Skip to content

Instantly share code, notes, and snippets.

@Seth-Johnson
Last active March 31, 2017 17:50
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save Seth-Johnson/3bc3f5a3e97acc4434025f108d7155bc to your computer and use it in GitHub Desktop.
Save Seth-Johnson/3bc3f5a3e97acc4434025f108d7155bc to your computer and use it in GitHub Desktop.
Give moderators better tooling to detect bad reviewers

Moderators have very little tooling to help deal with consistently bad reviewers. We have review audits & bans and some fancy stats pages. This is enough to stop true robo-reviewers, people that apply the same action to everything indiscriminately, but it breaks down when it comes to stopping just plain bad reviewers. I'm defining "bad" as reviewers who are otherwise paying attention (they pass all/most audits) but consistently apply wrong actions to things they review, actions they have to be corrected later. I've seen users that tend to "no action needed" most things, even if they are badly in need of an edit, comment, etc and other users who tend to "close & delete" everything, even if it's on-topic or otherwise acceptable. This is where it gets hairy.

I can look at a reviewer's stats and say "wow, 75% of this person's reviews are (no action needed/delete/close/etc) but moderators have no tooling to actually tell if this is pure coincidence or if bad actions are actually being taken, and no tools to find specific examples to approach the user with. This leaves me manually checking every single post a user reviews trying to figure out if this is a problem or not.

My proposal is this: Create a page (or amend the existing page) showing a user's reviews and the outcome of that post, whether it received bad flags, was closed, has a net negative score, was deleted, etc. This would be similar to the view moderators get on a user's profile where we get a list of a user's "bad" posts and a string of letters indicating which actions were taken on those posts.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment