<p>Each section has an individual curve that differs every time the test is given. In the simplest way possible to explain: the “smart” people get their scores magically translated to an 800, the “average” people get their scores to be about 500, and the “dumb” people get their scores to be about 300-400. The amount of rounding a “smart” person gets becomes the curve for everybody else’s section of that test administration. And CB just throws in the experimental section for fun/make the test even more grueling.</p>
<p>Here is a rough idea of how it works:
There are 100 test takers this SAT.
Smart Guy 1 gets 65/70 on the math section. Smart Guy 2 gets a 64/70 on the math section. CB rounds them both up to get an 800.</p>
<p>Above-Average Joes #1-30 all get a 45/70 on the math. These people are 30% (30/100) of the test takers, and thus did better than 68% of their peers (there were 2% “Smart Guys”). CB says that these guys deserve a 600.</p>
<p>Average Dudes #1-60 all get a 35/70. These guys deserve a 500 because they are just so darn average.</p>
<p>Dumb Guys #1-5 all get a 25/70. These people are way below the majority of their peers, they deserve a 450.</p>
<p>Uber dumb joes #1-3 all get a 20/70. These morons are only 3% of the test takers and did poorly. They deserve a 350.</p>