Vietnamese Esports in the LCP: An Empty Spreadsheet Does Not Mean No Problem

Vietnamese Esports in the LCP: An Empty Spreadsheet Does Not Mean No Problem

In October 2026, at the League of Legends World Championship group stage, GAM Esports walked into the final game of their group against Top Esports, the representative of the LPL, one of the most highly rated regions in the tournament. Before that game, almost every public prediction leaned toward the Chinese side. After it, GAM won, and Top Esports left the tournament. For an entire evening, Vietnamese forums were flooded with clips and status updates.

Two days later, I sat down and read everything written about that match in Vietnamese. Plenty of articles. Very few numbers. Nobody published an objective-control rate, nobody calculated the gold difference at minute fifteen, nobody placed that victory inside the tournament-wide probability distribution to see how rare it actually was. We had a historic moment and we let it pass without measuring it.

That is why I am writing this. To talk about what an entire esports scene is missing, and about how that absence is being read incorrectly.

Context: from VCS to LCP

Since 2026, the VCS has been the top tier of Vietnamese League of Legends, running two splits a year with promotion and relegation, and serving as the only gateway to international events. Throughout that period, GAM Esports emerged as the dominant force: multiple VCS titles, repeated appearances at Worlds and MSI. Levi, GAM's jungler, became the name most often invoked when discussing Vietnam's presence on the international stage.

In 2026, Riot Games announced a restructuring of the Asia-Pacific competitive system. From 2026, the LCP (League of Legends Championship Pacific) launched with eight partner teams, merging several previously separate regions, Vietnam among them. It is the largest change this title has seen in Asia since the VCS was carved out as its own league.

The significance cuts two ways. A partner league brings shared revenue, a more stable schedule, and a permanent international stage inside the region. The second, less discussed side is that the analytical bar rises. When your opponents are organisations with dedicated analytics departments, draft modellers, and pathing specialists, the fact that you employ none of that becomes an accumulating disadvantage rather than a single loss.

Over seven years of watching this region compete, I have noticed something: Vietnamese fans watch enormously and remember precisely, but they record almost nothing as data. Memory is rich. The spreadsheet is empty.

The dataset nobody opened

Let us be clear about one thing first: professional League of Legends data is not scarce. Riot Games publishes match data at several levels, and the international community maintains open datasets aggregating tens of thousands of professional games, downloadable as spreadsheets, complete with game duration, gold, minions, objectives, picks and bans. Those datasets are free and have existed for years.

The raw material sits on the table. The question is who sits down at it.

In Vietnamese-language analysis, the share of articles citing downloadable, verifiable data is very low. Most content describes what happened, offers impressions, and draws conclusions from exactly one game. I am not blaming writers. I am describing an incentive structure: emotional content spreads faster than verifiable content, and nobody is penalised for publishing a conclusion without a denominator.

There are matches the naked eye cannot see. The spreadsheet has to tell them.

Take a small, concrete example. A Vietnamese team plays roughly twenty to thirty official games in a VCS split, plus qualifiers and playoffs. That sounds like a lot. But if you want to evaluate one narrow tactical behaviour — the effectiveness of an early lane swap, or the win rate when taking the first Herald — the number of games containing that behaviour often drops to five or fifteen. At that sample size, every conclusion carries a confidence interval so wide it is useless for decision-making.

This is where most Vietnamese esports content collapses methodologically, and it has nothing to do with whether the team is good. It has to do with reading a tiny sample as though it were a natural law.

I once ran a simple exercise: I took every international game played by Vietnamese representatives over three years and split them by game-one result and series result. The question was basic — does winning the first game predict winning the series. The answer depended entirely on which format you filtered for, because format is the denominator, and when the denominator changes, the conclusion flips.

Patches, meta, and the single-game trap

League of Legends updates roughly every two weeks. Each patch adjusts champion strength, items, objectives, and game tempo. Statistically, this means your data has a very short shelf life. A model trained on first-half-of-season data may be obsolete before the season ends.

For large organisations this is an operational problem: they have staff who retrain models each patch. For a scene with few analysts, it is a survival problem — so teams read the meta by eye, by feel, by watching major leagues and copying.

Copying is not wrong in principle. It is wrong in timing. By the time an international team has demonstrated that a strategy works, the next patch has a decent probability of targeting it — the phenomenon the community calls nerfing the dominant playstyle. Copying a composition that already appeared at a major event usually means arriving one beat late.

For Vietnamese teams the story is more complicated, because the region's signature style leans toward early skirmishing and mechanical duels. That is an identity, and that identity has produced real victories. But two very different things must be separated: fighting a lot, and converting fights.

The right metric for this style is not kills per minute. The right metric is conversion rate — what share of kills leads to a major objective: a turret, a Herald, a dragon, or a valuable side-lane push. A team with many kills and no objectives is trading time for decorative numbers. A team with fewer kills where every kill opens an objective is playing more efficiently, even if the scoreboard looks less glamorous.

I built this table for several recent international series. The results, at medium confidence because the sample is small, suggest the gap between Vietnamese teams and deep-running teams is not in kill counts but in average gold difference at minute fifteen and neutral objectives secured in the first ten minutes. The issue is not that they fight. The issue is that most of the fighting is not converted into map assets.

I have to state my confidence level plainly: a few dozen games, spread across many patches, formats, and opponents. Enough to raise a question, not enough to conclude. Anyone asserting certainty from a sample that size is selling you confidence, not evidence.

Spreadsheets do not lie. Readers need to learn how to listen.

Format is the denominator, and it changes the conclusion

This is the least discussed element in every argument about Vietnamese esports, and it matters more than which team plays better.

A single game carries enormous variance. One individual play, one top-lane mistake, one objective contest at minute eighteen can flip a result. Best-of-one pushes variance to its maximum. Best-of-three reduces it. Best-of-five reduces it further. Swiss-format stages combine game types and produce an entirely different denominator from a classic group stage.

Which means: the same result has completely different statistical meaning depending on the format.

Here is a simple calculation to show the scale of the uncertainty. Assume a team's true win rate is fifty per cent — exactly average. If that team plays six group-stage games, the standard error of the observed win rate is approximately the square root of (0.5 times 0.5 divided by 6), which is about 0.204 — over twenty percentage points. An average team can finish a group stage 4-2 or 2-4 purely through luck, and both outcomes sit inside normal variation.

Yet online, a 4-2 team is called a breakthrough and a 2-4 team is called a crisis. Both claims come from a sample too small to distinguish them.

For Vietnam there is a direct consequence: most debate about whether the region improved or declined over recent years is, in effect, debate about noise rather than signal. Separating the two requires a long series of games, normalised by patch and by opponent quality.

From 2026, the LCP creates a structure that makes sample accumulation easier. But the capacity to accumulate and the act of accumulating are different things. Nobody automatically acquires data because a league changed its name.

There is a notable paradox here. In group stages, where the sample is smallest and variance highest, fans deliver their strongest verdicts. In knockout rounds, where the sample remains small but the pressure context is clearer, verdicts tend to be gentler. We are reading the statistical structure of our own sport backwards.

Players, metrics, and the heatmap

The most common tool in community player analysis is the position heatmap. It is attractive, it is intuitive, and it conveys almost nothing.

A heatmap tells you where a jungler spent time. It does not tell you why, what they missed getting there, whether that path correlated with lane state, or whether it correlated with outcomes. That is why I treat heatmaps as a new form of fortune-telling: they hide a player's real role inside a tactical system behind an image that looks objective.

More useful metrics are role-dependent, and applying one role's metric to another is a basic but common error. For a jungler, what matters is first-clear tempo, first-gank timing, and the share of ganks that produce something. For a mid laner, gold difference before minute fifteen and the number of roams into neutral objectives. For a bot laner, damage per unit of gold received — not total damage, which mostly indicates that the game ran long.

There is a structural blind spot here worth stating directly. Public data in this title is excellent at recording minions, gold, damage, and objectives. It is weak at recording vision and movement that produces no immediate result. Which means that if you rely only on public data, you will systematically undervalue the support role, and undervalue junglers who apply pressure without taking kills. An analytical scene built only on open data will inadvertently misprice exactly the positions that are hardest to measure.

I have seen the consequence repeatedly. When a team swaps its support and results do not change, the community concludes the position does not matter. The better conclusion is that the metrics used to evaluate the position are not sensitive enough.

The submerged part: money, rules, and unpublished figures

There is a portion of Vietnamese esports with almost no public data: club finance.

A professional club runs on sponsorship revenue, distributions from the publisher and tournament organiser, transfer or academy fees where they exist, and media activity. Revenue concentration is the single most important risk metric nobody measures. A team drawing seventy per cent from one sponsor carries structurally higher risk than a team with five smaller streams, even though the two look identical in the standings.

No financial statements are published, not even summaries. There are no salary figures. There is no information on sponsorship duration. This gap gets filled by fan speculation, and speculation always skews optimistic in winning periods and pessimistic in losing ones.

One publicly available source is notably underused in Vietnamese content: the player contract database Riot Games publishes for major leagues. It lists player names and contract durations. It lets anyone check whether a roster is about to expire or is committed long term. Very few domestic analyses cite it, even though it is free and takes thirty seconds to look up.

On the rules side, this title operates across multiple layers: publisher competition rules, regional league regulations, and provisions covering contracts, transfers, and the protection of minor players. When disputes arise, most fans only ever see one party's unilateral statement. There is no public record to cross-check against.

The point I want to stress: missing data is not evidence of health. If you have no figures on risk, you do not know that risk is zero. You simply do not know. The difference between those two statements is the entire content of this article.

A stray number can be a truth hiding where nobody looked. But a blank cell is not a number, and it is not hiding anywhere. It is just blank.

Transmission: from publisher down to the small stage

To understand why these gaps matter, esports has to be viewed as a transmission chain.

Upstream sits the publisher, deciding patches, calendars, league structure, and the commercial value of the title. Midstream sit clubs, tournament organisers, and streaming platforms. Downstream sit sponsorship, derivative products, and the mainstreaming of esports into general culture.

Every upstream change sends an impulse down the chain, and it deforms at each link. Consolidating regions into one partner league is one such impulse. It raises the competitive level, raises the potential audience, and raises the operational capability required.

In this chain, Vietnamese clubs sit in a particular position: they receive upstream impulses quickly, but their capacity to absorb and convert those impulses into competitive capability depends on something that cannot be bought in one season — analytics staff, coaching systems, and accumulated data.

The streaming platform is a link worth noting separately. In Vietnam, most viewership is distributed across social and video platforms where viewer figures are not independently audited to the same international standard. That makes a very basic question — how many real people does this league attract — hard to answer with comparable numbers. Sponsors want that number. Analysts want it too. For now both receive sets of figures that are methodologically inconsistent.

The contrarian angle: what we lack is not better numbers

Here I want to argue against the standard reflex.

The standard reflex says Vietnamese esports needs better data. I think that is only half the problem. The other half, and the more uncomfortable one, is the habit of reading the absence of data as a neutral or even positive signal.

When nobody publishes salaries, the default is that everything is fine. When nobody publishes contract durations, the default is roster stability. When nobody publishes sponsorship structure, the default is healthy finances. None of those defaults rest on evidence. They rest on the fact that no scandal has yet been loud enough to make the news.

For a scene that has already seen several generations of players pass through, the absence of a data system also creates a memory problem: there is no way to compare a 2026 player with a 2026 player, because the measurement standards of the two periods differ and nobody has normalised them. Every cross-era comparison in Vietnamese esports today is, in essence, a comparison of feelings.

Let me push this further. While the whole scene debates the lack of advanced metrics, a far more important investment is being entirely neglected: grassroots coaching education.

Famous retired players open academies, and most of these are commercial vehicles — a way to monetise a name, not a pedagogical system. Meanwhile there is no structured programme to train people to teach esports at the grassroots level: people who understand age-group psychology, coaching methodology, and how to build a long-term curriculum. As a result, the quality of youth development depends on whether a fourteen-year-old happens to meet a good teacher.

This is a data problem, not a moral one. You cannot measure the effectiveness of a development system without standardised inputs and tracked outputs. And you cannot compare two academies when neither publishes anything.

They told girls not to talk tactics. I drew a chart instead of answering.

Signals to watch next split

I am not predicting standings. Standings are outcomes, and outcomes carry high variance. I track input signals, because they forecast better and move more slowly.

The first signal is analytics staffing. A Vietnamese team adding a full-time analyst, or contracting an external data provider, is a far stronger signal than that team winning three games in a row.

The second is contract stability. The public contract database allows anyone to check how many key players at the top teams remain under contract beyond next season. A scene that retains its players across multiple seasons is accumulating, not rebuilding from scratch every year.

The third is disclosure. If a single Vietnamese team published its internal match data in anonymised form, or issued a seasonal report against a standard metric framework, the region's entire analytical baseline would shift within a year. That is the cheapest change with the largest leverage I can think of.

The fourth is youth development output. Not the number of students, but the number of certified coaches and the number of years they stay in the profession.

I know these signals are less exciting than a standings table. I also know they decide the standings table three years from now.

When I forecast, I do not look at emotion. I look at the indicators everyone else skips. That is why I am not predicting Vietnam will win the LCP next season, and why I am not ruling it out either. Both conclusions require the same thing we do not yet have: a dataset long enough to speak for us.

The only thing I am certain of is this. A blank cell in a spreadsheet never fills itself. It fills only when somebody decides that not knowing is a problem to be solved, rather than a normal state to be tolerated.