I've been at this for a while and I can't think clearly so I'm definitely doing something wrong.
The question:
Identify the symmetries of the curves in Exercises 1–12. Then sketch the curves.
$r = \sin (\frac{\theta}{2})$
In the book it states that first to find symmetry we have to check the following:
Symmetry about x-axis
$$( r ; - \theta )$$ $$( -r ; \pi-\theta )$$
Symmetry about y-axis
$$( -r ; -\theta )$$ $$( r ; \pi-\theta )$$
Symmetry about origin
$$( -r ; \theta )$$ $$( r ; \pi+\theta )$$
So I checked for the first one and this is what I got: For $-\theta$:
$r = \sin (\frac{-\theta}{2})$
$-r = \sin (\frac{\theta}{2})$
Which turns out to satisfy the y-axis.
Now here's what I don't understand. I want to check for $\pi-\theta$, So I replace that where $\theta$ would be.
$r = \sin (\frac{\pi-\theta}{2})$
$r = \sin (\frac{\pi}{2} - \frac{\theta}{2})$
Following the formula for sin(A-B), then:
$r = \sin (\frac{\pi}{2})\cos(\frac{\theta}{2}) - \sin (\frac{\theta}{2})\cos(\frac{\pi}{2})$
$\cos(\frac{\pi}{2}) = 0$ and $\sin (\frac{\pi}{2}) = 1$ then:
$r = \cos(\frac{\theta}{2})$
Which obviously doesn't make sense. So I checked online and found that it should actually be $r = \sin (\pi - \frac{\theta}{2})$ and then that it would satisfy the symmetry about x-axis. But I checked it (I'm sure by now I've done something major wrong):
$r = \sin (\pi - \frac{\theta}{2})$
$r = \sin (\pi)\cos(\frac{\theta}{2}) - \sin(\frac{\theta}{2})\cos(\pi)$
$\cos(\pi) = -1$ and $\sin (\pi) = 0$ then:
$r = \sin (\frac{\theta}{2})$
Which satisfies y-axis and not x-axis.
So my question is this, why is it that wherever I look it should be $r = \sin (\pi - \frac{\theta}{2})$ and not $r = \sin (\frac{\pi-\theta}{2})$ and what did I do wrong in the proofs and how can I make sure not to make the same mistakes again?
Thank you.

Your computation is correct, but your conclusion is wrong. The equation has failed the symmetry test, but that does not mean that the graph is not symmetric with respect to x-axis. As a matter of fact if passing a symmetry test verifies that symmetry will be exhibited in a graph, failing the symmetry tests does not necessarily indicate that a graph will not be symmetric. In others words, the graph of a polar equation can be symmetric with respect to one of these axes (or the pole) and not satisfy any of the test equations you wrote. This is because a graph can have many polar representations, so many tests are possible. There are many different ways of specifying a point in polar coordinates. You can realize this fact by thinking that every point $(r,\theta)$ could also be called $(r,\theta+2n \pi)$, with $n$ any integer. This fact can affect a symmetry test, even if I think it is possible to define alternative robust symmetry tests.