Why? Because bystanders don’t always have full context, we have a higher bar for bystander reports in order to find a violation. As a result, many reports of Tweets that in fact do violate our policies end up marked as non-violative on first review.