last day (79 days later) » 

12:34 AM
room topic changed to About HNQs: General discussion of HNQs, removal from HNQs, SEDE queries, feeds [hot-questions]
It seems that now there is quite lively discussion about HNQs (and specifically ofr removals from the HNQ list): Under what criteria should we remove questions from the HNQ list?
It seems that so far mainly Theo Bendit, quid and Asaf Karagila contributed to that discussion.
I have created this room partly because some part of the discussion might be too long to comments - so if something feels like more suitable for chat it can be discussed here.
And partly to look at various SEDE queries and other ways to check past HNQs.
5
Q: Under what criteria should we remove questions from the HNQ list?

Theo BenditSince the Hot Network Questions list was revamped 6 months ago, moderators have had the power to remove questions from the list. As Catija, the community manager for the SE network says, I recommend each site have a meta discussion with guidance for moderators about when - if ever - a questio...

I have already posted one such query in the comments:
I suppose that for the purposes of this discussion it might be useful to check which questions have been removed from the HNQ. Here is a SEDE query which lists such removals: data.stackexchange.com/math/query/1013246/…Martin Sleziak 13 hours ago
But perhaps let us start first with all questions that were in HNQ, not just the removed ones.
47
Q: How to determine if a question has ever been a "hot network question"?

neubertI've had at least one question featured as a "hot network question" and during that time it got thousands of views in the course of a few days. But if you look at that question now you won't have any way of knowing that that's where the bulk of its view count came from. Is there a way to find ou...

Since February 2019 there is Hot Network Questions room. It's based on a feed created by rene. I think it's relatively reliable.
You can find the questions from Mathematics by looking at the feed posting those questions (if you want to see more, you have to go to page 2, page 3, etc. by altering the url manually) or by searching for questions from this site.
Since end of February, becoming HNQ is now part of revision history and timeline and it is also possible to check this in SEDE: Questions which were in HNQ.
It's not difficult to restrict the list to some time interval: Questions which were in the HNQ between two dates
The discussion is around the questions which were removed from the HNQ list and the questions which were in the HNQ list but are of low quality.
Removal of a question from the HNQ by a moderator is recorded in post history, they can be found easily using SEDE: Questions removed from HNQ by mods
 
1:03 AM
It might be interesting to see also time between creation and removal. Here is a SEDE query which I tried: Questions removed from HNQ by mods - with time of entering end removal
I don't really see what mistake have I made - but in this query the username of the mods is not shown. (It is shown in the previous query.)
For comparison, from SEDE (=data since end of February 2019) we get 1658 questions in the HNQ list, 399 mod removals.
Oh, now I noticed what the mistake was. Here is the corrected query: Questions removed from HNQ by mods - with time of entering end removal
A question might be removed from the HNQ list by editing MathJax into the title. This can be more difficult to find using SEDE.
I have tried at least a rather simple query - I think it would be more complicated to show eliminate those ones where title was edited several times. But if somebody has some time, you can try to do some improvements.
It returns 397 results, but there are some false positives.
Here is one example where the title was edited before the question entered HNQ.
After adding the condition that the edit ahs to be after entering the HNQ list it is down to 392: Questions removed from HNQ by editing MathJax into the title
But title might have been edited several times, I am not sure whether this query is really reliable.
Question closure might indicate lower quality of a question.
So perhaps these queries might be interesting in connection with this discussion: Questions which are closed and were in the HNQ and (128 results) Questions which were in HNQ and were closed at some point (53 results)
Needless to say, SEDE does not have PostHistory data on deleted questions - so there might be a bunch of questions which were in HNQ and were later deleted.
Asaf wrote in his answer:
> I account for the majority of uses of this feature across the entire network (to the best of my knowledge).
I know that it is possible to write queries that look at all sites in the network. So it is possible to check this in SEDE, if somebody is interesting.
But I am not going to try this right now. (It's getting late and I have some other stuff to do.) If somebody does this, feel free post the query here.
Hi @quid - as I saw that there was quite extensive discussion in comments, I thought that it might be useful to create also a chatroom for this, I hope it is ok.
 
1:25 AM
@MartinSleziak well. Likely a better place for that. I think I'll leave soon for now.
I mean the chat is likely a better place the comments so it is a good idea.
 
I have posted one SEDE query related to HNQs in the above comment. You can find a few more queries related to this topic in this chatroom. Feel free to use the room also for any related discussion which seems more suitable for chat than for comments. — Martin Sleziak 10 secs ago
I left the above comment under the question so that Theo Bendit knows about this room.
Do you think I should notify also Asaf Karagila in some way? (E.g., inviting him in this room?) Although I'd guess that Asaf will notice the comment about chat when he gets back to the discussion on meta.
@quid Yes, it's already quite late in European timezones. (In fact, I fell asleep relatively soon, and I woke up in the middle of the night - that's why I'm here at rather unusual time.)
 
@MartinSleziak I'm aware of this chatroom's existence now. Thanks Martin!
 
Here is one more query which shows also the interval between entering HNQ and closure: Questions which were in HNQ and were later closed
I will stop playing with SEDE for today.
 
@quid I don't know why you think I'm not interested in how HNQs are handled? Maybe I should have explicitly mentioned this in my question, but I was only ever friendly and encouraging to Asaf posting his answer about the current state of play. It was a minor omission at worst.
 
@TheoBendit If you don't mind me asking what timezone are you in. I am in Slovakia and as far as I know quid is in a similar timezeno.
I am asking mainly because it is good to know when the others who entered the discussion have daytime (and are more likely to respond).
 
1:40 AM
I'm in GMT + 12:00, aka New Zealand. It's about 1:40 here. I encourage all interested parties to get some sleep. :-)
 
Although in the case of users who are rather active an chat - like quid or myself - one can guess also from the chat profile their most active hours.
On my part, I agree with that suggestion :-)
 
See you later!
 
@TheoBendit actually I think you are mostly right regarding this and I picked up on a detail that annoyed me and got hung up on this.
For reasons explained by Martin I'll leave now.
 
Possibly I should have included link to local time in New Zealand, too. (This is one of things that might make interaction between the users on the site complicated at times.)
 
 
2 hours later…
4:04 AM
There are various stats. And one answer also contains a SEDE query (created by Glorfindel) comparing various sites.
17
A: What is the impact of the recent changes to how Hot Network Questions work?

GlorfindelLet's start off with this SEDE query showing the number of times the ♦ moderators have used the ability to remove a question from the HNQ list. It has been used about 120 times in the past 6-8 weeks; more than half of it by the Mathematics mods, but to put things in perspective (thanks @BillyMail...

Mathematics has highest percent percentage (24.2%) and highest number of questions removed (399). At the same time, Mathematics is also the site with the highest number of HNQ questions (1649).
 
 
1 hour later…
5:09 AM
This is only tangential, since it is about another site - I have posted this question on MathOverflow Meta not too long ago: Should moderators in some cases remove a question from the network-wide hot questions list?
There was actually not much discussion - just a few comments. MathOverflow seems to be on the opposite end of spectrum - so far no MO question has been removed from the HNQ list. (Despite the fact that, in some past discussions, some MO users have complained about the influence of the HNQ questions on the site.)
 
5:43 AM
And here should be the same list, but with the ones that weren't removed left in, to give a better idea what proportion are being removed (not as many as the 75 suggests; they get about ten Hot posts a day, on average) — Billy Mailman May 8 at 12:36
This query shows both the question that were removed and that weren't removed: Questions removed from HNQ list.
@BillyMailman it's worth keeping in mind that Math.SE folks have (and use) yet another way to drop questions from HNQ - by editing MathJax into their titles. This can be roughly estimated by counting how many of their hot questions contain two dollar signs in the title — gnat May 14 at 13:20
I have used PostHistory table to check for the edits of the title. But using Posts.Title could be another way how to get some rough estimate of questions which were removed from the HNQ list by editing the title.
This comment also mentions MathJax in the titles:
as for many questions from Math.SE, I think I got it. Thing is, they are enthusiastic about improving titles. Questions at sidebar catch attention of title editors and if their changes introduce MathJax this drops question from the list, so that next "candidate" takes its place - which catches the eye of editors again and next change follows, and so on. You can verify my hypothesis by checking how many of their hot question titles contain two or more dollar signs (ie MathJax). Given site topic I would expect there will be a lot — gnat Jun 28 at 6:53
 
 
1 hour later…
6:52 AM
Theo, really, your example is "Discrete Math easy question"? That title sucks. That title tells you almost nothing about what's going on with the question. And lo and behold, the question was given a proper title which would have excluded it from the HNQ list anyway. As I said, sometimes I don't have the time to edit, and my reasoning for removing something from the HNQ list is due to the fact that a proper title would do that anyway. — Asaf Karagila ♦ 1 hour ago
3347226: Are $\{\{a\},\{b,c\},\{d\},\{e\}\}$ and $\{a,b,c,d,e\} $ the same set? - this one became HNQ with the title "a Discrete Math question on sets". That is certainly rather non-descriptive title. Any reasonable title I can think of contains MathJax.
This is example which I also had in mind as a question where I see nothing wrong about removing it from the HNQ list.
Although, it is certainly true that any user who noticed that question could have edited the title to something more sensible. (And as soon as the question became HNQ, many users must have seen it.)
One example from the "other side":
3346648: Why do we buy the Mazur Swindle in knot theory? - I would have no objections if this question remained in the HNQ list.
Of course, there are many questions which were in HNQ - so there is plenty of specific examples to look at.
 
 
3 hours later…
9:43 AM
@MartinSleziak When creating this query, I should have used ph2.UserId rather than ph.UserId - to show the user who edited title. So here's the corrected query: Questions removed from HNQ by editing MathJax into the title
If the title was edited several times after entering HNQ, it shows all edits - but probably it's better to have slightly simpler query - even though some questions will be shown twice.
 
 
1 hour later…
10:45 AM
They even created a separate room for HNQs: The Outreach Department
in The Outreach Department, Apr 23 at 8:57, by Wrzlprmft
room topic changed to The Outreach Department: … where a feed of hot network questions gets posted so we can edit, close, and flag them (as appropriate). (no tags)
It seems that the two most extreme answers (remove all/keep all) in that thread ended up with the score near zero:
1
A: Moderators can now exclude questions from the HNQ list – when should we use this?

Federico PoloniWhy not "all of them"? I'm not sure myself that this is the perfect solution, but take this answer as a way to think about the benefits of HNQ. What good comes to our community from a question being in the HNQ, usually? Most of the time, all that happens when a question enters it is that we get...

0
A: Moderators can now exclude questions from the HNQ list – when should we use this?

Massimo OrtolanoPlease, don't remove questions from the HNQ list at all I think that the HNQ is a non-issue and we should not exclude anything, just leave the algorithms do their work (whether good or not). So, I add this as an answer because I think that we should give the community also the possibility to cho...

 
 
1 hour later…
12:09 PM
Several comments mentioned that ideally HNQ should showcase good content available on the site. This reminded me of plans to use bounties to promote the best content, which was suggested by some users.
It seems that the plans are now on standby, probably those users have now other duties. But it would be nice if this happened one day.
 
12:59 PM
I've read the thread on meta, and am tempted to provide an answer, but I am not sure that I have an entirely cogent response.
My basic feeling is that when I click on HNQ math questions, I am almost always disappointed.
They almost never reflect what I think of as good mathematics, or good, interesting questions. Instead, they tend to be homework-style questions with quick, easy answers provided by half a dozen folk.
More often than not, I click an HNQ, am disappointed, and edit the title to add necessary TeX. :\
I think that HNQs should showcase the best of what a site has to offer. However, in general, I do not think that the majority of the questions that I see on the HNQ from Mathematics represent the best of what MSE has to offer.
I am all for aggressive pruning.
 
 
1 hour later…
2:17 PM
Is it possible to add questions on HNQ? I don't really know how the system works (in order for a question to arrive on HNQ), but perhaps there are also question worth finding there. Also since there are many disappointing questions that gets on HNQ, adding in exchange for a removal, a good question to HNQ might compensate things.
 
@Zacky The question that get into HNQ are not picked manually, they are chosen by an algorithm. I do not know the details of the algorithm - but you could certainly find some details on Meta Stack Exchange if you're really interested.
From that post:
> The data will help us see how effective changes to the formula are as we make them. You'll notice that we're not changing the formula itself for the time being and that's intentional. There have been some really great and interesting ideas floating around for a while, and we may try them out and see how they impact which questions are selected and how long those questions stick around.
 
I might be wrong, but I think I saw a mod (Asaf) add a question on HNQ recently, but maybe I didn't saw clearly (I can't find the question now).
Anyway this might not be a good idea.
 
Moderators can remove question from HNQ. Not add them.
 
Yeah, I'm probably blind.
 
Did you perhaps mean tag on meta. That can be added by mods and it adds the question to the . Here is list of the questions that have been featured: data.stackexchange.com/meta.math/revision/1066745/1318346/…
 
2:28 PM
Oh boy, Asaf made the question CW not HNQ. Aplogies as I thought this might be a mod option!
 
Yes, mods can make question CW. Any regular user can make their answer CW - if they wish to do so.
 
Thanks! Well, just ignore the above messages^.
 
3:00 PM
I know that I have posted already many SEDE queries, but perhaps it makes sense to look also at the original title. (In many cases, with this title the question enters the HNQ.)
 

  last day (79 days later) »