That makes perfect sense 
Now what's with all the talk I see on the forums about the top ranked people in the cohort getting lower-than-expected results, resulting in everyone elses' results being lowered?
well... now we're getting tricky. I will provide what is a little bit of a simplification of what happens because otherwise we would have to use equations and terms such as standard deviation and mean (which i'm no good at)- and also I don't know how to use latex.
Cohorts of a subject are not always equal. Lets use biology as an example. There will always be schools with very good biology teachers and bad biology teachers and there will always be schools with kids who are good at biology and schools with kids who are bad at biology. So it is nessicary to sort out who these schools are and make them equal. Because of the sac system, each school will end up with someone ranked first, seccond and so on. However because some schools are better than others, stuents will not always be equaly matched to their equaly ranked counterpart in another school.
Because of this the exam(s) is used to decide how one school's number 1 ranked student (and all other students) rank against the rest of the schools in the state's number 1 ranked students (and allother sutdents). by this measure, if one school is doing much better than the rest of the schools in the state, then their SAC scores should be pushed up as it is very likely that it was 'harder' to obtain number 1 rank in that shool than it would have been to obtain number 1 rank in another school, therefore the kid at that 'harder' school deserves to have his sac marks pushed up.
Real world example: My legal cohort this year was reasonaby good in the exam. This meant that my SAC scores which were originaly 92/100 and 91/100 got pushed up to 98/100 and 98/100.
note, this system has no impact on a studnet's EXAM score.
DANG! in after ATAR and swarley. Still posting cos it took me ages to write and it's from basics too.