It also plays with statistics quite a bit (lies, damn lies and government statistics been the old refrain).
All of that is putting aside that IQ itself is deeply flawed both because a) you can't reduce human intelligence down to a single number that way at best it's a weak measure of the type of intelligence been tested for b) IQ tests vary, both in type and scope and are deeply flawed when applied across socio-economic groups in the same society never mind across societies.
There is always variation within a population but I suspect that culture plays a much bigger role.
Case in point - Jewish people are 0.2% of the world population, they've won ~20% of all Nobel prizes awarded.
So the Nobel prize committee might not actually be racist ... :-)
But before measuring, some referee has to take the ball from the player who was holding it at the end of the play and put it down at its official spot from which to measure. But they do this part so quickly and carelessly. It's amusingly incongruous with the pomp of the measurement.
This makes me think of how data science operates on whichever data set happens to be conveniently available and then conducts very serious and literal downstream analysis as if the raw data is settled reality.