Would it be bad form for me to create a sockpuppet account so that I can add a "test" answer to my latest question/self-answer? I have a sneaking suspicion that questions with only self-answers are blocked from the Hot Network Questions list

The best (only?) way for me to confirm this, without help from a developer/someone in the know, would be to add a test answer from another account (causing my question to appear on the HNQ list) and then removing it (causing my question to disappear).

If this wouldn't be appropriate, that's fine; I can patiently wait for my question to get a canonical response. @Caleb @JonEricson

@Nathaniel Role number one of sock puppet accounts: ask for forgiveness not permission.

This must just be curiosity. How does knowing help you participate?

@fredsbend I'd like to confirm so that I can make an actual feature request on Meta if indeed questions with only self-answers are ignored

I'm not sure how such a proposal would fare, but I'd argue that a self-answer shouldn't be automatically disqualified; if the community thinks it's good then it should be on the HNQ list regardless of authorship.

@Nathaniel The second answer would boost its score, but I suppose you will take that into consideration.

Strictly speaking, the meta post explaining the formula doesn't say deleted answers don't count. One would think that to be the case, but you never know.

@Nathaniel You could try that. But it might be easier for me to look at the code. ;-)

@ThaddeusB I've been recording data on this for a few weeks, and I've seen several examples, on this site and other sites, where Q score = 3 and A score = 3, with one answer, and the "hotness points" come out to 13.118.

@JonEricson If that's easier, then great!

@ThaddeusB Right now the question on the list with the fewest hotness points is 10.9, so theoretically my question with 3/3 should be on the list

@Nathaniel What? Badges and rep ain't enough for you? You gotta game the hot list too?

@fredsbend Need... more... rep...

@Nathaniel You know, they proposed a 12 step SE once. You know, for people like you that need the help. Closed as self-defeating.

@fredsbend Yeah I think I saw that on Meta

But seriously, I like using any means we have to attract eyeballs and (hopefully) contributors (and I'm not the only one who writes self-answers :-p)

@Nathaniel For the most part, the hot list gets lots more views, but very few more votes. It has to actually be something of general interest to get the extra votes.

@ThaddeusB Yes, I've noticed that too. Ironically, if I'm right, that Xmas question that was so popular would have never gotten off the ground if there hadn't been a duplicate answer posted 10 minutes after the OP self-answered

@Nathaniel Also, my impression is that when this site gets eyes from across the network, it's usually a moderation burden more than anything else, rather than a draw for new users. But you're the joker with three weeks of data. Maybe you can report back in three weeks more.

^^^ Free flags.

@fredsbend Yes there's definitely some of that. I haven't noticed it being too much of an issue in the HNQs we've had in the last few weeks. And it's tough for mere users to be able to tell when new contributors are first exposed to the site (if by HNQ or otherwise).

@JimG. Shhhh, there are moderators in here :)

@Nathaniel It looks like Freemason posted the question and then the answer (i.e. didn't use the answer now check box). There is some chance that is also a factor (like maybe there is a bug that prevents the system from recognizing the answer that comes simultaneously.) @JonEricson if you are looking at the code, you might want to check that possibility too, if there is not an obvious "self answers don't count" component.

@ThaddeusB True, I haven't tested that possibility yet.

@fredsbend This post argues that if you get one new user out of a hot network question, it is worth the extra moderation b/c new users from other SEs are of much higher quality than new users randomly discovering the site.

@ThaddeusB That is certainly true. One thing I always check if I see a new user with 101 rep is their other SE profiles. If they are regulars on any other SE then I know they know how SE is supposed to work.

And if they find the topic interesting, then getting them to post regularly comes without a big learning curve.

@Nathaniel @ThaddeusB @curiousdannii @Caleb @El'endiaStarman @TRiG @MattGutting

@LeeWoofenden and I have been talking about the marriage tag, whose current usage guidelines and excerpt phrasing I spearheaded in a meta post last year.

We've come to an agreement that slight change may prove beneficial and would appreciate your opinions before we make a meta post.

Start at this chat message.

There's a lot of extra fluff, but the gist of what we get to ends about here.

@fredsbend Yes, to me just reading the end would have been sufficient. I don't have an objection to [marriage] being a high-level tag that often, but not always, refers to "traditional" marriage, so I'm fine with that modification.

@fredsbend Yeah, that makes sense to me too.

@ThaddeusB In the database, an answer is an answer. I don't think we even track self-answers-via-checkbox. (They should have the same creation date and time, though.) I think the issue is that the ways of the Hot Question list are fickle.

@JonEricson The formula seems straight forward... this one is x + y(7/5 +3)/12.286. X = 1.118 and Y=40.95 for small/medium sites. 12.286 is 1.4^6, since this is less than 6 hours old. So it calculates to 15.784, which is what appears in the tooltip of HNQ list.

Theoretically a question with 4/4 on a similar site (like ours) would be x + y(4/5 + 4)/12.286 = 17.117.

But I can drop it if I'm being a nuisance :)

Sorry, that's 6^1.4 = 12.286 (which is what is in the official formula), not 1.4^6 (typo).

@JonEricson That's why I phrased it as a (potential) "bug" - I certainly didn't think it would be intentional.

@Nathaniel Interesting; I had assumed each site had a different multiplier

You could go through the entire 100 question list & see if there are any self-answers

@ThaddeusB Our X and Y is the same as UX, Boardgames, and CompSci, at least

Have you verified that it counts as 2 answers if there is one self-answer and one regular?

@ThaddeusB Yes it does -- if a crap answer gets added to a good self-answer, the question shows up on the list

@Nathaniel With the full points for two answers, I assume?

@ThaddeusB Yes, it gets the bonus for two answers. I haven't been able to tell if the formula prevents a negatively scored second answer from reducing the overall score through the "total answer score" variable

@Nathaniel Hmm, that is rather strange - I could see potentially not counting it as an answer under either circumstance, but not counting it as an answer only in the case of 1 answer is odd.

@ThaddeusB I haven't seen any, and I've looked at a bunch, but that's not conclusive, since they are such a small minority (probably even more so on other sites).

@ThaddeusB Exactly

@Nathaniel Yah, not definitive since they are rare to begin with. Unless @JonEricson comes up with something, you'll just have to continue to reverse engineer it I guess. Next thing to try is a delay in answering, then the sock puppet.

@ThaddeusB Yeah, that's what I'm thinking.