The Games Without Scoresheets: The Trap Inside Vietnamese Chess's Data Gaps

On arbiter desk number four at a national youth chess championship, late afternoon in July, a game ended after 61 moves. The arbiter wrote four lines into the scoresheet: two names, the board number, the round, and the result, 1-0. Then he closed the folder and moved to the next board. Nobody copied out those 61 moves. There was no electronic board, no live transmission, no PGN file saved anywhere. By evening, on the national rating system, the game existed as a small number: the winner gained a few Elo points, the loser dropped roughly the same amount. The result survived. The game vanished.

I stayed behind after play, leafing through three days of scoresheets. Hundreds of games, thousands of moves, and most of them reduced to a few lines of digits. An eleven-year-old boy had just won the best game of his life in round five, and the decisive move will never be seen again. Someone else lost because of a mistake on move 40, and that mistake disappeared with the game. Every year, Vietnamese chess preserves the results of thousands of games and loses almost all of their content.

What troubles me lies somewhere else. Chess is a sport that runs on memory: a good game only truly exists when someone reads it back. When the data gap grows large enough, people start filling it with stories that sound entirely plausible. In sport, a plausible story that never happened is the most dangerous commodity there is.

The top tier has data. The tiers below have silence.

At the highest level of Vietnamese chess, records are not scarce. Le Quang Liem, who won the World Blitz Championship in 2026 in Khanty-Mansiysk (source: FIDE, June 2026), has virtually every game preserved, annotated and engine-analysed. Nguyen Ngoc Truong Son, who held the Vietnamese number one spot for years, also sits inside the coverage of international databases. Names such as Pham Le Thao Nguyen or Hoang Thanh Trang have records complete down to the round, the opponent, and the colour of pieces.

But that top tier represents a very small share of the chess world. Below it lie thousands of games each year: youth events, provincial events, amateur events, national qualifiers, training matches between academies. There are no electronic boards, no data entry team, no one copying moves. Organisers record results to rank players; the national system records results to calculate ratings; and that is the end of it.

Based on my experience following domestic tournaments across many seasons, the share of games preserved in chess notation below national level is effectively zero. Even at national championships, typically only the first few boards in the open and women's sections get live transmission. The rest — sometimes ninety percent of all games — exist for one evening and then dissolve.

This produces a concrete technical consequence. To evaluate a generation of players, you need data about how they play, not merely whether they won. We have outcomes. We lack process.

International events at home, and empty stands

The HDBank International Chess Tournament has been held in Ho Chi Minh City since 2026 and has run almost continuously for more than a decade (source: tournament organisers). It is one of the rare open international events in Southeast Asia that draws both players rated above 2600 and a large field of domestic juniors. Competitively, it matters.

In terms of spectators, it is an empty hall. I have spent many mornings at such events, listening to pieces striking the board, digital clocks ticking, someone pushing back a chair. There is no roaring. There is no wave. Nobody runs onto a balcony to shout when a player sacrifices a piece on move 28.

An empty stand still has an echo — you only need to be quiet enough to hear it. That echo says a mind sport survives on the effort of a small group of organisers, a handful of corporate sponsors, and parents who drive their children to the venue at six in the morning. It also says the sport has almost no mechanism for recording itself.

I once spent time reading through the sponsor lists of domestic chess events. Sponsor names appear on banners, on boards, on standings sheets. They do not appear at the local club where children play on Saturday evenings. That is the distance I believe every sport is now running into, as money follows exposure rather than community.

What Elo measures, and what it does not

Elo is a comparative rating system: a number reflecting results against opponents of different strength. A player who beats a stronger opponent gains more points than one who beats a weaker opponent. That design makes Elo an excellent measure of results and a poor measure of ability.

Elo answers who has won more often than whom. It does not answer who understands chess more deeply. A player who wins because an opponent collapses gains exactly the same points as one who wins with a seven-move combination. The system records results fairly, but it records results, not content.

Performance rating is a variant calculated for a single event. It indicates how a player performed relative to the field over a short window. It is still a measurement of results. To reach the content itself requires different indicators, and this is where Vietnamese records are almost entirely blank.

Average centipawn loss, commonly called ACPL, measures how far each move deviates from the best option an engine finds. The centipawn is one hundredth of a pawn's value. The closer a move sits to the engine's choice, the lower the figure, typically in the low double digits for elite players. This indicator requires move-by-move data.

Engine match rate is the share of a player's moves identical to the engine's first choice. It requires move-by-move data.

An opening novelty is a move that has never appeared in a database at a given moment. Establishing whether a move is new requires a database, and a recorded game to compare against.

Opening preparation is the study of specific variations for a specific opponent. A second is an assistant who helps a professional player analyse and prepare. Both concepts only become meaningful when game records exist, because only then can anyone check whether the preparation was actually used.

Over-the-board chess differs from online chess. Online results cannot be extrapolated directly into classical strength, because playing conditions, concentration levels and anti-cheating controls differ completely.

Every indicator listed above rests on one condition: the game must be recorded. In Vietnam, most games are not.

Judging a generation on too thin a surface

The consequence is not a few missing data files. It is that every conclusion about a Vietnamese generation is being built on a very thin surface: a few dozen games a year from the leading group, plus the memory of those who were in the room.

Memory is biased. It remembers the champion. It remembers the beautiful move. It remembers the win against a strong opponent. It forgets the player who finished fourth, the draw in round seven, the competitor who led until round nine and then lost three straight games to exhaustion.

When a record is silent, that silence is easy to misread. A file containing no controversy does not mean no controversy existed. It means nobody has read anything yet. This is the most common error in any analysis built on missing data, and chess is not exempt from the rule.

I have seen this at international scale. The 2026 affair between Magnus Carlsen and Hans Niemann, and the subsequent lawsuit in a United States federal court, generated thousands of pages of analysis. Much of it was written before FIDE's Fair Play Commission issued its formal findings. Even with extensive public data available, people reasoned well beyond the evidence.

Then, through 2026 and 2026, a former world champion, Vladimir Kramnik, repeatedly questioned multiple online players in public. Those accusations produced prolonged debate, and much of that debate concerned the persuasiveness of the statistical method used rather than any single game. That is the direct consequence of missing clean records, missing cross-checking procedures, and missing an authority that publishes raw data for the community to verify.

In Vietnam, the same question would be far harder to answer, simply because we have no data to begin the argument.

The blind spot of collective memory

Here I want to state a paradox plainly. Vietnamese chess does not lack talent. It lacks chroniclers. And that absence does not discriminate between strong and weak players; it erases evenly.

Over more than a decade of watching, I have come to see that we handle chess memory much like football memory: keeping the moment, discarding the process. A goal can fade, but the applause inside memory never ages. In chess, the only memorable moment is usually the final result line, and so the entire game is replaced by a score.

The real death of a golden generation happens when we stop remembering them. For Vietnamese chess, that risk does not come from natural forgetting. It comes from a flawed process: we have never recorded enough to be able to forget or remember properly.

We hunt for the sound of that old stand and find our own youth inside the noise. But at domestic chess events, there is no noise to hunt for. There are only pieces, clocks, and a folder of scoresheets closed in the late afternoon.

That places a clear professional demand on chess journalists, myself included. When there is no data, the correct choice is not to write with inspiration. The correct choice is to state plainly that we do not know, and then go find a source.

There is a subtler trap, and it relates directly to how automated analysis systems operate. When a processing pipeline receives an empty input, its default behaviour is often to output something that looks valid but is hollow: a correctly formatted report with all the right sections and tables, containing not one line of real content. With humans the phenomenon is more dangerous still. An analyst handed an empty file tends to fill it with plausible-sounding description: a solid opening, a tense middlegame, a technical endgame. Those descriptions are not wrong about chess. They are simply not true of a game that was never recorded.

This is why I treat record-keeping as an ethical professional act, not an administrative formality. Record-keeping determines what can be analysed a decade from now. It determines who will be remembered.

What remains after the folder closes

I do not believe the answer lies in digitising an entire competition system within a single season. Fitting an electronic transmission board to every table costs far more than most domestic organisers can afford. But there is a much cheaper solution: require both players to write out their own game on a paper form, submit it with the result, and have it digitised centrally after the event. That is how amateur tournaments in many countries have done it for decades. It requires no technology. It requires discipline.

Generations will pass, but the poems written about them keep being written. For those poems not to become fiction, the writer needs a record to lean on. And to have a record, someone has to stay behind after play, when the hall lights are off, to copy down 61 moves that nobody else cared to look at.

A game takes exactly one afternoon to be played. Keeping it from disappearing takes one person.