we pull data from public archives and run our own algorithm then return the top talked scores to users.