@Glen_b: I thought duplicates were frowned upon, because I thought the were percieved as a public display of laziness. But I totally agree with your view, when I think about it. I have indeed been helped more than I have been annoyed by a duplicate. Letting shoulders sink down again

@Erosennin: I sometimes up-vote duplicates: when they ask a question in a completely different way than it's been asked before. (But I would down-vote those that ask a question in exactly the same way as another answered question - I don't just because it's usually new users who do that & it could seem unfriendly.)

@Scortchi The ones that annoy me are when experienced users post answers to obvious duplicates. I don't downvote such answers, but I don't upvote them, either. I would much rather that people who know this site and have interacted for a long time share that experience by identifying duplicates and closing them as such.

It is by creating such a network of links that real information is built up.

@Scortchi , @whuber: I like how it seems like it is a genuine interest in spreading information in this community, and particularly how this materializes in a fine and gentle treatment of new users. But it surprises me that experienced users would post answers to obvious duplicates, in what I assume is an attempt to get upvotes. Considering how experienced users should be experienced because their passion for the field and not the points. But I guess this behaviour plays out everywhere...

@Erosennin In most cases I think these are not attempts to garner points. It's just easier to answer some questions than it is to find previous answers! The system offers no rewards of any kind for identifying duplicates, either.

@whuber: I hope you are right :)

@Erosennin: As well as being easier it's more fun. And frail memory can play a part. Once I read a new question, & had a vague recollection of a duplicate: I searched & found two, both answered by me.

Howdy gung!

Lol, hey @Scortchi. I've also thought questions seem familiar & found my own.

:)

There are times when I could swear a similar question was asked, but just can't find it, as @whuber notes.

("Howdy" because that's what sheriffs say, I've been led to believe.)

"Howdy, pardner" is presumably more correct.

I don't think crude reputation seeking is behind much repetition. People desperate to boost their reputation here and irritated by downvotes or other flak usually don't last long. It's like Hanlon's razor of assuming stupidity rather than malice. If anyone has posted a fair number of answers in various places over a fair period, and read many, many more it becomes difficult to remember specific answers. Naturally, one should always search.

I often see questions (especially on forecasting and time series) that are close analogues to earlier questions. However, there usually is some appreciable difference, at least to me. So I'll often answer such questions, linking back to earlier answers, while other people might close the questions as duplicates. Whether two questions are duplicates is often not completely clear-cut.

Especially not for less experienced questioners, who might well be scratching their head trying to map an identified duplicate to their question.

And to be honest, I think that a new answer to a slight variation of an existing question may well add more value than a straight-up closure as a duplicate. I know that I have often profited from having something explained twice in slightly different words.

(As long, of course, as the new answer links back to earlier answers.)

I'd understand it if other users don't agree with this point of view, though.

@StephanKolassa As someone who has closed many questions as duplicates, I find myself nevertheless gravitating towards your point of view. I now look for evidence of something a little new or different, and if I find it, I won't vote to close. I left several questions open on that basis just today.

To be honest, meditating on what I remember of the last couple of weeks, I can't recall when I last cast a first vote for a duplicate (I may have voted for duplicates through the review queue). This probably indicates too strong priors on "not a duplicate" on my part. I'll try to update them.

@StephanKolassa: I try to distinguish between when there's a slight difference in the questions, which can be dealt with by a giving a brief answer focusing on the difference & referencing the duplicate thread; & when the questions are similar enough that an answer to one would serve as an answer to the other, & if there's anything to add it might as well be added to the earlier.

But linking's the main thing.

Here is today's example: stats.stackexchange.com/q/214770/1352 is IMO certainly not a duplicate of stats.stackexchange.com/q/214118/1352, but it's close enough that, again IMO, the answer to the latter is quite sufficient as an answer to the former after the OP's clarifications. Looks like this would satisfy @Scortchi's conditions, would it?

In RMSE-land I'll take my stand, to live and die in RMSE.

Away, away, away down south in RMSE.

@Kodiologist You can't take the RMSE from me.