How is Game of the Year decided?

The short answer

There is no single Game of the Year, so there is no single decision. Every outlet that gives one decides its own: a magazine's editors, a website's staff, a readers' poll, a jury, an academy of developers. In the 2025 season 680 of them did, and 3,360 have since 1990.

The award most people mean is the one from The Game Awards. Its winner comes from a jury of more than 100 media outlets, blended with a public fan vote that counts for 10% — as the show itself explains it.

This site adds no vote of its own. It counts: the game named Game of the Year by the most outlets wins the season, and every award counts once. Why no outlet is weighted is a deliberate choice with a cost, explained there.

Five ways a Game of the Year gets decided

1. An editorial team. Most of the field works this way: a magazine or website argues it out internally and publishes one pick.
2. A readers' poll. The audience votes. 24.2% of every award in this database came from one.
3. A jury of outlets. The Game Awards is the press sitting as one panel.
4. A peer vote by developers. The D.I.C.E. Awards and the Game Developers Choice Awards are voted by people who make games.
5. An academy jury. BAFTA members vote through juried category panels.

For the big ceremonies, the method and the calendar look like this:

Does the method change the winner?

It does, measurably. The measure is simple: in how many census seasons did a ceremony's Game of the Year turn out to be the game the whole field rewarded most?

The Game Awards matched the field in 83.3% of the 12 seasons it has judged. The Golden Joysticks, decided by public vote, did in 30% of 20. BAFTA's academy sits at 36.4%.

Part of that is construction. The field this site counts is mostly press, and a jury made of press outlets is a small sample of the same population — it would be strange if it disagreed. A public vote or a developers' academy is a different electorate, and its pick carries information the count does not.

Agreement is not authority. A ceremony that always lands on the consensus tells you nothing you could not get by counting. The ones that diverge are the ones worth reading.

Only the jury or editorial award is counted here. A ceremony's separate fan award, such as The Game Awards' Players' Voice, is a different prize and is left out of this comparison. The full calendar and every winner are on the awards page.

Readers and critics do not always agree

Readers' polls and editorial picks usually crown the same game. When they do not, the season is worth a look. Every such season is listed in the seasons where players overruled the press, and the long-run pattern is in readers versus critics.

Does a late release have an advantage?

The usual theory is that voters remember December better than March, so autumn releases win. The winners are the test. Of the 22 census-season winners whose release date is on record, 14 came out between January and June and 8 between July and December.

If voters really forgot spring releases by the end of the year, the second number would be the larger one. Winners without a release date inside their season — early-access games, or dates missing from the record — are left out rather than guessed.

How this site settles a season

• Every award counts once, from any outlet in any country. The season goes to the game with the most.
• Totals are documented minimums. Every row has a source; an award nobody wrote down is not in the count.
• Only census seasons are compared. Before 2003 a season holds a handful of surviving awards, so it is a record, not a result.
• A tie has no winner. Two games level at the top are shown as level.
• A platform award is not a Game of the Year when the same outlet also gives an overall one.

The result for every season is on the list of every Game of the Year winner, and the full rules are in the methodology.