mathman said:
Is this your guess or is there some study which leads to this conclusion?
Well, as I said, without knowing how Hubble volumes like our own get started, we can't really know this for sure.
But, even in the standard model of particle physics there are spontaneous symmetry breaking events that would be different in different locations, and we expect there are likely more such events encoded in still higher-energy physics. Perhaps many more. This would be an indication that the low-energy physical laws which we experience would be a product of a peculiar series of random events in the early universe which would be different from place to place.
The second bit of suggestive evidence comes from dumb examinations of the physical constants: slight changes in most of them result in a universe which would be devoid of life. Now, these are rather dumb examinations, because without the correct theory explaining how new Hubble volumes get started, we don't have the probability distribution for these constants. But taking either a dumb flat prior or a logarithmic one seems to indicate that the region of parameter space available to life is likely to be a tiny fraction of the whole.
The final bit is a philosophical point. The second point above is basically a brute fact that must be dealt with (I'm sure I could come up with references if you really need them). This fact forces us into one of two possible situations:
1. In reality, only a minuscule fraction of the entire apparent parameter space is actually possible. It just so happens that this tiny bit of the apparent parameter space is also be the tiny bit that is conducive to life.
2. A significant fraction of the parameter space is possible, and the parameter space is explored.
The first possibility sounds to me quite magical, and from a philosophical standpoint is almost certain to require a theory with more parameters to specify the smaller parameter space (and thus made less likely due to Occam's Razor). In general, specifying a class of objects requires fewer postulates than specifying a single member of the class.
From this I conclude that a very large, mostly-uninhabitable universe, somewhat like a larger version of what we see in our own universe with its vast expanses of uninhabitable space, is by far the most likely possibility. The apparent discovery of spontaneous symmetry breaking in high-energy physics appears to be pointing us more strongly in this direction as well.