I'm just curious as to how the essays of people who get 50 are ranked against other people's.
Numerically, the people who get 50s have to score 9s and 10s (mostly 10s) on the exam, and being rank 1 helps a lot. But it's not like there's any hidden qualities in an essay written by someone with a 50 SS that aren't present in other pieces - it's just that those high-range essays do lots of different things well.
Consider:
Essay from Student 1: good use of evidence, and a strong contention, but the expression and grammar compromise clarity in some sections --> 8/10
Essay from Student 2: ideas are awesome, but only tangentially relevant to the prompt --> 6/10
Essay from Student 3: ideas are fairly simplistic, but well argued, and the expression is decent --> 7.5/10
Essay from Student 4: contention is relevant, ideas are well-supported and well-expressed, and the piece is very cohesive --> 10/10
Every one of these essays was doing at least one thing right; it's just that the 10/10 was able to combine a whole bunch of right things together.
What criteria do examiners follow?
Is it dependent on the average exam mark of the cohort? (if yes, what average mark is expected of the cohort?)
Sort of.
Explanation here. But I'm not sure what you mean by 'expected mark.' The average in the exam is around 5/10, but the average in your cohort (ie. the marks you get for SACs) will vary between schools.
Is it dependent on the score of the GAT? (if yes, what should be the average GAT mark of the cohort?)
Nope, not to my knowledge. Not unless you don't attend your English exam for some reason, in which case you'll get a derived score based on your GAT, SAC performance, and indicative grade from your teacher. It's in your best interests to give the GAT your best shot but you shouldn't take it too seriously or stress out about it - the marks you get for those Written sections are by no means predictive of your English results.