What the Engine Cannot Read: Gukesh, Ding and the Blind Spot of Elite Chess

What the Engine Cannot Read: Gukesh, Ding and the Blind Spot of Elite Chess

Singapore, the night of December 12, 2026. Game fourteen of the World Championship match. Ding Liren of China, thirty-two years old, the reigning champion. Gukesh Dommaraju of India, eighteen, the youngest challenger in recorded history. The board had reached a rook endgame so balanced that a club player could have seen the drawing path. Ding's clock ran more slowly than his opponent's. The hall was silent, and I sat in a small studio at Da Nang Radio and Television, a thousand kilometres away, my hand on the fader.

Then Ding pushed his rook to f2. A move that every engine on earth classifies as a serious error within a fraction of a millisecond. Gukesh did not need long. The eighteen-year-old's rook poured in, and Ding lowered his hand, signed the scoresheet, and accepted defeat. The crown changed hands.

On the analysis screen beside me, the evaluation bar leapt from equality to clear loss inside a single second. It was correct. Perfectly correct. And it was so empty that I turned the screen off.

Because that bar answers the question "how many points is this move worth" and never answers the question the entire Singapore hall was asking: why would a man who has played chess for twenty-five years, who sat at the board in Astana and survived the most tense tiebreak in history, place his rook on that square on the most important night of his life?

No metric answers that. Every metric we possess stands outside the question.

Forty years of listening to the board

I thought I had been exiled from chess; in truth, chess carried me deeper into myself.

I was born in India, in a town where chess was taught as a compulsory primary-school subject, and I came to the board not through talent but through stubbornness. In 2026, still a player-organiser, I began writing my first tournament reports. I have watched this industry for nearly four decades: from the era when Elo ratings were calculated by hand and published every six months, through the personal-computer age, through the period when engines became so strong that humans stopped dreaming of beating them, to today, when elite chess is a media industry with data streaming from every move.

In 2026 I left my newsroom after eighteen years because a twenty-seven-year-old editor cut my piece and told me I was writing literature in a sports report. I started a small blog. It drew ten thousand readers in its first week. Since then I have not written match minutes. I begin with an image: a defender's boot touching the ball, a coach shouting into an empty stand, a hand placed on the wrong square.

With chess the image is sharper still. Chess is the only sport in which every decision happens in silence and every consequence is recorded as notation. No impact, no breathing, no grass, no sweat on the touchline. Only a hand, a wooden piece, a long pause, and then a line of symbols in the scoresheet.

That is exactly why chess is the most measured and the most misunderstood of sports.

The language of the metrics

To discuss elite chess today you need a standardised global vocabulary. An Elo rating measures a player's relative strength; higher means stronger against the field. A live rating updates in real time from an ongoing event before official publication. A performance rating is the rating level matching a player's specific result at a specific tournament. ACPL, average centipawn loss, measures the average deviation of each move from the engine's best; lower is better. Engine match rate is the share of moves matching the engine's top choice.

Those four figures, plus the head-to-head record, are essentially the entire dataset a professional commentator holds when explaining a game to an audience.

I have used them for years. And I believe we have overused them.

Take ACPL. An elite player typically records somewhere between ten and twenty in a classical game. Score twelve and people say he played well. Score twenty-five and people say he blundered repeatedly. But imagine a game played under extreme pressure, in a state of sleep deprivation, against an opponent who has beaten you seven times in a row, in a hall where every cough lands in your ear. Your ACPL will be worse. The table records that as a dip in form.

What the table does not record is that it may have been the finest game of your career, measured by the difficulty of each decision.

A scoreboard says nothing about a person

I have kept private notes at major events for years. Not moves — other things: how long a player leaves the board, how many times he drinks water, how often he looks up at the gallery, how many times he straightens his pieces. None of it has ever appeared in a report of mine, but it is what lets me understand what is happening inside a game.

Based on my experience following these matches, I would say roughly seventy per cent of an elite game is time in which no move is played. It is time for thought, for doubt, for persuading yourself that the line you have just found contains no hole. That is precisely the part every metric skips, because machines cannot measure hesitation.

In rating coordinates, world chess divides into clear tiers. The 2700 tier is the world elite, where any open tournament may contain ten such players. The 2800 tier is the historical one: Garry Kasparov, Vladimir Kramnik, Viswanathan Anand, Veselin Topalov, Levon Aronian, Magnus Carlsen, Fabiano Caruana, Maxime Vachier-Lagrave, Wesley So, Ding Liren, Hikaru Nakamura, Ian Nepomniachtchi, Alireza Firouzja. Barely a dozen people in the entire history of the game have touched 2800, and each new arrival is an event.

But what does 2800 say about the person behind it? Very little.

Magnus Carlsen peaked at 2882 in May 2026, the highest in history, and held the world title from 2026 to 2026, defending it four times. Through 2026 and 2026 he remained the strongest player alive by every measure, yet he chose not to defend the world championship. No Elo figure explains that decision. Carlsen had lost interest in preparing fourteen classical games, and chose rapid, blitz and freestyle chess because he enjoyed them.

A man can be the strongest player in the world and simultaneously not wish to be world champion. That makes perfect sense for a human being and no sense at all for a ranking list.

The road to the throne and its narrow doors

To understand how an eighteen-year-old can reach a world title match, you must understand the structure that built the road.

FIDE's world championship cycle is multi-tiered. The Candidates Tournament is a closed eight-player, double round-robin event that produces the challenger. Entry comes through several routes: the World Cup, a vast knockout where a single mistake in one rapid game can erase a year of preparation; the Grand Swiss, a large Swiss-system event demanding eleven rounds of consistency against ever-stronger opposition; the FIDE Circuit, a complex points system only close followers fully understand; and a rating spot for the highest average rating over a defined period.

Each route rewards a different quality. The World Cup rewards knockout endurance and speed. The Grand Swiss rewards stability. The rating spot rewards persistence across many smaller events. And the Candidates itself rewards something else entirely: the capacity to survive fourteen consecutive classical games without losing composure once.

Gukesh came through that road in April 2026 in Toronto, winning the Candidates at seventeen to become the youngest challenger in history. Eight months later in Singapore he beat Ding Liren 7.5-6.5 and became world champion at eighteen.

But Ding's story is the part that makes me pause longer.

Ding won the title in 2026 in Astana, Kazakhstan, beating Ian Nepomniachtchi 2.5-1.5 in the rapid tiebreak after fourteen classical games finished 7-7. It was the first title match in decades without Carlsen, and it took place in a strange atmosphere: both players knew the strongest man on earth had walked away voluntarily.

After winning, Ding vanished from major events for months. He spoke of health, of exhaustion, of losing his love for the game. When he returned, his form was not what it had been. And in December 2026 he entered the defence of his title as a man holding himself together.

Nothing in Ding's file says he was holding himself together. His rating remained in the 2700s. His head-to-head with Gukesh was level. His engine match rate was that of a top player. And he lost.

When the draw rate becomes the audience's enemy

There is a problem elite chess discusses once a year, at world title time, then forgets: the draw rate.

At the highest level of classical chess, draws typically run between fifty and seventy per cent depending on the event. At the 2026 world title match in London between Carlsen and Caruana, all twelve classical games were drawn, and the champion was decided only in the rapid tiebreak. It was the peak of a trend organisers had spent years trying to reverse.

There are technical causes. Opening theory is now so complete that top players can follow analysed lines to move twenty-five without spending significant time. Defensive technique at the top is formidable, making a win from a balanced position extraordinarily hard. And risk is asymmetric: in a round-robin, a loss can cost you the title while a draw always preserves your position.

Organisers responded by shortening classical time controls to force faster decisions, by adding rapid and blitz games to event structures, by introducing Armageddon — where White gets more time but must win while Black needs only a draw — and by running knockouts so that every game eliminates someone.

Those changes work for audiences. They also produce a consequence rarely discussed: they change which kind of player is rewarded. Where rapid and blitz decide titles, fast reflexes and instant calculation gain; deep strategists who need long hours to build a position lose. That is not wrong as sport. But over time it means the definition of a great player is being rewritten by television's commercial demands.

Eighteen years old and the price of growing up too fast

This is the part requiring the most caution, because it touches something I have watched for years in another sport.

Gukesh won the world title at eighteen. Rameshbabu Praggnanandhaa reached the 2026 World Cup final in Baku at eighteen, losing to Carlsen only in the tiebreak. Alireza Firouzja, born in Iran and playing for France, entered the world's leading group as a teenager. Nodirbek Abdusattorov of Uzbekistan won the World Blitz Championship in 2026 at seventeen. Arjun Erigaisi crossed 2800 very young.

From one angle this is magnificent. From another it signals a system pushing unfinished bodies and minds into an adult tempo.

I watched the same thing in football for years, when seventeen-year-olds played thirty professional matches a season and by twenty-five their knees were gone. Chess does not ruin knees. It ruins something else.

An elite classical game can run six hours, sometimes seven. Throughout, a player must sustain intense concentration, calculate lines ten to twenty moves deep, constantly reassess, and endure the knowledge that millions are watching each move on a screen. A world title match is fourteen such games across three weeks, plus press conferences, analysis sessions and media obligations.

An eighteen-year-old body can take that for one match. The question is how many matches it can take.

What stands out is that chess has no protective system for this. No cap on how many events a young player may enter annually. No mandatory rest between major tournaments. No body monitoring the mental health of young professionals across a cycle. The system instead incentivises maximum participation, because ratings accumulate from results and major invitations are awarded by rating.

Ding Liren once spoke of losing his love for the game. He said it after becoming world champion. I think that sentence will be quoted for years, because it describes exactly what a metric-optimised system produces: the best people, at the very peak of their careers, no longer wanting to play.

Who keeps the rules, who keeps the audience

Over two decades, power in world chess has split into two poles that do not fully overlap.

One is FIDE, the global governing body: it writes the laws of play, recognises titles, runs world championships and administers the official rating system. The other is the online platforms, which hold enormous audiences, run commercially compelling online events, and have created new income streams for leading players.

The split generates real tension. A player can earn more from streaming rapid games online than from a multi-week official tournament. A platform can stage an event with a prize fund larger than many traditional tournaments and therefore has more say in which formats flourish. And when a new chess variant — freestyle, with randomised starting positions — is pushed by platforms and a group of elite players, the question of who may declare a world champion becomes ambiguous.

At national level another issue surfaces: federation transfers. When a player switches federations, representation in team events becomes complicated. Since 2026, Russian players have competed under the FIDE neutral flag at many international events, affecting team-event structures, Olympiad places and national squad ratings.

For a commentator, this creates a practical difficulty. When a player sits down, which flag do I introduce him under, which nationality, which title? The answer is no longer obvious the way it was twenty years ago.

Why India and Uzbekistan could do the unthinkable

I was born in India and still follow Indian chess with particular attention, not from sentiment but because it is a development model worth studying.

India won both the open and women's sections at the 2026 Chess Olympiad in Budapest, Hungary. It was India's first time atop an Olympiad, and winning both sections simultaneously revealed squad depth rather than one individual's peak. India's young generation — Gukesh, Praggnanandhaa, Arjun Erigaisi, Vaishali, Divya Deshmukh — arrived as a group, not as isolated cases.

Uzbekistan tells a similar story at smaller scale. At the 2026 Olympiad in Chennai, India, Uzbekistan stunned the field by winning the open section ahead of more experienced squads including Armenia and the host nation. With Abdusattorov and Sindarov, Uzbekistan showed a country does not need hundreds of thousands of players to produce a strong generation — only a focused training system, a few good coaches, and a small group playing each other often enough.

The race has turned. For decades chess was dominated by the former Soviet Union and then Russia. Then came the European era of Wijk aan Zee, Linares and Dortmund. Then the Norwegian era, built on one unrepeatable individual. Now comes the Asian era, with India, China and Uzbekistan sharing the centre.

China took a different path: Ding Liren gave it its first world champion, and Hou Yifan won four women's world titles. The Chinese model rests on state investment in training centres and domestic events plus systematic international exposure. That same model also places national expectation on a single individual — a weight no metric records.

The blind spot: what collective memory does not keep

When a match ends, what do we retain?

The score. The move list. The rating change. The ACPL, the engine match rate, the blunder count. All of it archived in databases that become the reference for every future argument. Ask who was strongest in 2026 and we open a database. Ask who played best at a given event and we open a database.

And we forget everything that could not be entered into one.

We forget that in London in 2026, twelve drawn games did not happen because two players were timid, but because they steered into positions both understood to the last move, where creating a difference at that level was nearly impossible. We forget that afterwards one of them said he did not know whether he wanted to continue. We forget that across those three weeks, some people cried in hotel rooms and some sat at a board until three in the morning preparing the next game.

We forget the human part of the board.

And this is not only true of outsiders. It is true of us, the people whose trade is telling the story.

I thought I had been pushed away from the board; in truth the board is where I have sat longest.

There is one memory I have never put into a report. In 2026, when events ran without spectators, I sat in an empty studio to commentate an online chess tournament. On screen were the board, the pieces, the movement symbols, the evaluation figures leaping continuously. The only other sounds were a fan and my own voice. When it ended I switched off and sat still for a long time.

It was one more game with a full set of metrics. And I cannot remember a single move.

What remains

I am not writing this to say data is useless. Data is indispensable to modern chess, and discarding it is voluntary blindness. Ratings let us compare players across eras. Metrics reveal what the eye misses. Engines let us verify analysis and understand what happened in a complex game.

What I am saying is this: a record with every field filled in can still be empty in every field.

There is one image I keep from that Singapore night. After Ding signed and left the board, there were a few seconds — perhaps only three or four — when Gukesh sat motionless and did not look up. He had won. He had become the youngest world champion in history. And he sat there, eighteen years old, hands on his thighs, eyes on the pieces that had finished their work.

No evaluation bar on my screen displayed that state. No database stored it. No analysis published the next day mentioned it.

And it is the only moment I remember clearly.

If there is one thing chess should do in this major-tournament cycle, it is to build a second kind of memory alongside the database — one that records what cannot be measured: a player's condition, the cost of each decision, the sleepless nights before an important game, the pause after a piece is released. Not to replace the metrics, but so that we know what we are missing.

When the youngest generation in history moves through careers at twice the speed of those before them, what we need is not only a rating updated in real time. What we need is a memory thick enough that when we look back we see not a string of symbols, but the people who wrote them.

Otherwise, decades from now, anyone searching for game fourteen in Singapore will find a flawless string of notation, an exact centipawn loss, and a short footnote about a blunder.

And they will never know that a thirty-two-year-old man sat there for a very long time before he put his rook on f2, and that he knew exactly what he was doing — until he no longer did.

That is the gap no engine can fill. And perhaps that is the part that still belongs to us.