I define predictable as "without randomness". If something is unpredictable, that means there is randomness involved. It might not be practical to try and make predictions in a chaotic system, but even chaotic systems follow laws.
Well I think you're one of the few who defines it that way (in a mathematical sense).
You can predict the probability of certain events occurring and given a large enough system in time or space, you can assert that an event is statistically certain to take place. This is the type of prediction I maintain could be made about the spontaneous creation of the universe. As Roundy passive aggressively pointed out, there would be no one around to make the prediction, but as a metaphysical possibility I think it stands that predicting a probable outcome, and given a system that makes the occurrence a statistical likelihood, renders it non-accidental. This might degenerate in to a semantic discussion of accidental, which is fine, I just wanted to clarify my original point.
My problem was with the fact that you need a set of laws to predict a possible set of outcomes.
You can still have a set of laws but you will never be able to predict the outcome, quantum mechanics is a prime example. I can't argue that you could give a probable outcome but that doesn't mean it's not an accident. A nuclear power station has a probability of blowing up but I would say it's an accident if it did.
Also what happens if our type of universe that came into being had an equal but remote possibility of being created as any other universes or non-existence? Is it an accident then?
What determines that probability? I would argue that in order to know that probability you would need to be independent of the universe itself, because anything you measure or interact with within the universe will influence the outcome and the probability you're trying to measure (which is shown to be the case in quantum mechanics).
I can't see how you can say the universe was created by accident or vice versa.