Methodology

What the win probabilities are

Each country's figure is an editorial estimate of its chance of winning Eurovision Asia 2026. It is a judgement, written down and argued for — not the output of a betting market, a poll or a statistical model. There is no voting history to model: this is a first edition and all eleven countries are debuting at once.

For every country we weigh three things that can actually be observed:

• Market and production strength — the size of the country's music industry and audience, and the resources behind the entry.
• How much of the entry is confirmed — an artist with a released song can be judged; an unannounced act can only be guessed at, so it is marked down for uncertainty rather than for quality.
• Promotional runway — how long the entry has to build an audience before 14 November. Bhutan's song has been public since 11 August; Cambodia's will exist for four weeks.

The reasoning under each country on the odds page says which of those is doing the work, so you can disagree with a specific step rather than with a bare number.

Why the numbers add up to 100

The eleven figures always sum to 100%. That has a consequence worth knowing: a country can drift down without anything about it getting worse. When Thailand finally had an entry to judge and moved up, the points it gained had to come from somewhere.

It is also why no country is rated much above one chance in five. With no form guide, a result split evenly between juries and the public, and several entries still unheard, a confident favourite would be an invented one.

How fractional odds are shown

Next to each percentage is the equivalent fractional price, rounded to the nearest conventional fraction — 18% is shown as 9/2, 4% as 24/1. These are "fair" odds with no bookmaker's margin built in. They are a second way of reading the same estimate, not prices anyone is offering.

Movement and revision history

The arrow beside each country shows the change in percentage points since the previous revision. We revise after every national final and every confirmed entry, not on a fixed schedule. There have been 4 published revisions so far — 26 Aug, 23 Sep, 7 Oct, 14 Oct — and every figure from each one is kept in the history table on the odds page. Nothing there is back-filled.

Why we don't use betting-market prices yet

Mainstream bookmakers have not priced Eurovision Asia. One prediction market has opened: Kalshi listed a "Eurovision Song Contest Asia Winner" market on 14 September 2026. We track it, and show a dated snapshot of it on the odds page, but we do not blend it into our figures.

The reason is in the numbers. In a functioning market the prices for all outcomes add up to roughly 100. When we took our snapshot on 14 October 2026, 09:21 UTC, the last-traded prices added up to 269, only two countries had anyone offering to buy, and there had been no trades at all in the previous day. Those are a few isolated transactions, not a crowd's estimate.

We will start blending market prices in once there is two-sided trading — real bids and offers on most countries, and prices that sum to something close to 100. Our explainer on betting and Eurovision Asia has the detail.

How we source facts

Dates, artists, songs, voting splits and results come from the contest's official channels and the participating broadcasters first, then from specialist Eurovision outlets — chiefly Eurovoix, Aussievision, ESCToday and ESCXTRA — and from the relevant Wikipedia articles, which we use as a cross-check rather than a primary source.

Three rules follow from the fact that this is a first edition where information arrives in bursts from eleven broadcasters:

• We don't fill gaps with patterns. If a broadcaster hasn't published a voting split or a jury, the page says "pending", even when every other country has done the same thing.
• We publish only what was published. Most national finals announced a winner or a top three without points. Our result tables show exactly that and no more.
• When sources disagree, we say so. Where outlets report different things — the order of a top three, a start time, the name of a broadcaster — we name the disagreement instead of quietly picking a side.

Corrections

We get things wrong, and when we do we correct the page in place and say that we have. Our running fact-check logs the notable ones, including our own. If you spot a mistake, tell us — a correction from a reader is worth more than a wrong page.

What this isn't

These are independent, fan-made estimates — not official data from the EBU, Voxovation or any broadcaster, not betting or financial advice, and not a guarantee of anything. Treat them the way you'd treat any other informed opinion.