Bumping posts

Question

Today, polarkernel added a new random bumping feature to remind readers of potentially interesting content of PO.The current bumping rules are as follows:

* A new bump is created at random time between 6 and 18 hours since the previous bump.

* Bumped posts are older than 90 days.

* They are selected among the following groups:
 - 30% unanswered questions (invitation to write an answer).
 - 30% posts with total Q&A votes >8 (reminder of goodies we have).
 - 40% all other posts from Q&A (show representative selection).

* To give more weight to rarer topics, the relative weight w for random selection is
 - w = max(0,netvote of question)+(sum(max(0,netvote) of answers)/(frequency of the most frequent tag)
 - where the (obsolete but still existing) tag 'research-level' is excluded from this computation.

* Questions once bumped are excluded from future selection

* Do bumped questions get activity (what we all hope) they appear as normal activity in the list and a new bump will be created within maximum one hour. 
 

Comments

This is pretty good, but I'd increase the representation of unanswered questions to 50% to improve our currently poor unanswered rates.

I also think that imported questions (questions with votes from SchrodingersCatVoter) shouldn't be displayed unless they're unanswered, as they aren't a reminder of our goodies or a representative selection of our posts.

---

@Dimension10

I disagree with excluding imported questions from the bumping mechanism. To me it seems not that important where the votes come from. To not exclusively represent the current PO community is a feature of the bumping mechanism not a bug, as it should also help enlarge the community.

At present, I would not increase the weight of the unanswered questions either. It is still the case that most of our current (imported or native) questions appearing on the main page stay unanswered. We should only adapt the bumping mechanism to bring up more unanswered questions when the answering capability of our community has significantly improved ...

---

@Dimension10: To improve our currently poor unanswered rates needs more committed users, not more displays. I think the likelihood that some new competent people will answer some questions is higher if they see more of the stuff that is here rather than if they see even more unanswered questions.

Most of the imported questions are from the now defunct SE-TheoreticalPhysics of which we are the continuation; so these are ''our'' posts. And for the more recent ones, if they are unanswered we need answers, and if they are answered, an answer is often form PO rather than from SE