This is why scientists try to control as many parameters as possible. When only one variable remains, an experiment can be performed again and again and the same result will be obtained every time. If different results are obtained, there may be some unseen variable, or in the case of quantum mechanics, the result may be random. If it is random, this can be repeatedly demonstrated as well.
This process of eliminating variables is so important that it is a requirement of the scientific method. Knowledge which is not gained in a manner including this process is not science, though I do not make judgment of its utility or value. It is something else - philosophy, common sense, whatever you want to call it, it is simply not scientific. I see no reason, therefore, to call creationist theory a science.