This question and its comments and answers inspired me to write this post.
There are lots of little rules and restrictions in place that can catch users by surprise and prove to be very annoying, such as:
- the resetting nature of the comment rate-limit timer,
- the six-character minimum and edit summary requirement for suggesting edits,
- the inaccessibility of deleted questions (more than 60 days old) to users, and
- the title word filter.
These restrictions are apparently intended to curb abuse, improve content quality, suppress "whining", and reduce the need for human moderation. However, do these restrictions genuinely reduce community moderation and reviewing burden? Do they actually improve the user experience for everyone?
Why don't we actually test these changes Google-style on select sites on the Stack Exchange network? Taking the title word filter as an example, it might be a good idea to run a test where the filter is disabled for a few days on Stack Overflow and collect some statistics on the quality of posts with otherwise filtered words (and "leet" equivalents like "pr0blem"). The data collected can be presented to the community afterwards, and the community can voice its opinion based on real data so that the right decision can be made.
It's my hope that this will allow for some serious improvements to the Stack Exchange experience without hurting content quality. Your thoughts?