Dear community,
first of all, thank you very much for your excellent work and extraordinary educational tool.
I have a question about the difference between "per submission" and "per reviewer" of the random allocation option of Workshop. I have looked for some information, and the potentially closest discussion is in this link: https://moodle.org/mod/forum/discuss.php?d=216305#p942023 Unfortunately, I have not been able to ask my question there.
The thing is that, to my understanding, it should be possible, and not very difficult, to make the random allocator to assign as many reviewers per submission as submissions per reviewer; so, if students that did not submit their work are not allowed to review other submissions, the option X reviews per submission or per reviewer could be ignored.
My idea is to generate X (from "X reviewes per...") permutations of the submissions (easy with the numpy library, for instance). Then, for each permutation, reviewers are assigned to the submissions according to the generated permutation. Since the number of submissions is equal to the number of reviewers, because just submitters are allowed to review, and permutations are bijective functions, each submission is assigned to one reviewer and each reviewer is associated to one submission. This procedure is repeated for the X-1 remaining permutations.
There is still a possibility of assigning the same submission more than once to the same reviewer, or even a submission to its submitter. However, with enough submissions (and thus, reviewers), I think that this is not very probable and the random allocator could replace problematic permutations with new randomly generated ones (within a while loop and a maximum number of draws).
In case non-submitters were allowed to review, then this solution should be fixed, but in any case, you could use your current allocator in this case.
That is all. I just wanted to share this possibility with you. My apologies in case I had missed any relevant piece of information that makes my solution invalid.
Best wishes!