From $40 Million to a Few Million: When Esports Money Moves Instead of Disappearing

The night of The International 2026 grand final, the arena in Bucharest stood empty because of the pandemic. But on screen, the prize pool jumped to 40 million US dollars — the highest peak ever recorded for a single esports event. I sat at my computer in Brisbane, logging every time that number moved. To me it was never just money. It was a metric measuring how tightly a community gripped its own game.

Three years later, at the same tournament, the prize pool fell to roughly 3.4 million dollars. More recently, only "a few million." A collapse of about 91 percent from peak.

But what made me sit down and write was not the number. It was Falcons — the team that had just won The International 2026 — announcing its withdrawal from Dota 2. And Dplus KIA — the team that had just won the League of Legends title at Esports World Cup 2026 — delaying salaries and searching for a new owner.

Two champions. Two disciplines. The same signal: winning is no longer enough to keep them in the game.

To understand what is happening, you have to separate two kinds of change that people habitually merge: gameplay balance shifts and business-model shifts.

The dataset I analyzed references no gameplay patch at all. No patch number, no hero changes, nothing touching the map. Which means "meta" here is not the in-game meta. It is an economic meta.

The only systemic change lies in Valve's Battle Pass rework. Previously, in-game item sales flowed directly into the TI prize pool. The community bought cosmetics, the pool swelled, and teams could see the number and know what they were investing into. When Valve cut that wire, the prize pool instantly became a figure decided by the publisher, no longer nurtured by the community.

This is the anchor point I want to set before going further: the TI prize pool did not collapse because Dota 2 players turned away. It collapsed because the fundraising mechanism was dismantled. Conflating those two things with "esports is dying" is precisely the reasoning error the source analysis itself warned against.

Meanwhile, money elsewhere is rising. Esports World Cup 2026 carries a 75 million dollar total prize pool spread across dozens of titles. Saudi eLeague 2026 gathers 37 clubs with more than 4 million riyals. The money has not vanished. It has moved.

And here is the pattern I have tracked across many seasons: the geography of money is forming two poles. Korea is self-correcting — maturing, tightening. Saudi Arabia is expanding — injecting capital, buying presence. The rest of the world, including China, Europe and North America, is almost absent from this dataset. For a topic called "global esports," that is a significant blind spot.

Start with Dplus KIA, because it is the cleanest case.

The organization won the League of Legends title at EWC 2026. Its predecessor, DAMWON Gaming, won the 2026 World Championship. On paper, it is one of the strongest names in Korea. But its LoL roster consumes roughly 3 billion won — nearly 2 million dollars — in salary alone. And it delayed payments and had to seek a new owner.

A team that won one of the biggest tournaments of the year still could not cover its payroll. That is the strongest evidence yet that competitive achievement and financial survival have decoupled.

I have tracked this kind of data for many seasons. In the 2026 A-League, I once wrote about Jamie Maclaren — 8 goals but an xG of 14.2, a striker squandering far too many clear chances. That piece had almost all of its numbers struck out by an editor. But it taught me something: numbers do not speak on their own. People only trust a number when they can see the human being behind it.

The human being behind Dplus KIA is an expensive, title-winning roster whose balance sheet could not keep pace. Its salary cost exceeded the commercial ceiling of the very discipline it competed in. Winning a trophy does not equal earning money.

Then Falcons.

Falcons won TI 2026. In 2026, they entered 18 tournaments at EWC. It is a multi-title organization, backed by heavy capital, at the peak of its form. And yet it withdrew from Dota 2.

Read their statement carefully — the only named source in the entire dataset I analyzed — and the phrase that stands out is "long-term sustainable operations." This is not an organization that lost and quit. This is an organization that won and recalculated.

When the TI prize pool falls to a few million while EWC pays 75 million across dozens of titles, maintaining an elite Dota 2 roster becomes a negative-ROI problem. Falcons did not withdraw because they were weak. They withdrew because they were calculating.

And at league level, the LCK introduced a salary cap with a luxury tax. This is a governance intervention, not a market outcome. Its meaning is clear: the league recognized that player prices had risen faster than revenue generation. A salary cap is not a punishment. It is a release valve.

The LCK salary cap is not merely a cost limit — it is a redistribution mechanism. The heaviest-spending teams pay a luxury tax to subsidize the rest of the league. It is a competitive-balance tool, and it arrived exactly on time.

These three data points — Dplus KIA, Falcons, the LCK — assemble into one picture: the money still exists, but it no longer flows evenly through the whole system. It flows toward major tournaments, titles with commercial value, and organizations with sustainable operations. Somewhere, money is still pouring in. Just not where it used to.

There is one more layer few notice: governance risk. When Valve unilaterally changed the Battle Pass, it was simultaneously rule-maker and commercial beneficiary. The entire Dota 2 prize economy depends on a product decision with no counterbalancing mechanism. A single product decision can demolish a funding channel worth tens of millions — and there is no shield between publishers. That is systemic risk no tournament can self-insure against.

Here I have to be careful about how the data is read.

The easiest thing to do is look at the 91 percent collapse in the TI prize pool and conclude Dota 2 is dying. I consider that a correlation fallacy. The pool collapsed because the fundraising mechanism was dismantled, not because player numbers fell. Those are not the same source.

The second dangerous move is looking at EWC's 75 million and concluding esports is ballooning. Also wrong. Concentrating money into a few super-events creates a new risk: mid-tier organizations grow increasingly dependent on appearance fees rather than performance-based prize money. That is a more dangerous dependency, because it does not encourage competition — it encourages attendance.

Concentrating money into super-events looks like growth, but it is actually a reduction in diversity. An ecosystem with fewer load-bearing points is more fragile when one of them shakes.

And here is the twist I believe the data is screaming while few care to listen: risk is not uniform. Dplus KIA and the Dota 2 ecosystem absorb losses. Entities tied to Gulf money expand. Same industry, same moment, two opposite directions. Calling it an "esports winter" without specifying who is cold and who is not skips the most important part of the story.

Here I want to recall something I learned in the summer of 2026, when COVID froze every tournament and I lost two broadcast contracts. I reopened the Liverpool 4-0 Barcelona match and built my own dataset on Andrew Robertson's movement — 12.4 kilometers, of which 2.1 kilometers were sprints. Writing about numbers nobody asked for. By morning, that piece had been shared more than 4,000 times. I learned that data only matters when it touches longing, when it answers "what does it feel like," not just "how many" and "where."

Every number has a story; my job is not to ruin it.

The signal I am watching in the next cycle is not prize pools. It is contracts. Will the Korean salary cap spread to other leagues, or will Korea lose its stars to uncapped leagues? And will the next champion be forced to choose between a trophy and an invoice?

When the spreadsheet speaks, the stadium must learn to stay silent. But this time, the spreadsheet is talking about rent, not form. At 39, I have learned that data also feels pain when it is distorted.