Nine Layers of Esports Analysis: When the First Layer Is Empty

A nine-layer analytical document. Thirty-two tables. Seven lines reading "insufficient information to assess." The conclusion compressed into a single sentence: the input payload was empty, and the entire analytical chain is blocked.

I read it at 6:40 in the morning, exactly as the summer transfer feed of the esports world dumped thirty new headlines into my inbox within the hour. A team swapping its mid laner. An organisation selling a regional franchise slot. A patch that completely rewrites how a core champion operates. And behind all of it, seventeen analysis pieces published within twelve hours, each one three thousand words long, each one confident to the point where not a single one left a line for its sources.

The document in my hands is the honest version of those seventeen pieces. It builds the full skeleton: patch tier, tournament system, roster and players, regional picture, club finances, rules compliance, risk profile, media narrative, and the transmission chain of the whole industry. Enough room for ten thousand words. But it fabricates nothing, so it points precisely at the place those seventeen pieces fabricated: layer one.

At three in the morning on 8 July 2026, I also stood on layer one without knowing it. I went on air at a Shanghai radio station and declared that the data said England would lose to Denmark. Denmark ran 118.7 km per match; England ran 112.3. Denmark took 18 shots per match; England took 11. I said it in the voice of a man who had just been vindicated by an entire country three years earlier. The next night, England won 2-1 after extra time, and I received the most expensive lesson of my career: a correct model cannot save a missing data layer.

On the night of the Shanghai derby, I chose the numbers over the entire city. That principle has not changed in nine years. What changed is that I now understand one more layer of it.

Context: the esports analysis industry is producing conclusions faster than it produces evidence

Over the past eighteen months, I have received roughly forty analysis packages from different content teams asking me to cross-check their figures. Eleven of them contained not one verifiable number. Six cited another article, and that original article cited a post that had since been deleted. Two built their entire conclusion on a single match, and in both cases the match was a pre-season friendly. Four asserted that a team had "transformed" after winning one group-stage game, with a sample size of three maps.

Esports has a structural feature that makes this kind of error spread faster than in football. The patch cycle is short. The season is fragmented into stages. A single patch can reorder the power hierarchy of an entire region within two weeks. That rhythm generates enormous demand for information, and enormous demand for information is always satisfied first by speed, and only later by accuracy.

Based on my experience tracking matches and cross-checking data across multiple seasons, the order of error is always the same. The writer locks in the conclusion first, then goes looking for the numbers to defend it. When the numbers cannot be found, the writer leaves the data section blank and keeps the conclusion intact. That is how a document can carry nine full analytical layers and still not contain a single information point.

This season of major tournaments pushes the pressure higher still. The 2026 Esports World Cup in Riyadh announced a total prize pool of 60 million USD, several times larger than any event before it, with team lists spanning more than twenty titles. VCT, Riot Games' Valorant circuit, has run on a franchised model across three major regions since 2026. The 2026 League of Legends World Championship recorded a peak concurrent viewership of roughly 6.4 million, per the organiser's own figures. Those numbers mean every newsroom has to publish something. And when every newsroom has to publish something, layer one is the first thing to get cut.

Layers one, two and three: patch, format and people

Layer one is the patch. On 27 September 2026, Valve replaced Counter-Strike: Global Offensive with Counter-Strike 2, ending a lifecycle spanning more than a decade. An engine-level change of that magnitude does not merely change the graphics. It changes weapon feel, hitbox latency, and the way players read movement rhythm. Anyone writing about a CS team in the three months that followed without specifying the build was writing about a different discipline from the one actually being played.

The same applies to every other title. A patch can push a champion from a 4% pick rate to 30% in two weeks, dragging the entire team-composition structure that surrounds it. The winners are the teams that already had a player proficient in that role. The losers are the teams that have to relearn it in four days. Good analysis at this layer must answer three questions: what did the patch change, which teams benefit, and where does the change sit on the scale — numerical tuning, mechanical adjustment, or playstyle rework?

The seven "insufficient information" lines in the document I was holding all begin here. No game title, no version number, no change list. Without a title and a change list, no layer behind it can function. Every metric downstream is bound to a specific build. A 54% win rate on the old patch does not carry over to the new one.

Layer two is the tournament system and format. This is the most misunderstood layer, because it sits between numbers and instinct. Single-game elimination formats carry a far higher upset probability than best-of-three and best-of-five. Swiss rounds force teams to adapt to the meta faster than fixed-pairing group stages. Double elimination completely changes the cost of a single loss.

The technical document states this plainly: without a tournament name, tier cannot be assigned; without a format, upset probability cannot be assessed; without schedule density, overload risk cannot be assessed. That is a technical admission, but it is also a frightening inventory. How many post-match analyses on the market contain all four of those elements? Of the forty packages I received, three.

Layer three is people. This is the layer where I made the worst mistake of my life, in the Euro 2026 semi-final. Average distance covered and shot volume said Denmark. I forgot that the equation contained one more unknown: bench depth. A team can run less, shoot less, and still win, if at minute 75 they bring on someone capable of producing a moment that the other side has nobody to match.

In esports, that unknown goes by other names. It is the number of players a team can substitute mid-series without collapsing its structure. It is a head coach who can call the right ban-pick in game five. It is a player in the final year of a contract, competing with an entirely different psychological posture. Roster depth, mental load tolerance, and the age curve of individual form are three metrics that appear in no statistics table, yet decide the statistics table.

Nine years ago, I wrote a model built purely on numbers, and it was smashed. Since then, every piece I write has two separate parts: the data section, and the reality check. The second part is usually shorter, and usually the decisive one.

Layers four, five and six: regions, money, and the limits of rules

Layer four is the regional picture. This is the most easily abused layer, because it touches emotion. A region that is strong in one title is not automatically strong in another. A region's standing in League of Legends says nothing about that region's position in Dota 2, Counter-Strike or Valorant. Every title has its own circuit, its own servers, its own calendar, and its own training culture.

Proper tiering must rest on international results within a defined time window, plus the quality of the talent pool and the output of the academy system. Those three do not always move in phase. A region can have an abundant talent pool yet lack a top-tier stage, and the result is that young players flow out to other regions before reaching peak form. Import and export flows are a better leading indicator than the standings.

Layer five is club finance. Here the technical document says something very few analysis pieces dare to say: without at least one figure or one named sponsor, nothing can be evaluated. Salary-to-revenue ratios, franchise-slot amortisation, and sponsor-concentration risk are all calculations that require concrete inputs.

But one detail in that document made me stop for a long time. It records that even when the source article carries an entirely positive tone, the risk review must be done first, and being unable to do it does not mean there is no risk. An empty checklist is not a clean bill of health. That is the sentence I want nailed to the wall of every sports desk.

In esports, that gap is more dangerous than in football. The revenue structure of most esports organisations depends on three sources: sponsorship, publisher revenue share, and prize money. All three swing hard with the cycle. An organisation can win big in one season and lose half its revenue the next if its lead sponsor walks. Stories of unpaid wages, dissolution and slot sales have surfaced across multiple regions over multiple years. No numbers, no story. Numbers without sources are rumours packaged as analysis.

Layer six is rules and governance. This is the layer I care about most, and the one I believe is falling furthest behind reality. The esports betting market has expanded far faster than the regulatory framework governing it. Rules on competitive integrity, transfer and registration, contracts, and player protection have not reached the maturity that corresponds to the money flowing through the system.

The consequence is that match-fixing, result manipulation and in-competition fraud appear in esports at a rate more alarming than in many traditional sports that have had decades of monitoring infrastructure. A monitoring system needs time, historical data and real enforcement power. Esports matured commercially before it matured institutionally. That gap is the opening.

One methodological note: when the governing body has not been identified, the alleged conduct has not been identified, and no official statement exists, then constructing a punishment scenario is a harmful act, not an analytical one. All three scenarios — severe, moderate and lenient — were left blank in the document I was holding, and that was the right call.

Layers seven, eight and nine: risk, narrative and the flow of the whole industry

Layer seven is the risk profile. Every risk item in a standard analytical framework is bound to a specific entity: a specific patch, a specific roster, a specific contract. No entity means no risk item. But the document I was holding points to a kind of risk that is real, measurable and rankable: process-level risk.

When an empty analysis package is passed downstream as a normal input, the recipient can read it as "the article contained nothing notable", when the truth is that nothing was ever extracted. Those two situations lead to two completely different decisions. In the first, people move on. In the second, people have to stop and start over.

In an ecosystem where content decisions, investment decisions and market information all run on input data, confusing "there is nothing" with "nothing has been read yet" is the most expensive class of error, because it makes no noise at all.

Layer eight is media narrative and expectation. This layer needs a sobering dose. Every sport generates narrative labels: new dynasty, succession, all-domestic roster, revenge arc, last dance, comeback. Narrative labels give audiences an anchor. But a narrative label is also a claim about the future, and it must withstand a sample-size check.

The case of Lee Sang-hyeok, known as Faker, is a rare example of a narrative label that has held for more than a decade, because it was built on repeated achievement, and because it never promised anything beyond what had already happened. By contrast, most "new dynasty" labels in esports last less than a single stage. They are attached after two wins and removed after two losses.

The gap between market expectation and objective assessment is measurable, provided both terms exist. Social media heat only means something when placed beside a fundamental base: head-to-head records, rankings, form. Without that base, heat is just noise.

Layer nine is the industry's transmission chain, from publishers upstream, through clubs, tournaments and streaming platforms midstream, down to sponsorship, derivative products and mainstream penetration downstream. Such a chain can only be analysed when a specific trigger event exists: a patch, a policy change, a sponsorship deal, a rights transaction.

Without a trigger event, the transmission chain is just a handsome diagram. And a handsome diagram without a trigger event is the thing this industry produces most of, every major tournament season.

The counterintuitive point: correlation is not causation, and an empty checklist is not a clean certificate

There is a misunderstanding I encounter in almost every conversation with young content teams. They believe an analysis with many numbers is an analysis with a foundation. That is systematically wrong.

If you take a hundred matches from any tournament, you will always find a metric that correlates with victory at a statistically significant level. That is the inevitable mathematical result of testing many metrics, not a discovery about the discipline. A team that wins more usually has more kills, more objectives, more gold, and naturally higher values in almost every column. Writing that such a metric causes victory reverses the causal order. Victory generates the metric. The metric does not generate victory.

In March 2026, I wrote a prophecy. The whole of Germany laughed. I analysed ten of Germany's qualifying matches and pointed out that their average pressing intensity stood at 11.3, far above the 8.5 to 9.5 range of leading pressing sides. On 27 June 2026, Germany lost 0-2 to South Korea and finished bottom of Group F. But I used that prophecy wrongly for years afterwards, until the Euro 2026 lesson hit me in the face. A correct prediction does not prove the model correct. It only proves that on one occasion, the model was not wrong.

Every prophecy carries its own probability of failure. An analyst who does not publish the failure probability of their forecast is not practising analysis. They are practising propaganda.

And here is the second counterintuitive point, the more important one. Failing to find risk does not mean risk does not exist. In the structure of the document I was holding, every blank cell means "insufficient information to assess". We state that in capital letters at the top of the file, because this is the kind of error that can lead an investment fund to pour money into an organisation three months behind on wages, or lead a newsroom to overlook a competitive-integrity case simply because a prior summary was blank.

In esports, this class of error is especially dangerous because entity lifespans are very short. A team can change owners twice in three years. A franchise slot can change hands without anyone outside knowing. A head coach can be replaced in silence. The data gap here is not a random gap. It is a structured gap, and structured gaps are usually where bad things happen.

The analyst's own risk, seen from the inside

I will say this part as a confession.

This profession has a particular temptation: once your model has been proven right in public, you develop a tendency to defend it beyond reason. I have been there. I defended a model based on distance covered for half a year after Euro 2026, instead of admitting that the very same model had ignored bench depth. Fans do not forgive that kind of defence, and they are right not to. An analyst has no right to stand above their own mistakes.

The second temptation is drowning the reader in context. I have made this mistake many times. I was so afraid of missing background information that I stacked a huge block of data at the top of the piece, and the reader left before reaching the conclusion. Context is a load-bearing column, but a column only works when built in the right place. The key line belongs in the first two lines. The detailed background belongs below, for the people who genuinely need it.

The third temptation is the "I told you so" voice. After the night of 27 June 2026, when my article was shared more than fifty thousand times, the feeling of vindication was larger than the feeling of joy. It took me a few years to understand that this feeling was a warning sign. It turns an analysis into a personal contest. I reframed it: whenever my data proves right in front of a crowd, I write about it as a lesson from the data, never as a personal victory.

Every crowd is wrong. The only thing that is not wrong is probability.

Signals for the next cycle

From the Bundesliga to Worlds, I look for the same thing: a truth that can be repeated. Over the next three months, as this major-tournament season moves into the knockout phase, I will track three specific signals.

First, the version question. Any major mechanical change within three weeks of the knockout stage will create a set of beneficiary teams and a set of penalised teams that current standings do not reflect. Comparing ban-pick rates between the tournament build and the practice build will reveal which teams have prepared.

Second, the gap between expected ranking and bench quality. Teams with depth but low standings are where the widest deviations appear in knockout formats.

Third, organisational flows. An unannounced transfer will be announced late eventually, and its value will only show in a later phase, once the contract takes effect.

The spreadsheet is an altar, and I offer myself to every number on it. But an empty spreadsheet is not an altar. It is just an empty room with a nameplate on the door.