From facebook/youtube it is known that if a video/post gets a distinct amount of likes, at some point the overall like number of this video is going to explode due to a network and avalanche effect.

I was recently wondering while thinking about bibliographic metrics like h-index, how much such effects apply in 21st academic publishing and also up to which decade (before internet existed, before strong networks in internet existed) this effect might also retroactive have pushed for instance publications of the 95-00 more than 90-95 years, also the real impact actually was not much higher. Probably this effect can be identified in time by looking when number of worldwide citations exploded, although the number of researchers just grew steadily.

Is the bibliopgraphic science investigating such questions? Any books or review papers I could read. I'm especially wondering which open source data sources there are to tackle such questions, if there are. Best I could google is this:


I'm asking as I'm wondering which citation count is actually necessary in some fields to see if a paper really had an impact on the community, didn't penetrate it, was mediocre. This will depend on community size and field. But I'm wondering how to estimate it, kind of Fermi-problem.

1 Answer 1


This seems like a naive view of citations. They aren't "likes" or even recommendations to read a paper. They reference things that actually support and form a background for a new work.

So, if citations suddenly explode within a field, and people are working honestly and not just gaming the system, then that field must be pretty hot with a lot of actual research going on.

And gaming the system outside a small working community is pretty hard since a lot of people need to cooperate.

But some fields are indeed hot for a while, with a lot of research, hence citations. But then they cool. Some future discovery may heat them up again and the cycle repeats. An a fresh discovery can often lead to new ways of looking at a problem and that, alone, will generate new ideas and new papers and new citations.

But a small citation count doesn't necessarily mean that a paper is unimportant. It may be incredibly important, but only to a small set of researchers. On the other hand, a less seminal paper might have a high count if it appears in a popular field.

And a there are a lot of people who aren't drawn to popular fields, thinking they will be just a face in the mob.

But research in general is hard. Facebook requires little actual intellectual effort. A paper in Nature requires quite a lot.

“If we knew what it was we were doing, it would not be called research, would it?” ~~Albert Einstein

  • I didn't compare academic publishing with facebook, but the technical mechanisms behind both social networks are probably very similar nowadays with papers, blogs, academic news sites etc interwoven... And this might have implications I'm trying to pose and question. You probably heard of citation cartels and you mention gaming of the system too. Even of academia.se you have strong avalance effects, for answers that seem important, but that are sometimes only very popular among the voters. If this sometimes becomes more and more often is rather my question and maybe there are bibliograpic Oct 30, 2019 at 22:47
  • Let's exclude niche and small research communities, I'm talking about the big questions/communities, condensed matter physics, quantum computing etc. where papers undergo avalanching to citation counts of 2000-3000 at some point. But where does this start? Oct 30, 2019 at 22:48
  • Do you really want to count "citations" in blogs and news sites as the equivalent of references in scientific papers? The frequently mentioned "Russian" troll farms can run up citation counts pretty fast, I'd guess.
    – Buffy
    Oct 30, 2019 at 22:53
  • 1
    When, for example, the Higgs Boson gets confirmed the paper causes a stir in the popular press. A lot of people will say "Hmmm, interesting." But how many people seeing that papular account actually go and read the paper? Understand it in detail? A dozen? So, is the paper "popular"? How many could even give its title the next day? How many actual researchers get on a plane to CERN? There is hardly ever an "avalanche" in research in such cases. Some, yes, and occasionally, yes, but not generally.
    – Buffy
    Oct 30, 2019 at 23:14
  • 3
    I think that's the ideal, but in reality, at least from what I've seen in my field, is that there is definitely an avalanche effect of sorts due to the way many authors write introductions. They are often too lengthy and too general. Authors will simply cite 5 to 10 papers which are not necessarily related to the problem they're tackling but rather serve as a general introduction to the field. These papers are generally highly cited and often not particularly groundbreaking.
    – Vibex
    Oct 31, 2019 at 10:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .