The Split Game: Wesley So, Awonder Liang and a Feed Item Out of Place

In a single line, a chess news feed declared that Wesley So and Awonder Liang were on perfect scores at the US Chess Championship, and in the same breath advertised six opening videos by an international master. Two stories, one source code. I read it at nearly three in the morning, and for the first time in years at the board, I did not know whether I was analysing a tournament or a sales page. The number was not wrong. The number was simply standing in the wrong place. That was the moment I understood why I still read data three times before writing, and once more after. The US Chess Championship is one of the strongest national events on the planet, gathering players such as Fabiano Caruana, Hikaru Nakamura, Wesley So and Leinier Dominguez. It is not a direct path to the Candidates — the world-title qualifying route — but a national stage of prestige, where reputation, sponsors and rating are protected at once. The opening rounds carry a kind of data that deceives easily: players on perfect scores after a few games. Wesley So is no longer a surprising name. He is a veteran grandmaster who held a rating above 2700 for years and has been a national champion. A perfect start by So is within expectation, and on its own it is not a newsworthy event. Awonder Liang is different. He is a young grandmaster of the next American generation, and if a perfect start by Liang is sustained, that is the more meaningful signal of the two. Beside that result line, another name appears: Mark Paragua, a veteran grandmaster, beating an opponent recorded only as "Yip". Placed in a wider context, that result line shows a United States still within the top three chess nations by elite depth. Caruana, So, Nakamura and Dominguez are four names enough to make any national event worth watching. But behind them a rising generation is growing, and Liang is one of its clearest representatives. The maturity of this generation comes largely from an online-platform ecosystem, where players both compete and broadcast. That is a different model from how India or China develops talent, and it is also why a start like Liang's deserves to be recorded carefully. And then, within the same source, the body turns. It does not speak of So, nor of Liang. It sells six instructional videos by International Master Valeri Lilov on how to play Black after 1.d4 d5. This is where the data begins to crack. Do not trust the headline yet. Look at the structure. A healthy sports item must have a single spine: either it reports a result, or it presents a product. When both exist in one line, we are no longer reading news — we are reading a data-pipeline fault. The evidence lies in the course's own contents list. Six videos, one plan per video for Black, covering the Colle System, the Trompowsky Attack, the Torre Attack and the London System — along with the Catalan, the Blackmar-Diemer Gambit and the Richter-Veresov Opening. This is not a theoretical invention. It is a revision package of the sidelines White chooses when avoiding main-line theory after 1.d4 d5, specifically when White declines 2.c4. To understand why this matters, one needs to know what those systems are. The Colle System is how White builds a solid set-up with a pawn on d4, a knight on f3, a bishop on d3 and a quiet centre. The Trompowsky begins with 1.d4 and then 2.Bg5, striking at Black's knight on f6. The Torre also sends a bishop to g5 but in a different move order. The London System is famous for its "pyramid" of pawns on d4, e3 and c3 with a bishop on f4, a structure club players around the world carry as a safe weapon. All four share one trait: White avoids main-line theory, choosing certainty over sharp contention. In other words, this product targets club and intermediate players — those who regularly meet the London, Colle and Trompowsky in amateur games, not top players, who rarely face these systems and prepare for them individually. What stands out is how the course positions the threat: it calls those sidelines "extremely dangerous". That is a classic sales hook for a defensive product, not an objective measurement of danger. There is a technical detail I want to linger on. Including both the Blackmar-Diemer Gambit and the Richter-Veresov Opening shows a completist ambition. The Blackmar-Diemer is a sharp but dubious pawn sacrifice favoured at club level. The Richter-Veresov is a low-frequency system. Both have very little practical value for a Black repertoire, while their study cost is high relative to the frequency actually encountered. For a club player, memorising the Richter-Veresov is like preparing a dictionary for a word an opponent almost never uses. On the tournament side, the data is even thinner. No rating, no round count, no opponent quality. "Perfect score" sounds loud, but a run of wins after a few opening rounds is a small sample, not a final result, and by nature a weak predictor of final standing. I have reminded myself of this many times whenever a beautiful set of numbers appears before my eyes. There is one small but notable industry signal sitting in the very part treated as out of place: the existence of a commercialised market for instructional opening videos. This is a mature niche within the chess content economy, where opening courses for Black live on subscription platforms. Its value lies not in inventing new theory but in packaging old theory into an easily consumed product. And in that niche, an over-strong promise is both a sales weapon and an honesty risk. This is where I want to invert an assumption. People often think a bad feed item means a bad source. Not quite. Here, both halves of the item can be true separately: the US Championship is under way, and Lilov's course exists. The error lies in joining them. This is a systemic reliability fault, and it is more dangerous than a single false report, because it does not create false data — it creates true data standing in the wrong place. I once believed in emotion, until a number knocked at three in the morning. But this time the number did not knock to bring good news. It knocked to remind me that even data can be placed in a story that is not its own. There are players who are forgotten, but data never forgets them — and that same data, mis-placed, can make us forget a truth that is present. This leads to another blind spot. An item merging two stories makes readers misjudge both. They may believe So and Liang are dominating an event they may not even be leading, while believing those sidelines are a life-or-death threat to Black. Both are conclusions built on a data sample too small to bear their weight. And one question remains unanswered: who is "Yip"? A player in the women's field, or a data-mapping error? The US Championship usually runs the open and women's events in parallel, so the likeliest explanation is that the result line mixes two brackets. When a result line mixes an open-field grandmaster with a name of unclear bracket, we are looking at the mark of an automated data pipeline short on verification. That is not chess's fault. It is the storyteller's fault. So what is the real signal of the next round? Not whether So or Liang stay perfect. It is whether, when the final standings are published, those two names still stand in a position that forces people to mention them. A perfect start is only a question. The final standings are the answer. And for those seeking how to play Black against 1.d4 d5, this market will not buy the past. It buys what the data has forgiven. An opening course is worth something only when it answers a question the learner is actually facing on the board — not when it calls a threat "extremely dangerous" to make a sale. When the stage is empty, the true value of people begins to speak. With a self-contradicting chess feed, the only thing I can do is take it apart, put each piece back on its own board, and read again from the start. I light a candle for data. But I always let the flame of emotion light the way to the question.