FRITZ 20 and the Data Void in the Chess World

I still keep a stack of old CDs in the drawer of my desk in Hai Phong. Fritz 8, Fritz 10, Fritz 13, one generation per disc. I bought them not really because I needed a stronger engine — free Stockfish had surpassed Fritz in strength years earlier — but because I wanted to see what story these software makers would tell next.

This time it is Fritz 20. The product announcement opens with the words "training revolution," and with a claim that made me put down my pen: it is meant for "ambitious players and professionals," yet also for "players taking their first steps." Reading that, I remembered an old principle of mine from the editing room: when a manufacturer cannot offer a single number, that is exactly when the real story begins.

After more than thirty years in the editing room, I have learned one simple thing. A sports documentary lives on measurable detail, not on adjectives. "Impressive" is not data. "Breakthrough" is not data. And "revolution" — a word worn thin by the marketing industry — is even less so. When I told the story of Nguyen Thi Oanh at SEA Games 29 in Kuala Lumpur, I did not write "she ran very fast." I wrote: 47 strides per minute in the final lap, a number I counted after rewinding the entire footage for three days. That number is what kept viewers in their seats until the last minute.

I trust footage over testimony, because footage does not know how to lie. A chess engine, after all, is also an athlete racing on a 64-square board. And if it claims to run faster than others, it must have a stopwatch.

To understand why Fritz 20's story deserves discussion, we need to look back at the chess market over the past fifteen years.

Fritz is not an unfamiliar name. It is the flagship software line of ChessBase, the German company famous since the 1980s for its game database. In an era when beating a grandmaster was a cultural event, Fritz was one of the most mentioned names. In 2026, in a historic match in Bonn against Vladimir Kramnik — then world champion — Fritz version 10 forced Kramnik to resign with a 4-2 score after six games. That was one of the milestones that brought chess computers out of the laboratory and into mainstream media.

But the world changed. In 2026 and 2026, DeepMind's AlphaZero caused a sensation by learning chess from scratch using only the rules and playing itself, then defeating Stockfish — then the strongest engine in the world — in an unofficial match. Since then, open-source engines like Stockfish and Leela Chess Zero (Lc0) have dominated the strength throne. Today, anyone with internet access can reach analysis power comparable to a supercomputer from two decades ago through Lichess or Chess.com, entirely free.

Integrated platforms like Chess.com, Lichess, and Chessable have turned learning chess into an experience with community, data, and a clear path. A young player in Hai Phong today can run Stockfish analysis on a phone, review their own games with ACPL metrics, study openings from an online database, and play online tournaments every evening. All free or nearly free.

In that picture, ChessBase faces a life-or-death question. When the strongest analysis tool in the world is free, what do you sell? The traditional answer — selling convenience, selling databases, selling an interface for coaches — is wobbling. And Fritz 20 appears as an attempt at repositioning: no longer an engine, but "a training revolution," "a personal chess trainer." That is a notable strategy, but also one very easy to collapse if people ask a single question: what exactly is it?

I read the Fritz 20 announcement three times. Each time I crossed out another line.

Here is what is in it: "you want to improve your chess," "efficient, intelligent, individualized training," "regardless of whether you are just starting or playing at tournament level," "a training revolution." Here is what is not there: no technical specifications. No ACPL metric — the standard measure of an engine's move quality, calculated in centipawns per average move. No move-match rate against leading engines. No performance comparison. No system requirements. No price. No release date.

And here is the most telling point: a chess product announcement with no single number about the tool's strength is an announcement selling sensation, not capability.

I say this not to disparage Fritz. I say it as someone who has spent hundreds of hours in the editing room, cross-checking every frame, to draw one lesson: a claim without evidence is the weakest kind of claim. When I edited the technical segment about Quach Thi Lan at the Tokyo 2026 Olympics — Vietnam's first athlete to reach the 400m hurdles Olympic semifinal with 55.71 seconds — I did not write "she ran perfectly." I counted every rhythm, every stride over the hurdle, and let the number 55.71 speak for itself. Chess is the same.

If Fritz 20 really is a revolution in training, give me one metric: how many ACPL points does an average player reduce, how much Elo do they gain after 100 hours of training with it? That is the question a coach genuinely seeking answers must ask. That is the question any parent considering buying software for their child should ask.

The silence about data in this announcement says more than any adjective. It suggests one of two possibilities. Either the company has no figures to offer — which is worrying for a product claiming to be for professional players. Or they have figures but choose not to disclose them, because the real numbers are not as pretty as the promises. With twenty years of experience reading sports press releases, I lean toward the second. In any industry, when people have a strong number, they show the number. When they do not, they show the adjective.

The second contradiction lies right in the line positioning the target customer.

Fritz 20 is introduced as being for "ambitious players and professionals," meaning those already playing at tournament level, with their own opening systems and serious training plans. But it is also introduced as being for "players taking their first steps." In marketing, targeting both ends of the spectrum is not inclusivity, but often a sign of confusion in positioning. A product serving everyone usually serves no one truly well.

Competitors do it differently. Stockfish does not need to advertise to beginners; it is an engine, a thing rough but astonishingly powerful. Chessable does not try to teach professional players; it builds a structured course system for those who want to learn. Chess.com splits the experience into tiers: beginners get elementary exercises, strong players get deep analysis, each tier with its own interface. Fritz 20, by claiming to serve from first-step beginners to tournament experts, is trying to occupy the same position — and that is precisely its weakness.

A tool that dares not choose its audience is usually one that has not found its own value.

I recall a time years ago, when I helped organize a youth chess tournament in Hai Phong. There was a student who had just learned to play three months before, and one already near 2026 Elo. Both were given the same training software. The newcomer opened it, saw an interface with hundreds of features, and closed it after ten minutes. The 2026 Elo player opened it, saw exercises too basic, and also closed it. That software failed with both — not because it was weak, but because it dared not choose its customer. I worry Fritz 20 is going down exactly that road.

Placing Fritz 20 on the competitive scale, the picture becomes clearer.

In engine strength, Stockfish and Leela Chess Zero dominate, and both are free, open-source, and continuously updated by the community. In training features, Chessable dominates the structured-course segment, with thousands of books adapted into interactive exercises. In integrated experience, Chess.com and Lichess offer everything from online tournaments and game analysis to courses, all within one account. In high-quality analysis, Lichess allows Stockfish to run free on the same webpage, not costing a cent.

ChessBase, and Fritz specifically, has only one remaining advantage: legacy and database. ChessBase's database contains millions of annotated games, a vast learning resource no competitor has intact. But that resource is only valuable if the user is guided through it effectively. And to do that, the product must prove itself with numbers: how much time does the user save, how many points do they improve, compared to self-study with free tools.

The Fritz 20 announcement offers no such numbers. It only speaks of "efficient, intelligent, individualized training." Three beautiful adjectives, but no measure. In the editing room, I have a rule: any scene that cannot be measured does not go into the film. A scene without purpose is a scene to cut. An advertisement without evidence is the same.

Now comes the hardest part — the part I always force myself to raise, because I do not want to merely stand on the opposite side of the crowd. Counter-argument must be applied to myself, not only to others.

What if I am wrong?

What if Fritz 20 really is a good training tool, and the poor announcement is merely a bad marketing campaign? This is a completely plausible hypothesis. I have witnessed many excellent sports products ruined by their own communications departments. A good product with a bad message can still survive; conversely, a bad product with a good message will not last.

So before concluding about Fritz 20, I need independent metrics: reviews from real users, public ACPL comparisons, testing videos from the community. Without those, any judgment — including critical judgment — is a gamble. And I do not want to write a critique just to sound sharp. I want to write a critique that can be verified.

But even if I give Fritz 20 the benefit of the doubt, one thing remains clear. The information gap in this announcement is not a minor incident. It reflects a larger trend in the chess industry: the shift from selling tools to selling experience, from selling strength to selling narrative. When the analysis tool becomes a commodity, value moves to the presentation stage.

ChessBase no longer competes by having a stronger engine — because that is impossible against open-source projects — but by packaging it into a more compelling story. That is why I find this announcement worth analyzing, rather than merely worth criticizing. It is a signal: even a long-established name like ChessBase is struggling to figure out who it is in a market where free tools have flattened every strength advantage.

And here is the most interesting counter-argument.

Perhaps the lack of data is itself a deliberate positioning choice. Perhaps ChessBase is targeting a user group that does not care about ACPL or benchmarks — casual players, people buying software as a gift, people buying the feeling that "I am studying chess seriously." For this group, "training revolution" is more appealing than "78 percent move-match rate." This is a reasonable marketing strategy in business terms, though it runs counter to the transparency standards of a professional product.

But if that is the real strategy, it places Fritz 20 in a fragile position. Because once you sell a feeling, you must continually sustain that feeling. And when users discover that the free tool on their phone is stronger than the tool they just paid for, the feeling will dissolve faster than any number.

In Vietnam, this story has its own dimension. Our chess world has made long strides over the past two decades, with names like Le Quang Liem — once in the world's top 30 — or Nguyen Ngoc Truong Son. But most youth training facilities still rely on free tools. Chess centers in Hanoi, Ho Chi Minh City, and Da Nang teach students with Stockfish and Lichess, because their budgets do not allow buying licenses for expensive software. This places Fritz 20 — and every commercial chess software — before a larger challenge: how to convince a market already accustomed to free that there is a reason to pay.

The answer cannot be "revolution." The answer must be numbers, specific experiences, measurable results. And that is what the Fritz 20 announcement, at least in the version I read, has not done.

I still keep the old stack of Fritz CDs in the drawer. Not because I use them anymore — my computer today runs free Stockfish, faster than anything I dreamed of in 2026. I keep them as a memory of a time when chess strength was something tangible, buyable, touchable. That time has passed.

Every documentary is a race: the audience sees the finish, I live at every starting line. And every chess product is the same — the user sees the promises, while those of us in the profession look for the data behind them.

What remains after reading the Fritz 20 announcement is not an evaluation of the product, because I do not yet have enough data to evaluate. What remains is a question I believe everyone in chess — from coaches to players to manufacturers — should ask themselves: in a world where the strongest tool is already free, where does the real value of chess software lie?

The answer, I think, does not lie in engine strength. It lies in the ability to guide a learner from point A to point B in a measurable way. And to prove that, the manufacturer must do exactly one thing the Fritz 20 announcement did not: provide a number. Because as I have learned after more than thirty years of making sports films, audiences can forgive a bad story. They never forgive an empty promise.