The 0% Team at TI 2026 Isn't the Story. The Price Grid Is.

The 0% Team at TI 2026 Isn't the Story. The Price Grid Is.

The most dishonest number on the board at The International 2026 is also the most honest one: 0%.

The fifteenth edition of Dota 2's world championship — sixteen teams, the Oriental Sports Center in Shanghai, the Aegis finals playing out this weekend — has at least one team priced with essentially no chance of taking the whole thing. Render it however a given board chose to render it: 0.0% on a chart, a 0.1-cent ticket on a prediction market, a flat 1% floor on a sportsbook grid. Good value for a headline, and almost certainly wrong. The useful question is what kind of wrong.

On the futures board before the tournament, the favorite was Team Vision, the team formerly known as PARIVISION and qualified through the European qualifiers, quoted at 2.80 — about a 36% implied chance if you naively invert the odds. Team Yandex, a direct invite, sat at 3.20. Then the field thinned out fast:

Team	Pre-event odds	Naive implied % (1/odds)
Team Vision (PARIVISION)	2.80	36
Team Yandex	3.20	31
BoomBoys	6.30	16
Aurora Gaming	8.00	13
Team Falcons	8.00	13
Team Spirit	11.00	9
Iron Wing	13.00	8
Xtreme Gaming	20.00	5
Nigma Galaxy	35.00	3
Vici Gaming	45.00	2
LGD Gaming	55.00	2
HULIGANI	100.00	1
GamerLegion	100.00	1
Team Resilience	100.00	1

Decimal odds carry the bookmaker's margin, so the real probabilities are lower at the top than the naive inversion. But the shape is what matters: the board simply refuses to go below 1%. The number the coverage reaches for — "0%" — never appears on that grid. It appears on the prediction books, where the same long tail is quoted in fractions of a cent. Once the main event was underway, the champion market had Team Vision at 66%, Team Spirit at 19%, Team Yandex at 12%, BoomBoys at about 4%, and Nigma Galaxy — the team that beat the two-time champions 2-0 in the group stage to walk straight into the playoffs — under 1%, a 0.1-cent share.

That price is not an estimate of Nigma's chances. It is the shape of the quoting machine. A sportsbook's odds grid and a prediction market's order book alike cannot represent "extremely unlikely but real." Everything past the deep tail gets clamped to the cheapest tick the machine can print: 1%, a tenth of a cent, zero. And there is a hard economic reason the machine stops there, which is that a number that cheap is not worth disputing. To argue with a 0.1-cent quote you would have to hold the opposite side for weeks — tying up a dollar for a scarce tenth of a cent — on an outcome that can flip on a single draft or a single pit fight. Nobody prices that. The market rounds it, and the round is the point.

The tell that you are looking at rounding rather than information is the one-sidedness. The market for "this longshot does not win" does not exist on the champion board; the No side of Nigma wasn't even quoted. What that says is not that the crowd believes the team cannot win. It says no one is willing to be the counterparty on a bet whose payoff is a fraction of a cent and whose settlement is weeks away. Information flows where counterparties are willing to stand. At the tail of a season-long futures board, neither side is willing to stand, so the market prints nothing.

Where the Money Actually Goes

Now watch where the same crowd puts its money when it wants to express an actual view. The platform that quoted Nigma at a tenth of a cent on the champion future moved $9.5 million across the markets on a single best-of-three between Team Yandex and Nigma Galaxy — $5.7 million of it on the moneyline alone. An earlier Nigma elimination series carried $2.05 million. Set that against the roughly $2.1 million the entire championship future had accumulated in volume since late June. One playoff series, one pair of teams, eclipsed what the season-long "who wins it all" market managed in seven weeks.

That is the basis-risk pattern stated in real numbers. The hedging that the betting public actually wants is specific and near-term: who wins this series tonight, not who survives a sixteen-team gauntlet three weeks from now. The instruments that let people transact the specific hedge are liquid precisely because they are specific; the long-horizon winner future is a speculative accessory with a month of dead time, coarse resolution, and no counterparty appetite at the tail. The market is not mispricing the bottom of the board out of ignorance. It is not pricing it at all, because the structure rewards every dollar that moves toward the near-term, specific bet and punishes every dollar parked in the long capsule. The deep product is a rounding error waiting to be reported as a forecast.

The tournament itself keeps supplying the empirical check on what the machine cannot express. The odds-on second favorite, Team Yandex at 3.20, went 2-3 through the group stage — a couple of bad series from going home. HULIGANI, ranked 24th in the world and priced at the 1% floor, went out first. Iron Wing, quoted at 13.00, played its way into the final eight. None of this proves any particular longshot will lift the Aegis. It proves the tail cannot be priced at the resolution the machine offers, because the variance of a sixteen-team double-elimination is exponentially bigger than the grid's deepest tick. The modelers know it even where the boards won't admit it: one widely shared 10-million-iteration Monte Carlo of the event plots all sixteen teams on a 0-to-30% axis — the entire field wedged into less than a third of the chart, with the bottom of the field rendered as a straight line at zero.

Read the Machine, Not the Team

The same structural indifference sits inside the champion market's own rulebook. It resolves against a third-party stats database rather than an official settlement source. In the unlikely event of a first-place tie it breaks the tie alphabetically, and anything not decided by early September falls into an "Other" bucket. That is acceptable plumbing for a market nobody trades at the tail. It would never survive contact with the money that flows through the single-series markets — which is the point. On the short horizon the market bothers to be precise, because the incentives to be precise are real. On the long horizon it settles for coarse governance, because no one is there to demand better. The resolution mechanism is exactly as careful as the capital structure beneath it.

Verdict: A 0% printed on a board is never information about the world. It is information about the machine — its grid, its cheapest tick, and the people it could and could not attract to the other side of a trade. The teams at the bottom of the TI board know the number is a lie, because none of them read it as permission. Every longshot ever quoted at zero has the same job: win enough series that the machine is forced to reprice.

When someone shows you a market quoting 0%, or 100%, the disciplined move is not to argue with the number. It is to ask who could act on it. If the answer is no one, you are not looking at a market. You are looking at a rounding error with a price tag — true at TI in Shanghai this weekend, and just as true in every futures board, crypto or otherwise, that prints probabilities finer than its structure can carry.