Wesley So's Fourth US Title, and a Headline Where Only One of Nine Data Points Was About the Chess

I counted before I opened the file. The headline carried three names — Wesley So, Fabiano Caruana, tiebreaks. Inside the article there were nine information points. I numbered them the way I number moves. Point one: So beat Caruana in the tiebreak and took his fourth US Chess Championship title. Points two through nine described a video course — eight paragraphs, and not once in those eight did the name Wesley So appear.

One hundred per cent of the headline's claim rested on eleven per cent of the payload. I built this ledger before anyone asked for it, which is why I can write that sentence. Calling it a scandal is easy; in my paperwork it is an ordinary commercial structure — bring in the news traffic, then walk it through the door of a course. My job is not moral commentary. It is to record the count.

The claim about the event itself runs out in a single line. No score, no round, no date, no rating, no game scores. My own register says So had already won this title in 2026, 2026 and 2026; the fourth sits on the same shelf. Caruana's name carries several titles from the same event. The peak ratings in my reference note sit at roughly 2820 for So (2026) and 2844 for Caruana (2026) — both flagged pending verification, because the source recorded no rating at all.

I do not read this pairing as a bogey relationship. It is a mirror. Both are classical technicians; both draw strength from structure and the endgame; both win by slow squeeze. In that kind of match the tiebreak is usually decided by fatigue and the clock, not technique. The US Championship tiebreak ladder runs rapid, then blitz, then Armageddon. A tiebreak result is one data point, not a trend. Reading it as a rating signal is the same mistake I made in 2026.

That summer I was building half-time xG boards for all sixty-four matches of the Russia World Cup on a two-hour turnaround. In the Round of 16 my graphic showed Spain had completed 1,114 passes against Russia — a World Cup record — held 79 per cent possession, scored once from open play, and lost 4-3 on penalties. It was the most shared thing I made that summer and the most misread. Since that day every number I publish carries a paragraph on where it fails. Here is that paragraph — eight of the nine headline points were volume; one was value.

Now the audit of the material that is actually the article's cargo. The course premise is that White players decline 2.c4 to dodge the heavy theory of the Slav, the Semi-Slav, the Queen's Gambit Accepted and the Orthodox Queen's Gambit, and therefore choose systems like the London, the Colle, the Trompowsky and the Torre. That description is accurate. Below 2200 it is everyday reality — time-limited players rationally pick low-theory systems. The author is IM Valeri Lilov, a Bulgarian trainer.

The counting starts. The syllabus lists seven names: Colle, Trompowsky, Torre, London, Catalan, Blackmar-Diemer Gambit and Richter-Veresov. The article says six videos. Seven openings and six videos is a small inconsistency, and it is the most honest quality signal the piece offers. One video may cover two systems; the count may be loose marketing. The difference matters to a buyer, and the article did not resolve it.

The second inconsistency is serious. Placing the Catalan (1.d4 Nf6 2.c4 e6 3.g3) in this list is a category error. It is a mainline contested at 2700+, with its own enormous theoretical tradition. Its theory burden is not lighter, it is differently shaped. Filing it beside the London and the Torre gives a buyer the false expectation that he is purchasing a low-effort system.

The third inconsistency is move-order, and it is the most useful one. The whole course is framed around the 1.d4 d5 structure. Exactly three systems genuinely sit inside it — the London (1.d4 d5 2.Bf4, or 2.Nf3 Nf6 3.Bf4), the Colle (1.d4 d5 2.Nf3 Nf6 3.e3) and the Blackmar-Diemer Gambit (1.d4 d5 2.e4). The other four arrive via 1.d4 Nf6: the Trompowsky (1.d4 Nf6 2.Bg5), the Richter-Veresov (1.d4 Nf6 2.Nc3 d5 3.Bg5), the Catalan (1.d4 Nf6 2.c4 e6 3.g3) and the Torre (1.d4 Nf6 2.Nf3 e6 3.Bg5). A buyer who came for a clean answer to 1.d4 d5 will receive a mixture of four different move orders.

The fourth issue is language. All four systems are called "extremely dangerous," and not one sample line is offered in support. "Black has to be prepared for each of them" is a purchase motivator, not an analytical claim. A pressure phrase cannot stand in for technical demonstration.

This is where Wesley So returns to the file, and where the piece can say something useful. So is the most famous London player in the world. For any club player that is magnetic logic: if the top of the world wins with the London, the London is the low-theory road to winning.

At 2750 the London is not a theory-avoidance weapon. So's London builds closed structures, and then the slow squeeze begins. His conversion comes from endgame technique, and his greatest asset is how rarely he loses. The London reduces his memorisation load and raises his understanding load — which piece to trade and when, which closed structure is winnable and which is not. The London looks easy because So does not make it easy; he makes it hard to beat. The cost of theory shifts from recall to comprehension. It does not fall. It changes shape.

I have seen this pattern before. Through 2026-18 I hand-coded all eighteen of FC Pune City's league matches off broadcast footage, logging every defensive action into a PPDA sheet alongside shot-quality data I built myself. The model showed Pune's high press conceded 0.38 xG per game once the first line broke. I sent the club a nine-page PDF; the staff used two pages. Pune reached the playoffs and lost the semifinal. Payload ratio and payload quality are different things, and that season taught me the difference.

Now the part where I have to doubt my own conclusion. The plain explanation is probably true. Lilov is an IM and a trainer; authors with a coaching profile naturally emphasise plans over forced variations. For a player between 1400 and 2100 that is the correct teaching sequence — understand what the structure asks of you, then memorise lines. The Bulgarian coaching tradition has a long record of plan-based training. My count does not measure the quality of the course; it measures the ratio of the cargo. Treating those as the same thing is the oldest error of my own habit.

The second trap I keep falling into is assuming a link between purchase and improvement. There is no pre/post rating data, no sample video, no runtime, no level rating, no engine-verification claim. That last absence is information in itself: repertoire products that have been engine-checked almost always advertise the fact. Here there is silence.

The third trap belongs to my identity. I was born in the United States and I work an Indian chess beat, and I am permanently tempted to be the man who counted correctly. The record discipline built over decades by the chess desk at The Hindu is the foundation of my work. My numbers are additions to that lineage, not corrections of it.

So what survives in the file? A title that is correct — So won in the tiebreak, his fourth US title — and a body of content on a different subject. Both are intact. They simply arrived in the same envelope.

Three things are worth watching, and I will log them in this column. First, whether the full crosstab of the fourth title is published with tiebreak scores, or whether the headline becomes the record. Second, whether this course publishes one sample line — just one — so a buyer can see what "extremely dangerous" actually means. Third, whether results for systems like the London and the Torre rise at 2700, or whether So remains the exception.

The next time a headline arrives, I will count again. A record does not begin on the day somebody wins. It begins on the day somebody decides to count.