FRITZ 20 and the Void Behind ChessBase's Training Revolution Claim

In August 2026, in Khanty-Mansiysk, a twenty-two-year-old Vietnamese man rose from the board after the final round of the World Blitz Championship. There were no Vietnamese spectators in the hall. No domestic television cameras. Le Quang Liem became world champion, and it took nearly a day for the news to crawl back to Saigon through a short wire dispatch from a foreign agency.

I sat in a small editorial room at VTC that day, rewinding a tape that held only the sound of chess pieces striking the board, and thinking something rather cold: Vietnamese chess has never had the habit of recording its own golden moments. An empty stadium still holds the applause of the dead, but for Vietnamese chess, even the stands have never been full.

Twelve years after that story, I sat down in front of an installation file called FRITZ 20.

This week, ChessBase — the chess software company founded in Hamburg in 2026 by Frederic Friedel and Matthias Wüllenweber — released a press statement introducing the latest version of its flagship line. The text describes FRITZ 20 as a "personal chess trainer," aimed at both players taking their first steps and players already competing at tournament level. ChessBase calls it a revolution in training.

I read that material three times. First, I looked for a number about engine strength. Nothing. Second, I looked for a feature comparison with rivals. Nothing. Third, I looked for a release date. Nothing there either.

A chess product promoted without a single technical metric is a notable thing, because this is a discipline where everything is measurable. Engine strength is measured in Elo. Move quality is measured in ACPL, average centipawn loss per move. Speed is measured in kilonodes per second. Chess is a sport that can be measured down to the percentage point. Yet a statement of that length could not produce one measurement.

To understand why that detail matters, one has to remember what the name Fritz once meant.

In 2026, the first version of Fritz was released. In October 2026, Deep Fritz drew world champion Vladimir Kramnik 4-4 over eight games in Bahrain. In 2026, X3D Fritz drew Garry Kasparov 4-4 in New York. And in November 2026, in Bonn, Deep Fritz defeated Kramnik 4-2 — the first time in history that a computer program beat a reigning world champion in an official match.

Thirty-five years after its birth, the name Fritz reappears in a statement containing not one indicator of strength. The distance between those two events is the real story of modern chess.

The engine market has been transformed from the ground up over the past fifteen years. Stockfish, the free open-source engine, is now the benchmark for calculating strength. Leela Chess Zero, built on neural networks, takes a different path but is likewise free and likewise strong. Lichess offers Stockfish-powered analysis to anyone with an internet connection, at no cost. Chess.com has integrated artificial intelligence across its entire learning and playing ecosystem, with tens of millions of active users.

What Fritz once sold — superior calculating strength — is now free merchandise.

That is why ChessBase's statement does not mention Elo. You cannot compete on numbers against a free rival that is stronger. So the company moved to selling something else: experience, a learning path, an interface, the feeling of being coached.

ChessBase is shifting Fritz from an analysis engine toward an integrated training platform, and the absence of published strength figures is not a communications oversight but a strategic choice. It is a reasonable business move. It also raises a hard question: if engine strength is no longer the product, what exactly is being sold?

Reading the description closely, one meets words like "personalized," "efficient," "intelligent." That is marketing language, not technical language. A genuinely personalized training program would specify how it personalizes: through the user's own game database, through repeated-error analysis, through an individual opening model. The statement says none of that.

The second notable point is the contradiction about the audience. The same product is presented to beginners and to professionals. In the chess software market, that segmentation is very clear. Stockfish is for people who know how to use it. Chessable is for people who want structured courses. Chess.com is for people who want a complete ecosystem. Very few products survive by standing in the middle of all those groups.

For Vietnamese readers, this story has its own cross-section, and that is the part I care about most.

Based on my experience following the matches and training sessions of Vietnamese players over more than two decades, I can state something few documents record: in Vietnam, chess software was once an asset, not a commodity. In the early 2000s, a licensed installation at training centers in Hanoi or Ho Chi Minh City was something shared, carefully backed up, passed from class to class. When a seventeen-year-old prodigy does not yet know he is about to become a legend, what he needs most is not a stronger engine but a teacher who knows how to ask the right question.

The generation of Le Quang Liem and Nguyen Ngoc Truong Son grew up under those conditions. Nguyen Ngoc Truong Son became an international grandmaster at fourteen, in 2026 — one of the youngest players in chess history to earn that title. Le Quang Liem entered the 2700 Elo club and won the World Blitz Championship in 2026. Neither of them had a Western-style integrated training platform at the time.

They had fewer tools, and they went much further than those tools allowed. That detail deserves remembering when we discuss the power of software.

Now the next generation — Nguyen Anh Khoi, Tran Tuan Minh, and thousands of children playing chess every evening at clubs across the country — holds more free tools than any generation before. They have Stockfish. They have Lichess. They have an enormous library of lectures on YouTube. What they do not have is a curriculum in Vietnamese, designed for them, by people who understand the domestic tournament system and understand the specific psychological gaps of a young player growing up in a country where chess has never been a sport with money.

This is where the FRITZ 20 story touches something larger than a software product.

If an integrated training platform actually works, it must answer three questions. First, what does it measure? A serious program measures ACPL by position type, by game phase, by opening pattern — not a meaningless aggregate score. Second, what does it fix? A serious program identifies recurring errors over time, not just a wrong move in a single game. Third, how does it progress? A serious program needs a path from one week to the next, not a set of disconnected exercises repackaged behind a new interface.

The FRITZ 20 statement answers none of these three questions.

That does not mean the product is bad. It only means this statement is not a technical document. It is a sales document. And a sales document issued by the manufacturer itself cannot serve as a basis for evaluating a product — no matter how reputable the manufacturer. ChessBase has a thirty-five-year history and a real technical legacy. But past credibility does not substitute for present data.

In sport, we are used to verifying every claim. When a club says its player has recovered from injury, we wait to see whether he starts. When an athlete says he is ready to break a record, we wait for the stopwatch. Chess needs the same discipline. A claim about a "training revolution" should be treated like a claim about form: wait for independent verification before believing.

And there is something more troubling than the missing numbers.

The counterintuitive view here is this: the dominance of free engines does not make paid software obsolete. It makes paid software harder to fake — and that is good news for users.

For fifteen years, the chess world has celebrated the democratization of analysis tools. That joy is real. But there is a blind spot in the collective memory that few are willing to name: democratizing tools is not the same as democratizing knowledge. Giving a twelve-year-old in Hanoi access to Stockfish does not help that child know what to study first, what to fix afterward, and why move seventeen in his own game was worse than move eighteen. The child has an expensive instrument and no music teacher.

What a paid training platform can sell lives exactly there: curriculum, not strength. If FRITZ 20 does that seriously, it deserves the money users pay — no matter how many Elo stronger Stockfish is. A teacher's value is not that he plays better than his student.

But precisely for that reason, the product's audience contradiction becomes worrying rather than merely clumsy. Teaching a beginner and teaching a professional are two fundamentally different design tasks. The beginner needs to be guided step by step, blocked when going astray, given exercises curated to the point of boredom. The professional needs raw data, unlimited depth of self-exploration, a tool that does not interfere with judgment. One interface cannot serve both equally well, and a statement that refuses to pick a side is often the mark of a product that has not found its identity.

There is another possibility, and I believe it is more accurate. ChessBase is not hesitating. It is trying to sell to beginners in the language of experts, and sell to experts with a promise of convenience. That is a familiar marketing tactic in the software industry, and it works only when the buyer has no way to verify. In chess, the buyer always has a way to verify. They simply have not done it yet.

I am not writing this to say FRITZ 20 is bad. I have never run it. I am writing for a different reason, closer to my trade.

In many years as a commentator, I learned that sports fans are led by two kinds of information: numbers and stories. When numbers are missing, stories fill the gap. A statement with no technical indicators gets read through brand trust, through the memory of those Deep Fritz matches against Kramnik, through the feeling that a big name cannot be lying. That is how a marketing claim becomes a collective fact that nobody verifies.

Vietnamese chess has been through exactly that mechanism, only in reverse. We once had world champions whom very few people could name, because no story was told loudly enough. The same void, two opposite outcomes.

This leads me to what I consider the most important point in this whole affair.

After three decades of writing, I have realized I never wrote about sport. I wrote about people forced to choose a move in silence, and about stands that were never full.

The clatter of esports keyboards is also a heartbeat of the modern field. But that heartbeat only means something when behind it stands a training system that knows what it is teaching. FRITZ 20 may be part of the answer, or it may be only a beautiful echo of an era when a software title was itself an argument.

What I know for certain is this. If Vietnamese chess wants a serious place in that arena, we will have to learn to write our own curriculum, in Vietnamese, for Vietnamese children — instead of waiting for an installation file from Hamburg to answer on our behalf. A product can be bought. An education cannot.