big changes are never loud. They happen in silence, between unnoticed updates, between nights when no one stays awake for a move anymore. The strongest chess engine in the world today is downloaded for free, and few remember the names of those who wrote it. And what we called legends fifteen years ago is now only a faint line in an undated release. The question I leave behind is not whether FRITZ 20 is worth buying. The question I leave behind is this: when a generation of chess machines stops being mentioned, are we still quiet enough to hear its echo, or have we gone so far that we only know how to wait for the next update? Generations will pass, but the poem about them keeps being written. And sometimes, the one writing that poem is just a person sitting in an old café, watching a computer still run outdated software, listening to its hesitation, and recording everything before it disappears.