I mark all my students work here, which has proven an interesting challenge since it is all in Mongolian. Also I had to make an adjustment to treat 65% as a passing grade, and deal with what I perceive as very high grade inflation. My students showed me their results from an intensive with a visiting teacher, in which they mostly received high 90s. I wonder if this is an American influence, but this could just be my prejudices.
I do read my students work, but I can usually make a solid estimate of what mark they will get based on two primary factors: if I look at the student's name and the length of response, I have a pretty good idea of what their work will be like. Of course, that is not entirely unexpected, but sometimes it makes me very self-conscious about my marking practices! I do try and mark to some kinds of rubrics, and at times I mark without looking at student names beforehand (alas, there is no 'student number blinding' system going on here), but I think this just reflects the reality that I am pre-aware of my students' abilities and have a pretty good idea of the caliber of work they are going to produce.
I was reading about both the history of grading systems in US higher education, and US colleges that do not have ranked grading recently. I think that the latter systems, which utilise intense written evaluation, would probably be a marked improvement, but I also think that I couldn't operate that way in Mongolian.
Besides which, there are certain requirements here of our accreditation, including 50% of a course's evaluation must be through examinations.
There is a lot more for me to think about and process in the field of assessing work, and if you have any thoughts they would be welcome.