Auditing the Turlov File: One Wrong Title, Five Adjectives, Zero Chess

I opened the source and counted two columns first. The headline carried a governance claim — FIDE President Timur Turlov meeting Cyprus President Nikos Christodoulides. The body carried five lines of copy for a chess program called FRITZ 20. Two names, one office, one country, one product — and not a single connective tissue between the two halves.

The arithmetic is plain. The headline contains no chess move; the body contains no human name. The headline contains no date; the body contains no engine figure. For seventeen years I have read chess news the way an accountant reads a ledger — the claim on one side, the count on the other. This time the count ended before it began, because the second side was not there to count.

The method was built in 2026 in the Salt Lake Stadium press box in Kolkata, when I hand-charted 240 shots across nine ISL matches and found the team had produced 24.6 expected goals from that volume and scored 18. The 6.6-goal gap went untouched on every broadcast. In 2026 I put all 64 World Cup matches into the same workbook — 169 goals, roughly 1,500 shots, every set-piece routine — and in November carried the template to the Tata Steel Chess India in Kolkata, timing seconds instead of shots. The top half of the field spent 40 percent more clock in the first fifteen moves than the bottom half. In 2026 the club contract lapsed, the ISL folded into a Goa bubble, and I regressed home advantage across 380 matches from four seasons: home win rate fell from 44.8 to 36.5 percent, home goals per game from 1.52 to 1.31. No outlet commissioned that report. I published it myself.

From that same ledger I run four checks on any item. One, verify the office attached to the protagonist. Two, classify the body — news, analysis, or advertising. Three, confirm the claims are falsifiable. Four, locate the source and the date. This item failed three checks outright and declined to answer the fourth.

Check one: the office. The public record places the FIDE presidency with Arkady Dvorkovich, elected in 2026 and returned in 2026. Timur Turlov is a businessman tied to the Kazakh financial group Freedom Holding, whose name surfaces in chess sponsorship. A sponsor and a federation president are two different chairs, and seating them in the same line is the kind of error any regular chess reader catches in the first sentence. Where a headline attaches an office to a name, that office becomes a join key — and on this file the key did not fit the door.

Check two: classification of the body. Five lines call the product a “personal chess trainer, toughest opponent and strongest ally,” announce a “training revolution,” and promise that a user can “train more efficiently, more intelligently, more individually.” Beginner-to-tournament positioning is attached. Those five lines contain not one number: no release date, no system requirement, no size for the opening database, no measure of engine strength, no performance indicator like ACPL. ACPL — average centipawn loss per move — is the cheapest instrument we own for grading quality in chess. It is absent here, which leaves no scale against which the claim can be measured.

Check three: falsifiability. “Revolution” and “strongest” are words that cannot be proven wrong. Not one competitor is named for comparison. The free engines — Stockfish and Leela Chess Zero — have sat at the top of raw playing-strength lists for years; fighting on that front is not a rational commercial move. The FRITZ line's real differentiation has always lived in the interface, the training GUI, the opening database and structured repertoire — the ground where it can charge money. The article names none of it. FRITZ is not an invented name: the ChessBase line has run for decades, and in 2026 Deep Fritz won the match against Kramnik 4-2 in what was billed as Man versus Machine. My point is that the product may well be real; its advertisement is a template.

Check four: source and date. No date in the headline, none in the body. No reporter's name, no FIDE press office citation, no Cyprus government communiqué. In a ledger, this is an expense without a voucher.

Had such a meeting genuinely taken place in Cyprus, the weight of the story would sit in three places: hosting rights for a major event, chess-in-schools policy, and the effect on federation relations. None of the three appears. That is where I think the damage lies — less in the falsehood than in the emptiness. A meeting that was real news left no policy trace in this file.

Now add it up. The headline holds two names, one office, one country, zero chess moves. The body holds zero names, zero numbers, five adjectives. The imbalance is structural, not accidental — a pipeline stitched two unrelated files together. I opened the spreadsheet expecting an answer and found a better question: on what diet of search traffic does an item like this survive? In Kolkata I once saw a scoresheet that held more silence than the sixty-four matches. Here the silence sits in the gap between headline and body.

I state the counter-hypothesis first, because when a pattern appears my instinct is to call it a cause. It is possible a meeting did happen in Cyprus and only the office in the headline is wrong. It is possible FRITZ 20 is genuinely good and the writer simply chose poor language. I cannot dismiss either possibility, because I lack the instruments to test them. So I have to be plain about what I checked and what I did not: FIDE's official site, ChessBase's product page, Cyprus government releases — on none of them did I confirm such a meeting. My verdict is unverified, not false. The distinction matters and it is worth protecting in this trade.

Second, the point I owe against my own side. I sit in India, I have no line into FIDE's press office, and my daily reference is the chess media operation run out of Chennai. That dual position buys an outsider's eye, and an outsider sometimes misses the ordinary thing insiders know. The heaviest evidence in front of me is that this item found no place in that large Indian chess network, which prints dozens of stories a day. Silence is not proof. But when a headline misprints the name of the head of the world's largest chess organisation, that organisation's media wing does not sit quiet — that is an expectation, not a record.

Third. I treat every product claim as a missing value wearing a headline, and a missing value is never entered as zero. Product review and advertising are separate instruments, and in the chess-software market that boundary has been worn thin for years. A real review of FRITZ 20 will exist when someone opens the spec sheet, measures the depth of the opening database, and explains why a club-level player should pay for it in an era of free engines. That review has not been written. What has been written is standing in its place. I trust a model more after it survives a dull, reproducible afternoon; this item never reaches the afternoon.

The industry effect is measurable, and small. The growth axis of chess software is platform-integrated training — Chess.com, Chessable — and AI-coach features, not standalone desktop engines. What this file can deliver to the FRITZ brand is brief, minor visibility; what it delivers to the public image is mildly negative, because a misattributed governance headline breaks trust faster than ad copy builds it. The youth training pipeline registers nothing.

Four tracking signals for the next round. One, watch which name FIDE's official releases attach to the presidency. Two, watch ChessBase's product page for a verified FRITZ 20 spec and release. Three, watch for confirmation of any bilateral meeting from the Cyprus government. Four, trace the domain and metadata behind this item. Any one of them lands and the file can move from unverified to verified, and I will be the first to print the correction. Until then the line stays in the ledger exactly where it is — without a voucher. Sixty-four matches taught me which columns cannot hold a crowd. This file added a new column: the one where an office is written, and which has to be checked before it is read.