Forum:Almost Impossible For A New Question To Get To The Top

Question

[Michael]

I have just made a new test question to see where new questions arrive at the moment.

It actually arrived on position 5, while adding 2 comments + receiving 2 views didn't bring on position 1 which is occupied by a bad question with 5 downvotes, but upvotes on comments and answers plue 34 views.

I think this is due to the ranking by mixture of votes, edits and views. I have the impression that this system is very unfair towards promoting only old questions. In fact I want the old stack behaviour back that while being not really documented seems much more transparent to me: new questions and edits put the question on top of the stack again, anything else has no effect. I think adding comments an maybe even voting to this is ok, but giving viewing any significant influence is odd imo.

That in consequence leads to top questions staying at the top forever because of curiosity, and it might also happen that new questions arrive very lowly ranked in the list, just because there are a lot of views accumulated to the top posts.

I think this needs an urgent tweak.

I think new questions should be always at the top, to give them a good chance to be seen.

[Simon Cockell]

cannot agree strongly enough with this.... The ordering of questions has to be transparent, obvious and (ideally) customisable.

[Istvan Albert]

but it did go to the top soon in a few hours. which is the right behavior in my opinion, once sufficient people looked at it, and some number of votes were cast on the answers and comments - I think it worked perhaps not as optimally as it should. A question with some level of activity will shoot to the top.

Answer 1

[Istvan Albert]

I can tweak it so that new questions go immediately to the top - also the formulas for gaining reputation on views/votes/answers/comments probably need to be toned down a bit. Getting these right is not as straightforward and some experimentation is necessary. Both of these changes will go in soon, in a few days.

More importantly I want to add (this almost done) sorting to each page - that will also allow people to simply order by creation date, votes, rank, etc and will remember your choice between visits. I will also add the rank to be visible for moderators, that way we can better understand where the disparities come from.

Now getting a question on top just because someone edited it always irked me as a wrong behavior. I always felt that a badly written questions now gets to the top again just just because someone took the effort to fix a typo or a link.

[Neilfws]

My opinion: moving questions to the top should occur when there is any activity on them, including editing. People just want to see easily what has changed; there's no judgement on the value of that change. Currently, all I see is old questions, giving the impression that nothing is new.

[Sean Davis]

+1. I would like new stuff at the top. Even if age were the only sorting factor, I think that would be pretty close to optimal. Searching and "similar questions" gets the old stuff when necessary.

Answer 2

[Istvan Albert]

Also added as:

https://github.com/ialbert/biostar-central/issues/134

the new behavior will be activated upon the next server update, sometime next week.