Build an A/B test that will test a feature which detects the user's language and re-sorts the links around the globe on the Wikipedia portal in line with their preferred language settings. If they do not have as many language preferences as links, fill the remainder with the "top" links that are not in their language preferences. If they do not have a header full stop (or we cannot retrieve it), give them the default experience.
Sample logic should look like the following:
1. 1 in 200 people are included in EventLogging
2. Of those 1 in 200 people, 1 in 10 are included in the test
3. Of those 10 people, 5 go in a test group, with the cohort "language-detection-b", and 5 go in a control group, with the cohort name "language-detection-a"
4. The other chunk of the 200 people gets a NULL (the string null, or the MySQL null, we can detect either).