The Empty Cell on the Data Sheet: Why Esports Needs an Analytics Culture That Can Say 'I Don't Know'

In the summer of 2026, inside a press room in central Seoul, I sat next to a Korean colleague who worked for an independent analytics organization. On the big screen, the organizers had just announced a figure for the peak concurrent viewership of the final. That number was shared by hundreds of accounts within minutes. What caught my attention was not the number itself, but my colleague's question: 'Where does this number come from?' Nobody in the room could answer. I myself, after years of working in club finance and fan data analysis, could only say that the figure 'seemed reasonable'. That was the moment I understood something I have carried throughout my career: most public esports data circulates because it sounds plausible, not because it has been verified.

This article is not meant to accuse anyone. It is meant to dissect a structure. After nearly two decades of observing this industry — as a player, a tournament organizer, and a club financial analyst in Boston — I have concluded that esports' biggest problem is not a lack of money, fans, or tournaments. Its biggest problem is that the entire media, sponsorship, and valuation system has been built on a foundation of data with gaps nobody has ever measured.

I want to tell this story as someone who has sat inside press areas, retyped spreadsheet after spreadsheet, and had to admit to leadership that there was not enough data to conclude. It is an uncomfortable story, but a necessary one.

When a number is born without a midwife

Let us begin with something very concrete: the peak concurrent viewership of a major tournament. This is the most quoted metric in commercial press releases, sponsorship pitches, and club valuations. It appears everywhere. But where does it come from?

In most cases, it comes from third-party measurement platforms such as Esports Charts or Streams Charts. These platforms collect data from the public APIs of Twitch, YouTube, and a handful of regional platforms. They add it up, average it out, and produce a peak figure. The problem is this: they cannot measure viewership on China's domestic streaming platforms such as Bilibili, Douyu, or Huya, which together form the largest market for League of Legends. Even when they estimate, it is still an estimate, not a measurement.

Yet in sponsorship negotiations, that figure is still presented as fact. I once watched a sponsorship deal inflate in value simply because one side presented a viewership comparison table the other side could not verify. Nobody was lying. Nobody was checking either.

This is the crux I want to stress: the danger of esports data is not that it is wrong, but that it cannot be verified — and this industry has grown used to that. Once unverifiability becomes the norm, any number can be born without a midwife.

Context: an ecosystem of three misaligned data layers

To understand why this happens, you have to understand who operates esports data, and for what purpose. I divide it into three layers.

The first layer is the publisher. Riot Games, Valve, Blizzard, Tencent — each giant owns part or all of a game, and therefore owns the source data. They provide APIs to third parties, but always with limits. Riot, for example, tightly controls data on player behavior and internal revenue. Valve is relatively open with Dota 2 match data, but almost closed on business data.

The second layer is independent measurement. This is where Esports Charts, Streams Charts, and a handful of smaller platforms operate. They are useful, but they only measure what is measurable: Western-platform viewership, broadcast hours, co-stream counts. They have no access to source game data, no ticket sales, no merchandise sales. In other words, they measure the visible tip of the iceberg, but the big business decisions rest on the submerged mass they cannot see.

The third layer, and the most neglected, is the internal layer of organizations. This is where data truly lives: sponsorship contracts, ticket revenue, salary costs, internally analyzed performance metrics. But this layer is almost never published, except during a crisis. As a result, most of what the public knows about the esports economy is external guesswork.

These three layers do not align. Layer one holds source data but does not share it. Layer two measures what it can but cannot reach the source. Layer three knows the truth but does not speak. Fans and investors sit in the middle, forced to assemble a picture from fragments. And when no fragment fits, the industry tends to display the prettiest one.

Based on my experience watching matches and announcements over the years, I have found that fans approach esports data in two ways: either absolute belief in the number presented, or dismissal of all of it as fabricated. Both are dangerous. The right way is to distinguish signal from noise, and to accept that some questions simply do not yet have answers.

Core analysis: dissecting three types of empty cells in esports data

This is the section I want to dwell on most, because it is where I have made mistakes and where I have learned the most. Not all missing data is the same. I divide it into three types, and each requires a completely different response.

The first empty cell: data never collected

Some things have never been measured, simply because nobody considered them important. The clearest example is player mental health data. For years, esports operated as if players were machines without nervous systems. Major organizations began hiring sports psychologists roughly from 2026 onward, but to this day there is almost no standard dataset on the relationship between congested schedules and long-term performance. We have schedules, we have results, but we do not have health.

I once proposed building a 'cognitive load' index for a youth team I advised. The idea was to measure quality practice time versus forced practice time. No database existed. We could measure hours in front of a screen, but not the submerged part. As a result, every conclusion about a young player's form curve was a story, not science.

This is the point I want to stress: missing data is not useless; it is a map pointing to where nobody has measured. When you find an empty cell, you should not fill it with speculation. You should mark it and ask why it is empty. The answer usually tells you where the industry's value is being shaped with no oversight.

The second empty cell: data collected but with the wrong unit of measure

A more dangerous type. An example is 'effort' metrics in player analysis. We have grown used to looking at distance covered and sprint counts and calling them effort indicators. But as I have said many times in internal analysis sessions, ineffective running also produces impressive distance. A player who moves a lot may be running chaotically because he cannot read the rhythm of the match. A player who moves less may be controlling space better.

The problem is not the number. The problem is that it is packaged as an effort metric without a proper unit of measure. Effort is multidimensional: it includes intensity, timing, and effectiveness. Distance measures only one of the three. Yet in transfer reports, distance covered still appears as an evaluation criterion. That is a unit error disguised as a metric.

I remember a transfer meeting where a coaching staff member presented a list of players ranked by fight participation rate. The list looked persuasive. But when I asked on what basis that rate was calculated, the answer was that it came from a community stats site, not from the publisher's source data. We were about to spend millions of dollars based on a table of unknown origin. We stopped and started over. The time invested later saved the club a much bigger mistake.

The third empty cell: data collected correctly but interpreted against the wrong denominator

This is the subtlest type, and the one I have personally committed. It happens when you have a correct number but compare it against an unsuitable sample. For example, when evaluating a young player, you compare his performance in a short tournament against the performance of stars across a dense season. The denominators are entirely different. A BO1 tournament differs in nature from a round-robin season.

Another common error: opponent adjustment. A player who posts high metrics against weak teams will not necessarily post the same metrics against strong teams. But public stat leaderboards often lack opponent-strength adjustment. As a result, players who shine in easy matches are pushed to valuations above their true worth.

During a period when I led transfer strategy for a club, I built an opponent-strength-adjusted model for pressure metrics. The idea was: if a player sustains high pressure metrics even against strong opponents, only then is he genuinely valuable. That model was imperfect, but it helped me eliminate several names that public leaderboards were celebrating. I believe what we call a 'star' is often just someone who appeared exactly when the system needed them — and that system, in most cases, is an unverified measurement system.

The power structure behind the numbers: who benefits when data is blurry?

Now step back and look at the structure. If esports data is this blurry, who benefits?

First, organizations that need to raise capital. A club seeking investors has an incentive to present the highest possible viewership. This is not exactly lying — they simply choose the most favorable numbers from an imperfect source. But when every club does this, the industry's shared standard rises, and nobody can bring it down.

Second, the measurement platforms. They provide a useful service, but their business model encourages impressive numbers. A leaderboard where every tournament has enormous viewership gets shared more than an honest but less flashy one.

Third, the publishers themselves. They have an interest in their ecosystem being perceived as growing. This does not mean they manipulate data — only that they have no incentive to publish figures showing decline.

Fourth, and this is what I want to state clearly, is the media. Esports articles, including respectable ones, often cite numbers without verifying their origin. Not out of laziness. It is because verification takes time, and esports media operates far faster than verification allows. When speed is priority number one, accuracy is the first thing sacrificed.

This creates a loop. A number appears in one article. It is cited in another. After a few rounds, it becomes 'common fact'. Nobody remembers the original source. And once it is common fact, challenging it becomes eccentric. Every transfer bubble starts with a beautiful story and ends with a balance sheet. This is as true of esports as of football or any other entertainment industry.

A counterintuitive angle: the limits of analysis are not its failure

This is the part I want to say plainly, because it runs against most esports readers' expectations. We are used to the analyst as someone with answers to every question. On talk shows, in prediction pieces, in transfer breakdowns, the analyst is expected to deliver conclusions. Silence is read as weakness.

I believe the opposite is true. The true value of a deal only reveals itself when the market stops making noise. And to wait until the noise dies down, the analyst must accept that while it is loud, he does not know. He must endure the discomfort of having no answer.

There is enormous pressure in this industry to conclude early. When a young player shines in a tournament, someone immediately calls him 'the future of esports'. Such statements are not wrong emotionally, but they violate denominator discipline. A tournament is a small sample. A season is a larger sample. A career is an entirely different sample. Blending these three levels is the most basic error in esports analysis, and it happens daily.

I was once in a situation where I had to choose between giving a prediction I did not believe and admitting I lacked enough data. I had pushed to pursue a transfer target across three windows, building an analytical framework so thorough that I forgot time itself is a variable. As a result, we lost the player to another club within 48 hours. My board told me something I have carried since: a perfect model never exists, and being on time is also a variable.

That lesson does not contradict what I am saying. It complements it. If you lack data, you must decide with the information you have. But if you have data and you know it is insufficient, you must not invent the missing part. Those are two different situations, and distinguishing them is the core skill of a mature analyst.

I want to extend this to an area esports rarely discusses: the social impact of bad data. When a player is overvalued based on a misunderstood metric, the consequence is not just a bad deal. It is tremendous pressure on a young person, expected to live up to a number that does not reflect his ability. I have watched inflated players be criticized, then disappear from the system. Looking back, the fault was not theirs. The fault was in the measurement system.

That is why I believe in investing in analytical infrastructure over flashy investments. A club can buy a star, but a club that builds a talent-detection system will never be left behind. A system does not create genius; it only creates space so genius is not stifled. And in esports, where a player's career span is far shorter than in traditional sports, creating that space matters more than anything else.

What esports data can say, and what it cannot

Having criticized, I want to be fair. Not all esports data is bad. Some parts of this ecosystem are well run and deserve recognition.

Game-level match data is fairly strong. For Dota 2, Valve provides match data detailed enough for the community to build deep analysis. For League of Legends, post-match data from Riot has become far richer than a decade ago. For Counter-Strike, platforms such as HLTV have built a relatively credible data standard, with metrics like Rating — though these too are constantly updated and debated.

But there is a large gap between match data and business data. And that gap is where most valuation errors occur.

Take franchise fees. When major leagues moved to franchising, the numbers presented were impressive. But those numbers are usually initial slot purchase prices, not the true value of the organization. An expensive slot does not mean its owner is expensive. This is a common confusion among new investors. A slot's value depends on the cash flow it generates in the future, not on the amount paid to acquire it.

Another example is media-rights revenue. For years, esports was expected to follow traditional sports, with huge rights contracts. Reality is more complex. Because games are owned by publishers rather than independent federations, negotiating power sits on two different sides. Publishers have an incentive to distribute content freely to grow the player base, while leagues want exclusivity to sell rights. This tension has never been fully resolved, and it means esports media-rights numbers cannot be directly compared with traditional sports.

I once spent three weeks building a cost-benefit model for a potential sponsor, based on data collected at a major tournament. After three weeks, I decided to abandon it. Not because it was wrong, but because the dataset was too small to guarantee reliability. That was one of the correct decisions of my career. I learned that abandoning a model when data is insufficient is a professional act, not a weak one.

Crisis as a demolition contractor

There is an aspect of the data problem I want to view from a long-term angle. Crisis is not the industry's enemy; it is the demolition contractor for what has already rotted. Esports has weathered at least two major shocks in the past decade: the collapse of some North American leagues and the stagnation caused by a global pandemic. Both made some organizations disappear. But they also exposed data gaps that had previously been hidden.

When leagues were cancelled in the pandemic's first year, clubs had to decide whether to keep or release players. Clubs with clear fan-retention data and contract structures could decide with less damage. Clubs running on inspiration collapsed. The difference was not the money in the bank, but the quality of the internal dataset.

As the person responsible for a club's financial model, I proposed three contract-restructuring scenarios. We saved a considerable amount, but we also sold one of our key players due to internal conflict. I spent four months convincing leadership that the long-term consequences of selling that player were more serious than the short-term savings. That was a lesson in how financial data, however accurate, cannot replace judgment about people and intangible value.

This is where I want to talk about intangible assets — what inexperienced operators often overlook. The value of an esports organization is not the money in its account. It is in things hard to measure: fan loyalty, organizational culture, talent-detection ability, sponsor relationships. These assets often do not appear on the balance sheet, but they decide whether an organization survives the next season.

The problem is that intangible assets cannot be managed without being measured. And to measure them, you need something esports lacks: a sufficiently refined fan-behavior data system. Not viewership counts, but engagement depth, reasons for leaving, and the lifetime value of a fan. Today, almost no mid-tier organization has enough data to answer these questions.

What I learned from a player no leaderboard could see

I want to tell a story to illustrate everything I have said. Not a story about esports, but a story about method — a lesson from football that applies to esports.

During a major tournament cycle, I built a database tracking young players with low minutes but high pressing metrics. I found a young midfielder at a small club. He appeared on no leaderboard. I wrote a long report and sent it to three big clubs. Only one replied. Two years later, that player moved to a top European league, and my report was recognized as visionary.

The point is not that I was clever. The point is this: he appeared on no leaderboard not because he was not good, but because those leaderboards were built on the wrong denominator. They measured players at big clubs, in big tournaments, with big minutes. Players at small clubs, with few minutes, were excluded from the sample — even when they had qualities the big clubs needed.

Missing data is not useless; it is a map pointing to where nobody has measured. In esports this is especially true. Countless talented players compete in regional leagues, on teams the media does not cover, with data no major platform collects. They exist. The system does not see them. And every time a major organization buys an established name for a high price, it means they are paying for fame built by an imperfect measurement system rather than for genuine ability.

That is why I always stress timing and opportunity cost in transfer analysis, rather than only a player's technical value. A perfect model that arrives too late is a useless model. Meanwhile, an imperfect model executed on time can change an organization's fortunes.

The right questions matter more than new data

I want to close the analysis with a thought I believe is central to every data problem in esports. We do not lack data. We lack the right questions to make old data speak. We do not need more data. We need better questions so old data can speak.

Think about this. Every esports match has been recorded. Every action has been stored. But most of that data has never been analyzed because nobody has asked the right question. We ask 'who won' when we should ask 'why they won'. We ask 'who has the highest metric' when we should ask 'what does this metric measure and what does it omit'. We ask 'who is the star' when we should ask 'which system allowed this person to shine'.

The shift from transaction questions to structural questions is the most important shift esports analytics needs to make. And it does not require more money, more technology, or more people. It requires a change in mindset: from wanting an immediate answer to accepting uncertainty until the data is strong enough to speak.

What does this mean for fans?

I am not writing this for executives or investors. I am writing for fans. Because in the end, it is fans who pay for the blurriness of data.

When a club misvalues a player because of bad data, fans watch an unbalanced roster. When a league inflates viewership to attract sponsorship, fans face format changes that do not serve them. When a star is inflated then criticized, fans lose a player they loved, and sometimes lose faith in the sport they follow.

But fans also have power. Every time you share a number without checking its source, you feed the loop. Every time you ask 'where does this number come from?', you make this industry better. It is a small act, but it has power.

Based on my experience watching matches and announcements, I believe esports fans are among the most analytically capable communities in the world. Communities like Dota 2 and Counter-Strike have built detailed databases so comprehensive that professional organizations consult them. That power should be directed at challenging public numbers, not only at defending legends.

Closing: value lies in what has not been measured

When I sat in that Seoul press room in the summer of 2026 and heard the question about the number's origin, I thought of all the times I had to abandon a conclusion for lack of data. I thought of the models I built and threw away. I thought of the times I was wrong, and the times I was right because I was patient.

Esports stands at a crossroads. It can keep operating on a blurred data foundation, keep building bubbles on numbers nobody verifies. Or it can choose a harder path: accept that most questions have no answers yet, invest in measurement infrastructure from the ground up, and build an analytical culture where saying 'I don't know' is not a sign of weakness but a sign of maturity.

I choose the second path, not because it is easy, but because it is right. And I believe this industry, with all its dynamism and youth, will choose it too — not immediately, but gradually, as operators understand that true value lies in what has not yet been measured, not in what has been painted over.

The question I leave readers with is not 'which number is correct'. It is: if esports can only move forward by measuring more honestly, who among us will be the first to refuse a beautiful number with no source?