In 2014 Nature looked at the top 10 and top 100 papers with most citations.

Has similar research been done more recently?

I am particularly interested in whether the relative numbers have changed: In the image above 1000 citations would put a paper in the top 14499 of 57798126 papers = 0.025%. Since 2014 a lot of papers have been published (thus growing the pile), but more papers have also been cited (thus growing the part with 1000 citations). Do they grow equally? In other words: Would 1000 citations today still mean it is cited more than 99.975% of other papers? And if not: What are the numbers today?

Ideally the answer would include a percentile function that given a year and a percentage returns the number of citations to be in that percentile that year: percentile(year,pct) = number of citations.

  • +1 Very good question.
    – user135405
    Apr 2, 2021 at 3:33
  • For what it's worth, I suspect there is a huge difference from mid 1990s to mid 2000s to mid 2010s for percentages at the non-stellar level, simply because of being able to search online for relevant papers. But at the stratospheric levels you're talking about, my guess is that this is a lot less pronounced. Apr 2, 2021 at 15:45
  • @DaveLRenfro I have the same feeling, but I would like to know if the feeling is supported by actual measurable data.
    – Ole Tange
    Apr 4, 2021 at 7:47
  • That might not be the best methodology. Papers that have been out for a long time (decades, centuries) can accrue citations over long periods of time. you ight want to look at citations within the first year, within the first five years and within the first ten years after publication, as well as citations overall, to see if citation behavior changes. Otherwise, older papers will always bee cited more often, just because they had a longer time to accrue citations.
    – Polygnome
    Apr 6, 2021 at 7:23
  • "Has similar research been done more recently?" recognition.webofscience.com/awards/highly-cited/2020/… Apr 6, 2021 at 9:55

1 Answer 1


That's a really fascinating question. I looked at the original Nature article (https://www.nature.com/news/the-top-100-papers-1.16224) - my initial intuition would be that the proportion shouldn't differ to significantly from the earlier one (it will most likely differ to some extent as there is no reason to expect the proportional growth of citations and the growth of papers to correlate), but we would of course need to calculate it.

Wouldn't the best course of action be to use a research database where you can filter by number of accrued citations, set that to display papers with > 1000 citations and then divide the retrieved number with the total number of papers in that database? If you you wish to be consistent with the Nature article, I believe they did their analysis solely on Web of Science. Having that said, I'm not sure they provide these figures out of the box, it looks like Nature got special access to these figures for their article.

You must log in to answer this question.

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