How does the contest work?

The contest is divided into three different phases: submission phase, qualification phase and final voting phase.

During the submission phase, contestants propose their translations of any of the source texts provided into any language. There is a submission deadline for all language pairs and if a pair receives fewer than 3 entries, it does not go into final voting rounds and so no winner is selected. However, feedback is enabled for the benefit of the contestants and translators that work in the pair can tag the sections of the entries they like or dislike and provide a comment.

Note that the submission deadline may be extended for some language pairs at the discretion of the contest organizer.

For pairs with a large number of entries (more than 7), a qualification round is opened to ensure that an appropriate number of entries reaches the finals, and that the entries selected are the best and most accurate. Voters are asked to rate entries in two categories: "Quality of writing" and "Accuracy of translation" on an absolute scale of 1 ("poor") to 5 ("perfect"). During this phase, entries are removed from contention when it becomes clear that they have no reasonable chance of winning. Such assessment is made based on the average rating received from peers (relative to that assigned to other entries), after some minimum number of votes.

Winnowing out entries in this way helps focus voter attention on the leading entries, increasing the likelihood that the best entries will be identified for inclusion in the final round. The 3-7 entries that receive the highest scores will pass into the final round.

In final voting rounds, voters are asked to select the entry that they consider the best of the group. Second and third best entries may also be designated, with points being assigned to entries in amounts of four (4), two (2) and one (1), respectively. The entry that collects the most points in final round voting, wins.