FRITZ 20: When the King of Chess Software Struggles to Redraw Its Own Map

A young coach in Chiang Mai opens the FRITZ 20 product page, reads it from the first line to the last, then turns to me and asks: "How exactly is it stronger than the free version?" I have no answer. The promotional page talks about a "training revolution", a "personal chess trainer", a journey stretching from beginners taking their first steps to professional players. But there is not a single number, not a single specification, not a single comparison. Years of analyzing positions taught me that the empty space matters more than the move. Not the goal, but the gap before the goal appears. And here, that gap sits right in the middle of the ornate lines of a brand that once set the standard for the chess world. ChessBase is no stranger. For more than three decades, it has been the leading provider of chess databases and analysis software, the cradle of Fritz — an engine once synonymous with raw computational power. But the market has changed beyond recognition. Stockfish, an open-source, free engine, now crushes nearly every commercial rival in pure strength. Leela Chess Zero, a neural-network engine, brings a different style of calculation, also completely free. Chess.com and Lichess have turned engine analysis into a default feature embedded in online play. What does that mean? For years, a strong engine was a proprietary asset. Today, it is a commodity. Being good at analysis is no longer enough to make a player reach for their wallet. The question — and I felt this clearly while tracking international tournaments for many seasons — is whether a commercial product still has any room to survive when computational strength is no longer an advantage. Since the moment Deep Blue beat Garry Kasparov in 2026, chess entered an era in which machines are no longer the opponent of humans, but their tool. Thirty years later, that tool has become free. FRITZ 20 appears against that backdrop, positioning itself no longer as a pure engine, but as "a training revolution". This is where the story gets interesting. FRITZ 20 is presented as a total solution: personal, efficient, intelligent training, suited to both ambitious players and beginners taking their first steps. But that very breadth is a suspicious sign. By the standards of a training product, what is needed is methodology. A serious chess teaching tool must answer: how does it correct a player's mistakes, based on which metrics, with what feedback. The ACPL index — average centipawn loss — the rate of moves matching the engine's top choice, progress charts over time: these are things that can be measured. None of them appears in the FRITZ 20 introduction. To be clear: I am not concluding the product is weak. I am pointing out that the claim has no evidence. In any field, and especially in chess — where everything can be quantified — a product that wants to convince professional users must speak in numbers. Here, we only have adjectives. The spatial map never lies — it only exposes what we want to believe. And the map FRITZ 20 draws for its users has a hole: it wants to stand in every position at once. On one hand, it targets "players already competing at tournament level", meaning an audience with a foundation, an understanding of positions, and their own analysis tools. On the other, it beckons to "those taking their first steps" — a group that needs something entirely different: basic guidance, simplification, not deep analysis. These two groups differ not only in level. They differ in needs, in how they learn, in how they make a buying decision. In chess, as on a football pitch, no single system serves both a beginner jogger and an Olympic athlete. A product claiming to do both is usually a sign that it has not yet found its true identity. Looking at its rivals reveals the difference. Stockfish dominates pure strength. Chessable dominates structured learning, with opening courses designed by grandmasters. Chess.com dominates the integrated experience — play, study, analysis in one platform. Each product picks one arena and masters it. Specialization is the only way to create value in a market where free tools have already claimed the foundation. FRITZ 20, positioned as a "training revolution", seems to be trying to avoid a head-on clash over engine strength. That is strategically reasonable. But avoiding one clash does not mean winning another. It only means you have not yet entered the board. In 120 pages of report, I found what the season never records: repetition. Here, the notable repetition is the way ChessBase markets its core product. Not by proving it is stronger, smarter, more efficient — but by calling it a revolution. The language of transformation, when unaccompanied by evidence of transformation, becomes a signal of the opposite. This is the blind spot most readers overlook: a product that refuses comparison often does so not because it has nothing to compare, but because comparison would expose the gap. When a manufacturer is itself the source of information about its product, every adjective must be multiplied by a factor of suspicion. But there is another, more counter-intuitive view. ChessBase moving Fritz from a pure engine to a training platform may not be a sign of weakness, but of awakening. They realized the game of computational strength is over, and they are not the winner. The question is no longer who is stronger, but who teaches better. That is the right direction. The problem lies in the gap between awareness and execution. Realizing you need to pivot is one thing. Proving you have pivoted successfully is another, and that is precisely the missing part. Every lineup is a hypothesis until the ball rolls — and here, the ball has not yet been kicked. We should also remember that the chess software buyer of today is not the buyer of ten years ago. They grew up with Lichess, with YouTube, with thousands of hours of free analysis. They demand proof before belief. A product that wants to survive must speak their language: numbers, comparisons, transparency. When a brand that once defined the standard of chess has to write an advertisement without a single number, that is a signal more telling than any product review. The question for next season is not how strong FRITZ 20 is, but whether ChessBase dares to put out numbers — and whether those numbers can stand against a free Stockfish. The spatial map has been drawn. Now we wait to see if it is real.