Effectively, pratically impossible. 2**50 is a very very very very large number.
Any sequence has the same probability. Remember, you agreed to this! Therefore any sequence is practically impossible, according to you ...
Yes, but "effectively, pratically impossible" not impossible. Every time you flip a coin 50 times, the odds of the particular resulting sequence resulting is the same as any other; one chance in 2**50, regardless of what that sequence is. So, you "beat the odds" so to speak, every time you do it.
No matter what happens in a series of coin tosses, the odds are the same. We (very wisely) suspect fifty heads in a row to indicate a rigged coin.
But the coin toss series is irrelevant to predicting what is likely or unlikely in a biological system, because no series is predicted. If we think of mutations as coin tosses, each toss survives to be replicated or it does not. There is no predicted direction of the series and no required outcome.
If we look at the evolution of a complex structure, such as an eye, there is no prediction or requirement that eyes evolve. In fact, 99.999999 percent of all living things do not have eyes and get along just fine without them.
More importantly, the evolution of eyes is not at all like requiring all heads (or any SPECIFIC series, for that matter). There are many different degrees of eyeness, from photropisms in plants, through photosensivity in single celled organisms, through lensless eyes, compound eyes, and eyes that are sharper than human eyes.
Even among humans there are vast differences in the effectiveness of eyes, based on genetic differences. There are many kinds and degrees of color blindness, for example.
So calculation of the probability of evolution (after the fact) makes no sense. It ignores the simple fact that prior to a change or modification or "improvement", there is no prediction that such and event will occur.
I'm going to toss in one more monkey wrench. Some Freepers have posted assertions that bacteria respond to environmental stress -- say the presence of an antibiotic -- by mutating to survive. This implies that mutation is an "intelligent" behavior rather than a fortutious event.
I'm wondering whether ID proponents have proposed or conducted any research on this phenomenon, and whether ID makes any testable predictions that would differentiate it from mainstream biology. I'm not aware of any useful research being conducted outside mainstream biology.