Basically, the paradox works like this:
Two entangled particles (A and B) are a light year apart.
1. Measure the X spin of A 2. Measure the Z spin of B
You now know (the paradox claims):
a. B's Z spin, because you just measured it. b. B's X spin, as it must be opposite of A's.
Of course, this turns out not to be true. However-- what if, after measuring B's Z spin, I immediately measure the B's X spin. What is the probability of it being, at that point, the opposite A's X spin (as the paradox claims to be the case)?
If that probability is not 100% (as the paradox's refutation claims), then I can determine that B's X spin is in the same direction as what I had originally measured A's X spin to be.
So then, what would happen if, at that point, I then measured A's X spin (a second time)?
My point -- Is it possible to measure A and observe its X spin, and then flip A's X spin by merely measuring B's Z spin? Has this been recorded experimentally?