The International from $40M to $3.4M: The Migration of Esports Capital

Hook

The International 2026 paid out $40 million. The International 2026 paid out roughly $3.4 million. Esports World Cup 2026 paid out $75 million spread across dozens of titles. These three data points sit on three different axes, belong to three different funding models, and reflect three different stages of the same process. When I plotted them on a single chart, what appeared was not a single-line collapse. It was a migration of capital — from a pipeline pumped by the community to a pipeline pumped by states and multi-title organizations.

I spent nearly a week rebuilding this chart. I reviewed prize-pool data for every edition of The International from 2026 onward, cross-referenced it with multi-title events of comparable scale, and checked it against internal reports from the platform I work for. What I wanted to answer was simple: if the total volume of money in the global esports ecosystem has not declined, where did it flow?

Context

To answer that, I needed a method. In football analysis, there is a basic principle I learned in my earliest days working with data: never read a single metric in isolation. You cannot judge a team by goals scored — you look at xG, at PPDA, at distance covered, at the spatial structure of pressing phases. You cannot judge a club by roster value — you look at salary structure, cash flow, and contract trajectories.

In esports, I apply the same logic. Prize pool is a surface metric. It is like goals scored in football: easy to read, easy to impress with, but only the end result of a deeper chain of decisions. At a deeper level, I need to read three things: the funding structure (where money comes from), the cost structure (where money goes), and the authority structure (who has the power to change the rules of the money game).

The International is a perfect subject for this kind of analysis. It is the most transparent funding model in esports history, tied directly to in-game item sales through the Battle Pass. When Valve changed that mechanism, the impact of a single product decision on an entire ecosystem could be observed and measured. Something football never gives you — football has dozens of rights holders, hundreds of leagues, and no single entity can change the money rules with one patch.

The current cycle is a transfer window, and this is when information noise peaks. Rumors of transfers, withdrawals, investments, and bankruptcies appear at dense frequency. In an environment that noisy, method matters more than conclusion. What I need to do is separate signal from noise, and the signal in this case lies in structure, not in rumor.

A note on data limitations. Not every piece of information I present here has independent verification. Some figures come from official statements; others come from internal sources not yet cross-checked. For unverified data, I label it "pending verification" and do not use it as the basis for final conclusions. This is the discipline I carry from football analytics: a model is only as good as its inputs.

Core

In 2026, The International 10 reached a prize pool of $40 million — the peak of the crowdfunding model. The mechanism was clear: 25 percent of Battle Pass and in-game item sales went directly into the prize pool. Fans did not just watch; they spent, and that spending became prize money for players. There is no precedent for this model in traditional sports. In football, you cannot buy a digital item and turn it into prize money for players. In Dota 2, you can.

In 2026, the pool fell to $18.9 million. Still high by esports standards, but already down more than half from the peak. In 2026, it dropped to roughly $3.4 million — a 91 percent decline from 2026. In the most recent editions, the pool has continued to sit in the low millions. This sequence is not random. It is a trend line with a defined cause.

The cause lies in Valve's decision: a rework of the Battle Pass model that severed the link between item sales and the prize pool. Once that link was cut, The International's prize pool shifted from a community-determined metric to a publisher-determined reward.

This is where many readers misread the data. The drop from $40 million to a few million does not mean community interest in Dota 2 fell 91 percent. It is the arithmetic consequence of removing a funding mechanism, not a measure of a community's health. Data does not lie; only the reading is wrong. If you look at the prize pool and conclude Dota 2 is dying, you are taking temperature with a scale.

But the story does not stop at Dota 2. What matters more are the events happening simultaneously in other ecosystems.

Esports World Cup 2026 — a multi-title event backed by Saudi Arabia — was announced with a total prize pool of $75 million, spread across dozens of titles. Saudi eLeague 2026 brought together 37 clubs with a total prize value exceeding 4 million SAR. On a different timeline, the LCK — Korea's top League of Legends league — formally introduced a salary cap and luxury tax.

These three events are not random pieces. They form a three-tier picture. Tier one: capital is concentrating into a small number of large multi-title events, rather than spreading across single-title events. Tier two: decision-making power over capital is shifting from fan communities to financially capable organizations. Tier three: traditional leagues are having to self-correct with cost-control mechanisms to preserve competitiveness.

The Falcons case deserves close analysis. This is the organization that won The International 2026. They are not a weak team. They are the team at the very top of world Dota 2, and also an organization that entered as many as 18 tournaments within Esports World Cup 2026. Falcons formally announced its withdrawal from the Dota 2 ecosystem. The statement cited "ensuring long-term sustainable operations." Suspicious corporate language, but the logic is clear.

An organization at the peak of competitive achievement willingly narrowing its investment portfolio. They are not underperforming, not bankrupt. They chose to leave. In football analysis, I once read Josef Martinez's xG and saw a revolution brewing at Atlanta. In the Falcons case, I read the opposite signal: a calculated withdrawal of capital from an ecosystem where prizes no longer cover operating costs.

At the same time, Dplus KIA — the organization that just won the League of Legends title at Esports World Cup 2026 — faces serious liquidity problems. Internal sources say the team has been delaying player salary payments and is in the process of seeking a new owner. Their League of Legends roster costs an estimated 3 billion KRW, roughly $2 million in player salaries alone. Their predecessor organization, DAMWON Gaming, won Worlds 2026.

Reading these two events side by side reveals a paradox. I call it the competitive-efficiency paradox. An organization can win a world-class title and still face the risk of bankruptcy. In traditional sports, winning usually comes with rising revenue — broadcast rights rise, sponsorship deals rise, brand value rises. In esports today, winning and financial sustainability are drifting apart.

To understand why, I need to dissect the cost structure. Esports organization costs come from three main sources: player salaries, operational overhead, and tournament participation costs. Revenue comes from four sources: sponsorship, publisher royalties, tournament prizes, and commercial revenue. The problem lies in the growth rates of the two sides. During the growth phase, player prices rose faster than revenue generation. Clubs signed contracts based on growth expectations, not actual cash flow.

When growth slows, the cost structure is already locked at a high level. A $2 million salary roster still needs to be paid, while sponsorship revenue does not rise correspondingly. The LCK salary cap emerged from this pressure. It is not a punitive measure, but a necessary one to prevent an uncontrolled spending race.

The LCK luxury tax is notable for another reason. It is not merely a spending ceiling but a redistribution mechanism. Organizations spending above a threshold pay extra, and that money goes into the league's common fund. In football, I have seen similar mechanisms in the English Premier League and North American leagues — where revenue is shared to protect overall competitiveness. The LCK is moving in that direction, but in the esports environment this is a step with few precedents.

Reading the three cases — The International, Falcons, Dplus KIA — side by side, a three-tier model emerges. The first tier is the ecosystem tier: the funding model changes, and the prize pool contracts. The second tier is the organization tier: costs exceed revenue, forcing organizations to restructure or withdraw. The third tier is the governance tier: leagues and publishers must intervene with cost-control mechanisms.

Each tier reacts at a different speed. The ecosystem tier reacts fastest — Valve only needs one product change. The organization tier reacts more slowly — Falcons took months to decide to withdraw. The governance tier reacts slowest — the LCK needs time to discuss, agree, and enact. The delay between tiers is precisely the window in which risk accumulates.

To clarify this three-tier model, I need an example illustrating how a shock propagates. Imagine Valve's Battle Pass rework decision. It is a single decision at the ecosystem tier. Within weeks, The International's prize pool changes. Within months, Dota 2 organizations begin recalculating operating costs. Within one to two years, decisions about rosters, investment, and withdrawal begin to appear. By the end of the cycle, you see moves like the Falcons withdrawal. A decision at the top tier takes nearly two years to fully manifest at the bottom tier.

Based on my first-hand experience watching matches — from packed The International arenas to sparsely attended group stages — I noticed something about the relationship between applause and spending. In 2026, when European football returned to empty stadiums, I measured average PPDA falling from 10.8 to 9.7. When a stadium falls silent, the only thing left is the honesty of pressing. Players pressed more truthfully, because there was no cheering to hide hesitation behind.

In esports, this relationship has a different variant. Dota 2 viewers do not need to be in an arena to spend. They only need to buy a Battle Pass. So the breaking of the crowdfunding model did not come from viewers losing interest, but from the publisher changing the collection mechanism. This is a fundamental difference between the esports model and the traditional sports model.

I want to go deeper into the $75 million figure for Esports World Cup 2026. It is a large number, but reading it requires care. It spans dozens of titles. If divided equally, each title would receive a few million dollars — comparable to or below the prize pool of an average The International. But the actual distribution is not equal. Titles with larger viewership and stronger participating organizations receive a larger share.

The key point lies in the origin of the money. The International 2026 drew money from the player community. Esports World Cup 2026 draws money from the state and strategic investment funds. These two sources have completely different characteristics. Community money is dispersed, relatively stable by season, and tied to product popularity. State money is concentrated, dependent on policy decisions, and can shift quickly if strategic priorities change.

No source is absolutely better. But the shift from a dispersed source to a concentrated one changes the risk structure of the entire ecosystem. When money concentrates, the system's ability to absorb losses falls. A single decision by one sponsor can affect dozens of organizations. The International 2026 showed source diversity — millions of players contributing. Esports World Cup 2026 shows concentration — a handful of strategic sponsors.

The degree of capital concentration matters more than the total volume of capital. An ecosystem with $40 million from 5 million contributors has a different risk structure than one with $75 million from 5 sponsors. A larger sum does not mean a healthier ecosystem.

The Saudi eLeague 2026 case deserves to be placed alongside Esports World Cup. With 37 clubs, it is one of the largest domestic leagues by participating organizations in the region. That number needs careful analysis, because it reflects different dynamics between expanding supply and expanding demand. When the number of clubs rises, it means more capital is being injected into the system. But if viewer numbers, player numbers, and spending numbers do not rise correspondingly, you are watching supply growth. And supply growth can be financed with money, while demand growth cannot.

Contrarian

Now to the part where I want to challenge the popular reading.

The popular reading of esports' current situation is the "esports winter" — a comprehensive downturn caused by over-fast growth during the pandemic and a collapse afterward. I do not agree with that simple reading. But I also do not agree with the opposite reading — that everything is fine, it is just reallocation.

Both readings make the same mistake: treating the esports ecosystem as a uniform entity. It is not. It is a set of ecosystems with different dynamics, affected by different capital flows, and with different resilience. When you aggregate everything into one total metric, you lose the most important information — internal differentiation.

The data I presented shows a more precise pattern: polarization. A small group of multi-title organizations with financial strength tied to state-backed events is expanding. A large group of single-title organizations dependent on prize pools and traditional sponsorship is contracting. The gap between the two groups widens.

But there is one point I want to question about the polarization model itself. Correlation between multi-title scale and survival does not mean causation. A multi-title organization is not automatically more sustainable than a single-title one. Falcons is an example. They are multi-title, entered 18 events, and still willingly withdrew from a core title. Dplus KIA won a major title and still needs a new owner.

If you read my data and conclude that diversifying titles is the way to survive, you are reading too quickly. The true variable is not the number of titles, but the cost structure relative to the revenue mix. A multi-title organization with high salary costs in every title will face exactly the same problem as a single-title organization with high salary costs. Diversification only helps when it comes with cost discipline and a diversified revenue mix.

The Arda Güler case I once analyzed is a lesson about the relationship between timing and value. In early 2026, I analyzed the data of this 16-year-old midfielder at Fenerbahçe — 3.4 successful dribbles per 90 minutes, creativity index in the top 5 percent. I delayed my report by 10 days to wait for additional data from three other leagues. By the time I sent a 5 million euro valuation proposal, the transfer window had closed. In summer 2026, Güler moved to Real Madrid for 20 million euros.

That lesson applies directly to the current moment. When an ecosystem is restructuring, asset values shift faster than organizations can react. A roster once valued at 3 billion KRW can become a burden after a single season. Conversely, an undervalued organization can become an opportunity if its cost structure is restructured correctly.

I want to push the skepticism one step further. The $75 million figure for Esports World Cup 2026 is often cited as evidence that esports remains healthy, that capital still flows strongly, that the winter is only a local phenomenon. But that data needs to be read alongside another data point: 37 clubs in Saudi eLeague 2026, a number unusually high for a domestic league. If the number of clubs rises faster than the number of viewers, you are watching supply growth, not demand growth.

This is an important distinction. Esports supply can be expanded with money. Esports demand — viewers, players, spenders — cannot be expanded with money at the same speed. As the gap between the two widens, the business model becomes dependent on external capital rather than endogenous revenue. And external capital, like any investment flow, can be withdrawn.

The transfer market is where emotions get priced, and I only stand outside that room. In the current transfer window, I see many rumors about large organizations looking to acquire financially struggling rosters. The logic of these deals is simple: buy assets at a discount, restructure costs, and wait for the cycle to recover. But that logic assumes the cycle will recover. If the ecosystem's structure has changed permanently, then acquiring an expensive roster is just buying debt at a discount.

One final point I want to raise. The story of Falcons' withdrawal and the Dplus KIA crisis is usually told as two separate events. But placed side by side, they form a shared pattern: organizations are re-evaluating the value of each title based on actual profitability. In the growth phase, the criterion was presence in as many titles as possible. In the current phase, the criterion is efficiency per unit of capital invested. That change in criteria, not any single event, is the biggest story of this cycle.

Takeaway

So which signals should we track in the next cycle?

I will track three metrics. The first is the shift of money between titles within multi-title events — whether Dota 2 continues to lose its position in allocation, or whether it finds a new funding model. The second is the number of organizations entering single-title events — if this number keeps falling, the traditional league model has a structural problem, not a cyclical one. The third is the speed at which the LCK salary-cap mechanism spreads to other leagues — if it spreads quickly, that signals league administrators see it as a necessary fix; if it spreads slowly, that signals the salary race continues.

What I have not yet satisfactorily answered: can an esports ecosystem exist sustainably without a massive prize pool as its driver? Or is a large prize pool only the early stage of a maturing model, and true stability lies in other revenue sources like sponsorship, rights, and commerce?

Data is where I take shelter, but also where I learn to distrust every claim. I once believed that prize pools were the measure of an esports ecosystem's health. After rebuilding the entire chart from 2026 to now and placing it alongside data on salaries, ownership structure, and capital origin, I am no longer sure of that. Perhaps we are witnessing a transformation that will only be fully legible a few years from now. And perhaps what looks like decline is only the shape of something else being born.