That assumes "rigging" is the only non-random component. There's also unintentional bias; when asked to name a random number between 1-100, multiples of ten are underrepresented due to our bias.
It solves the rigging problem, it doesn't necessarily make claims as to whether that is the only problem that exists.
It solves the rigging problem, it doesn't necessarily make claims as to whether that is the only problem that exists.