Unlike the Oct 2020 Reading section, the Math sections fell within expected ranges and distributions of Q types.
The way we think about SAT math is this: it’s a surprisingly closed universe (with very tight boundaries), and what you get on any given SAT is a permutation of these expected Q types.
What we’ve been able to do beyond that, however, is to cluster (group, connect different Q’s through shared principles or visual cues) SAT Math questions to such an extent that 2/3 of SAT Math Q’s for every administered test since 2018 fall within 10 tightly held groups.
For those keeping score, that means that we’re able to predict roughly 40 of the 58 math questions (on both No Calc & Calc sections) from a tightly held rotation of 10 Q Types. That’s not much bigger than an NBA playoff bench rotation.
So – from a test that seems messy and unpredictable – our data can bring a high degree of clarification and predictability.
Now – two quick points:
These SAT math patterns (reflected in Q Types) need to be applied/practiced/drilled to get to the place where a student feels a high degree of confidence in recognizing – and almost automatically – responding to them.
There is variance in the Q forms within each group. For example, some of the harder Q’s on the Oct SAT asked you to recognize a simpler math principle when it was presented in what looked like a complicated form. But that’s what our visual cues come in – when you see the format, you cut through the static and look for the simple, automated response.