I’m a little confused about /admin/reports/user_flagging_ratio, either I’m doing something wrong or I think it’s bugged. I’ve tested it with someone who flagged 5 posts, agreed and took no action, this didn’t update so I thought maybe “agreed” needs to be “agreed and took action”, ran the test again silencing, suspending and hiding posts and no update.
It appears to be yes, this section believes our site has had about 25 flags since the new year, anything else that refers to flags on the dash appears ok.
The game plan as we grow is to pay the top 10 “most helpful” flaggers in the community to incentivise and outsource moderating by tweaking flags… That’s how I ended up on that report which I suppose is rarely looked at otherwise.
Le problème que nous avions était que, pour éviter certaines erreurs, nous avions cette clause HAVING #{disagreed} > #{agreed}, ce qui empêchait essentiellement l’affichage de nombreux résultats.
Si vous avez 5 désaccords et 20 accords, vous n’apparaîtriez pas dans les résultats. Cette correction permet également d’inclure les personnes ayant 0 drapeaux de désaccord ; leur score sera égal à agreed^2.