How we determine facts.

[Jenny A]

<snip>
For example, suppose the bag has 3 marbles in it. W stands for a white marble. B stands for a non-white(black) marble. If 3 marbles are in the bag the possible starting conditions are:
WWW (3 white marbles)
WWB (2 white marbles and black marble)
WBB (1 white marbles and 2 black marbles)
BBB (3 black marbles)

Before we start drawing marbles the probability that all the marbles are white is .25. Now suppose we draw a white marble. Now only two marbles are left in the bag and we've eliminated one possible initial starting condition. The remaining possible starting conditions are:
WWW
WWB
WBB.

Hang on a minute.
I'm going either to be pedantic or dim.....or dimly pedantic.
But you have to specify something about the initial conditions for the marbles, for example, that the probability of any single marble of being white is 0.5 and black is the other 0.5.

No, it's an attempt to explain that we never know anything for sure unless we've seen it. It's disingenuous though because it suggests we make our scientific conclusions about the world based upon samples sizes as small as one or two. But you're quite right that assuming from the get go the the chances of black or white is 50/50 is unwarranted.