I have read the descriptions (https://docs.moodle.org/26/en/Using_Workshop#Grade_for_assessment) on how frade for grading is determined. Having done a workshop with my class now, I'm eager to understand more about how it works, so that I can explain this to my students. There can be some bizarre behavior, and this erodes the credibility of this system in the students' eyes.

I can understand how a "best" paper is identified, and this is assigned 100%. I can also understand that the settings of strictness determines how the grades fall from that. Is there a way to understand how this strictness works? I had students grading a number of errors strategy with 10 components (pretty much equal weighting). For those who aligned perfectly with my grading (and I weighted mine higher), they got 100. But if others deviated just a small amount, some of them got zeros. I don't understand how this happens.

The documentation is clear on how it assigns 100%. But I don't understand how it assigns a zero - I suppose a theoretical terrible response is one that is completely opposite of how the average is graded. Say I graded a paper with only 1 other student, but I weighed my paper 10 times the weight. It seems that only mine can be assigned a grade of 100, and this student will have a lower grade. If he grades completely opposite from me (as far as determining the binary decision of whether or not an error in the assertion is marked), then I suppose that is a theoretical zero. But How else does the algorithm determine how low to grade this student? There must be a logic of how strictness is tied to the lowest grade possible among the bunch. Can this be described with greater clarity?