Wesley So's fourth US title and a video course: auditing nine information points

Hook

For three decades my habit has been the same: watch from the tribune, come back to the desk, and measure what the eyes missed. This time there was not a single move to measure. A report sat on my desk in Kolkata with a headline claiming Wesley So had beaten Fabiano Caruana in tiebreaks to win a fourth US Chess Championship. I started counting. The report carried nine information points. One was about So and Caruana. The other eight were about IM Valeri Lilov's video course on what Black should play against sidelines of 1.d4. The entire headline claim rested on eleven per cent of the payload. No score, no round, no date, no rating, no tiebreak format.

For a reader that is a disappointment. For an auditor it is data. And when a file cannot carry its own title, the question is not whether the news is true. The question is whose news it is.

Context: where the story comes from, where the money goes

Chess media economics changed after 2026. Before that, a newspaper had one desk — a sports editor, two columns a week, and a tournament made print only if someone fought for it in an editorial meeting. When ChessBase India launched in 2026, news and product merged into a single pipeline. Event traffic arrives, traffic sells courses, course revenue funds the next event's coverage. This is not a conspiracy; it is a model, and the model has a name: news-hook-led native advertising.

When I sent ChessBase India a nine-page column proposal in 2026, I learned that evidence needs a door. Without one, the best audit gathers dust on somebody's table. The lesson has an inverse, and this report is the inverse: when the door is built out of a sales counter, everything that walks through it has to become a package first.

Lilov is not an invented name. He is a Bulgarian IM and trainer, and the product is real. My objection is not to the product's existence but to its packaging — and to the fact that no reader was told whether they were reading a news item or an invoice.

| Category | Information points | Share | |---|---|---| | The event (So–Caruana, US Championship) | 1 | 11% | | The product (Lilov's course) | 8 | 89% |

That table is the whole article. Everything else is its expansion.

Core: two ledgers in one file

The event ledger. I fixed the minimum for a tiebreak report on my Kolkata performance desk at the 2026 Tata Steel Chess India: both players' classical scores, the tiebreak format (two rapids, then blitz, then Armageddon — which one?), the final tiebreak score, the date, and the round-by-round scores. This file has none of them. That is not a small omission. The day the title was decided does not exist in this report.

Four titles is a career aggregate, and aggregates can be checked against the record book. My condition is blunt: a report that puts a count in the headline must put the years in the body. By withholding the list of four, the report left its most checkable number unchecked. That is the first missing table.

The product ledger. Here the substance is real, so the questions are real too.

| Metric | Status | |---|---| | Number of videos | 6 | | Openings claimed | 7 | | Engine-verification claim | None | | Sample video or runtime | None | | Mainline theory (QGD, Slav, Semi-Slav, QGA) | Excluded | | Any sample line or move sequence | None |

Six videos and seven openings means either one video carries two systems or the count is loose marketing. Either way it is a buying signal, because when you purchase a course you are buying hours, not titles — and not one hour is disclosed here.

The real problem sits in the syllabus grouping. The Catalan (1.d4 Nf6 2.c4 e6 3.g3) is filed alongside systems the article itself calls theory-avoidance. The Catalan is contested at 2700+, carries several times the theory burden of a London or a Colle, and has a different strategic character altogether. Listing it here is like putting an SUV photograph in a compact car brochure: it attracts the buyer and misdirects the driver.

The second flaw is move order, and it is the most measurable thing in the file.

| Opening | Principal move order | Frame | |---|---|---| | Colle | 1.d4 d5 | d5 | | London | 1.d4 d5 | d5 | | Blackmar-Diemer Gambit | 1.d4 d5 2.e4 | d5 | | Catalan | 1.d4 Nf6 2.c4 e6 3.g3 | Nf6 | | Trompowsky | 1.d4 Nf6 2.Bg5 | Nf6 | | Torre | 1.d4 Nf6 2.Nf3 e6 3.Bg5 | Nf6 | | Richter-Veresov | 1.d4 Nf6 2.Nc3 d5 3.Bg5 | Nf6 |

Only three of seven genuinely sit inside the 1.d4 d5 frame. That is 43 per cent. The other four arrive through 1.d4 Nf6 move orders. A buyer expecting a coherent 1.d4 d5 antidote repertoire receives four different frames. The advertisement's premise was one question; the syllabus answers a different one.

And the single sentence the pitch stands on — that these systems are extremely dangerous — arrives without one illustrative line. Chess instruction has always measured danger the same way: a sample line, an engine assessment, a database percentage. None of the three is here. A coach who will not show you how strong his line is has told you it was never measured.

In November 2026 I watched six games of Anand–Carlsen from the Chennai tribune across ten days. One image survives: a player's hand hovering over a knight for three or four seconds, then moving elsewhere. By the end of that year I had run a small pilot on fourteen players at the Kolkata Open — mean heart rate 112, peaks of 138 in time trouble, fewer than six hours' reported sleep in nine of fourteen cases. I have attached a physical-cost line to every match report since. This file has no such line, because it contains no hands.

Contrarian: the tiebreak is not the smaller title

The easy complaint is clickbait. Mine is sharper.

The part of the US Championship that was actually decided was the tiebreak, and that is precisely the part with zero data. That is not coincidence. Chess journalism treats anything settled in hours as an appendix, and quietly assumes the shorter format carries luck. For a player whose signature for years has been shorter-format steadiness — fewer errors per hour — a fourth title arriving in that format is not accident but a separate ledger nobody built. I trust the eye for the question and the instrument for the answer; when the instrument has no reading at all, the answer is always borrowed.

The real blind spot is not the advertisement but the buyer. The reader who complains about the funnel is the reader who does not pay for long-form tournament reporting. The seller did not choose the shape of the story; he chose the only shape with existing demand. The coaching market reached the same settlement, which is why it sells plans rather than forced variations — plans do not have to be measured.

Takeaway

From the next file I want four things: the tiebreak format and score, both players' classical round-by-round scores, the four title years listed, and from the course one sample line per opening, the sample runtime, and a plain statement on engine checking. Without them, the fourth title stays a headline and the course stays an invoice.

Next time you open a championship report, ask one question — in which paragraph does this file keep the score?