Actually I wasn't thinking about "physically impossible things happening very rarely" (QM) but only about regular physics vs probability of things happening.
If you consider quantum mechanics you are right in an infinite universe there could be areas in which physics just happens to work very differently, people there would formulate very different physical laws (if people could evolve, or spontaneously appear). So if the universe is infinite, it doesn't make much sense to talk about laws of physics. Still there need to be some fundamental rules that never change, for example the fact that something exists. You can't have areas of the universe in which the universe itself does note exist (I think). Frankly I don't believe the universe is infinite, occam's razor says it's just very big. Last month there was a report about someone finding a pattern in galaxies that would suggest the universe is much smaller than we thought but light "wraps around" making it appear infinite... the theory was discarded very soon after more experiments were carried out, but it reminded me of that star trek episode.. "state the nature of the universe" - "the universe is a hollow sphere 12 km in diameter" ... or something. Infinity is just our perception of things very big... something that originates from nothingness and expands has very little chances of becoming infinite in finite time. mirai++ > > I think two things are being confused. First, the laws of physics, second, > > the laws of probability. A gas particle follow physical rules (movement, > > bumping, thermal vibrations) and lots of gas particles together follow > > probability rules (low probability of people suffocating in rooms). > > The problem is that all the laws of physics have been found observing the > world around us in an experimental way. But all the outcomes of an > experiment are probabilistc and we know the low of physics only with a > certain error. So the paradox in the laws of probability is a paradox in > laws of physics too. The whole physics is probabilistic. >

