Here is my theory of #Reviewer2 – I’m curious to hear whether you find plausible!

It is commonly claimed that in any set of reviewers, #Reviewer2 is the really mean one. Here’s why:

You submit a paper to a journal. The journal uses an algorithm to assign reviewers. The algorithm is based on a comparison between the paper’s reference list and reference lists of published papers. It results in a ranked list of reviewers, with the best matches at the top. (1/4)

@academicchatter

Taking into account the typical “decline” response rate, at least 6 very suitable reviewers from the top of the list will be invited. These are busy people, so only one of them accepts the invitation. This is #Reviewer1, who will write a helpful and constructive review. (2/4)
People lower down on the ranked list, who are even less likely to accept, are invited to review. At some point, one of them will accept the review for reasons that have nothing to do with your paper (boredom, feeling guilty, bad day…). This is #Reviewer2, who will not like your paper and tell you that you must change absolutely everything in it – to turn it into a paper they would have liked to read. (3/4)

Next, the editorial office checks the status of your paper and finds out that only two reviewers have accepted the invitation although the journal policy requires three. They contact someone who owes them a favour or a member of their editorial board. This person will take only a quick glance at the paper and write a positive, but generic review. This is #Reviewer3.

So – what do you think of my theory?? (4/4)

@landwehr_c not my experience as editor. I rarely used algorithms to select reviewers and the few times I did it was only for the last reviewer (but really very rare). Often the perceived friends of the authors (the ones they recommend) are the harshest reviewers.

you are right on one thing, though: it is very hard to get anyone to accept reviewing. Most of the time a paper spends “being reviewed” is the editor trying to get anyone to accept and then chasing the ones who did…

@gepasi Interesting to hear the editor perspective! But maybe the argument would work without the algorithm: the point is that you have a ranked list of reviewers and that the first to accept is more likely to be helpful and constructive, because they are really interested in the paper, while those who accept later and reluctantly are also less interested and more likely to behave like #Reviewer2
@landwehr_c this may be true, but I think there is a large random component here
@landwehr_c @gepasi An editor’s point of view: reviewers do not know whether they are 1 or 2, whether they were selected first, second or at last.
@marianallanos @gepasi
Not when you are invited, but I usually receive the editorial decision along with the reviews afterwards, so I can see which of the reviewers I was.