The Nine Data Layers of an Esports Analysis: A Beautiful Skeleton, an Empty Core

I received an esports analysis nine pages long. Every page had a heading. Every section had a table. Every column had a label. Not one data cell contained a number.

The "Patch and Meta" section read: insufficient information. The "Teams and Players" section read: insufficient information. The "Risk" section read: insufficient information. Nine layers of analysis, nine blanks. The report was not wrong in a single word of its shape. It was so structurally correct that if you skimmed it, you would think you had just read a professional document.

That is what I want to talk about today.

In Vietnam, esports draws an audience that ranks among the largest in the region. The VCS has run a steady schedule since 2026. The Arena of Valor circuit is tied to names like Team Flash, champions of AIC 2026 and AIC 2026. GAM Esports, with Đỗ Duy Khánh (Levi) in the jungle, played MSI 2026 on home soil in Hanoi and returned to MSI in 2026. But the data infrastructure behind that audience is still thin. Most Vietnamese-language esports content stops at the commentary layer: who won, who looked good, who deserved it. Commentary answers the question "what happened." Analysis has to answer a harder one: why is that rate 62 percent rather than 50, and how many more matches are needed before 62 percent means anything at all.

I started this work in a rented room in Nha Trang in 2026, logging V-League metrics by hand, four hours per match. I wrote a blog from a rented room in Nha Trang; now probability takes me everywhere. That experience taught me one thing: the hardest part of sports analysis is not reaching a conclusion. It is knowing what you are missing.

The nine layers in that report form a correct skeleton. I will walk through it, but differently: each layer is a question, and if you cannot answer it, every conclusion stacked above it is decoration.

The first layer is the patch. This is what separates esports from football most clearly. Football changes its rules once a season, and slowly. Esports changes mid-season. A champion can move from a 47 percent win rate to 54 percent after a single update, and the team that happens to employ the player who mains that champion suddenly gets stronger without anyone training an extra hour. Without patch data, every claim about "form" is a guess. You are praising a team for improving when in fact they simply caught a good patch.

The second layer is tournament format. BO1, BO3 and BO5 are three different sports in probabilistic terms. In a BO1, variance is wide enough that the weaker team wins roughly one time in three. In a BO5, that number shrinks to about one in ten. Upper bracket, lower bracket, Swiss, round robin — each format produces a different distribution. People like to say a team "shows up on the big stage," as if that were a mental quality. Most of the time, it is just a longer series.

The third layer is the subject: teams, players, coaching staff. Based on my experience tracking matches, I hold a position I have kept for years. Transfer valuation models overprice young potential and underprice locker-room chemistry. An 18-year-old with a clean top-lane stat line can be valued level with a champion, but what decides a team's results over the next six months is whether five people will call for each other in a teamfight. Chemistry appears in no stat table. That is exactly why it is mispriced.

And here is the mathematical problem you cannot route around. A professional League of Legends player plays about thirty matches a season. Thirty matches. To distinguish a player whose true win rate is 55 percent from one whose true rate is 50 percent, thirty matches is not enough. Every player ranking you read is ranking on a sample far too small for its own uncertainty. The match ends, but the data remains — and that data is often screaming that it has nothing to say yet.

The fourth layer is regional context. Comparing the VCS with the LPL or the LCK is a story about density of opposition more than a story about talent. A VCS team plays roughly eighteen matches a season against opponents of comparable level. An LCK team plays three times that many against comparable opponents, plus high-quality internal scrims. The gap between regions grows inside practice sessions nobody streams. Reading international results to measure that gap means reading only the visible tip.

The fifth layer is finance. The loan-with-obligation-to-buy mechanism sounds fair, but it pushes risk toward small clubs. A small club takes on a player, pays the salary, develops him, starts him, and then at season's end must choose between buying him outright with money it does not have or returning an appreciated asset to a big club. The result is that small clubs spend the year producing finished goods for the wealthy. A one-way financial flow disguised as partnership.

The sixth layer is rules and governance: player age, contract length, transfer conditions, and disciplinary rulings. This layer rarely comes up in commentary, yet it decides who is allowed on stage. A six-month suspension can reshape a season more than any patch.

The remaining three layers — risk, public narrative, and industry transmission — are the ones that nine-page report also left blank, and I do not blame it. They are hard. The narrative layer requires you to measure crowd expectation and compare it with reality. The transmission layer requires you to look from publisher down to streaming platforms, down to sponsors, down to derivative markets. Those are questions for an industry, not for a single match.

An analysis earns trust only when the writer can say which layer was skipped, and why. Skipping because data is missing is honest. Skipping out of laziness is dishonest. Skipping without saying so is worst of all.

What brings me back to that nine-page report is not that it was empty. It was empty in a highly organized way.

There is a trap in analysis that I have fallen into more than once: the more complete the framework, the easier it is to believe you understand. A handsome skeleton manufactures false confidence. A reader skims nine headings, sees tables, sees terminology, and automatically assigns more credibility than to a short piece with nothing in it. That report fooled at least one person, and that person was almost me.

There is a subtler error I have to remind myself of every week. Correlation is not causation, and in esports the causal order is often reversed. Winning teams tend to have high objective-control rates, high gold rates, high kill rates. It is very easy to write a piece claiming objective control is the key to victory. But winning teams control objectives because they are winning. That metric is a consequence of advantage, not its cause. When you read an analysis claiming a team "wins because of metric X," ask yourself which other hypothesis would explain the same data.

And there is one thing I do not want to skip: fan emotion. It is usually treated as noise. To me it is a variable. When a Vietnamese team exits an international event at the group stage, the community reaction is still data — data about expectation. And expectation, measured properly, is one of the best early indicators that a team is being mispriced.

People call me a number-obsessed type; I take that as a compliment. But I am not obsessed with numbers. I am obsessed with what the numbers are trying to say.

That nine-page report will be filled in someday, and when it is, it will be a good document. The skeleton is there. What remains is data, and data in Vietnamese esports is being generated every week by people who do not know they are generating it.

The signal I am watching in the next round is not which team wins. It is whether anyone starts writing down the thing that next week nobody will remember.