In general, people don’t request a review of marking (a) if the mark is better than expected, (b) if it is much as expected and well below the grade boundary for the next grade up, (c) if the mark is a few marks below the boundary, but it’s clear there are no errors, (d) if they underperform a bit, but in similar ways on two or more papers.
People request a review of marking (e) if they, their parents, teacher or school spot a specific error - say a missed question, addition error or a question marked wrong that is clearly correct, (f) if they did really well on one or more papers, but scored very low on one paper, or (g) if they are one or two marks below a grade boundary. Some of (e) and (f) will be close to a boundary. There are more requests for reviews from fee-paying schools than from state schools, and more from high-achieving schools than from more average schools.
For these reasons, where a review of marking is requested, there is more likelihood that there is a genuine error, particularly in the case of (e) or (f), and a mark is more likely to go up than down. Then people post to social media, tell all their friends, and generally put it about that the marking is rubbish, and attribute it to lack of training, inadequate pay for examiners or some other problem with the system.
In fact, examiners do have to do training. Experienced examiners still have to retrain every year. New examiners have a reduced workload so that they can spend a little more time on each question until they build up experience. All marking is supervised, monitored and corrected if necessary by senior examiners, with one to one feedback throughout the marking period, and the payment is, if not generous, certainly satisfactory. In addition, exam marking provides teachers with excellent professional development experience at no cost to them or their school, and they get paid for doing it. (CPD, which is generally compulsory, can cost anything upwards of £250 per course, so examiners are effectively saving the school money.). Many examiners mark more than one paper (eg, GCSE and A Level for their subject) and return year after year. The exam boards and senior examiners are very supportive, provide performance feedback, and encourage examiners to develop their examining career alongside their day job.
Last year, 4.6% of grades were challenged. Of those, 75.9% were NOT changed. That means that just under 1.1% of all grades were changed after a review of marking - that’s just a wee bit over one exam for one student out of 100 exams. In my view, just under 99% accurate marking isn’t bad.
Oh, and while posters are being dismissive of examiners’ competence, do try to remember that some of your own children’s teachers may well be GCSE or A Level examiners. They can’t disclose that to you because examiners are supposed to keep their role confidential except on a need-to-know basis where their head of department/ SLT need to be aware.