If you’re not a paid editor yet or you’re looking to regain paid task access, we take a random sample of half of the first ten training tasks you’ve completed since registering. Any tasks you complete after the first ten are not necessary, but will be included in the pool of potential tasks we can select.
If you’ve been working on paid tasks, those we select for evaluation are a random sample of five tasks, taken from a pool of all the work you’ve completed since your last evaluation. The vast majority of tasks can be evaluated, but there are some exemptions for particularly difficult content. Please note we don’t change this on a case by case basis and therefore we’re not able to discount tasks from evaluation.