I didn't forget to take a class, I had statistics for my Bachelor's and my Master's. Statistics and statistics manipulation should be at the fore of classroom learning IMHO, as statistics (and statistic cherry picking) are vital to better understanding law, forensics, intelligence-led policing, politics and propaganda, media advertising, biology, phylogeny, etc., etc.
Here's what you're laying out before me:
1) I don't understand statistics
2) I propose raising the odds to the 32nd power for 32 separate iterations of the same evolutionary trait
3) You feel important factors modify the odds, and rightly so, e.g., easy to evolve something over and again (that is circular reasoning on your part, since you accept mechanistic and not theistic evolution as axiomatic, but okay) and I'll add to your assumption natural selection, enhanced survivability (despite claims the thing becomes vestigial!), etc.
I'm willing to evaluate the odds, I learned something new this week--no skeptic at this forum will say anything but "incalcuable odds" or "your odds are creation bs" or "you are stacking the odds, man"!
If I take your factors into account, and reduce the numbers vastly, from X^32 to X^16, do you see we are dealing with laws of large numbers? If I accept odds of 10^50 as technically null, impossible, do you understand I see X^32 as possible, but highly unlikely? That is, I'm open-minded, especially as I better understand evolution, since all learning should tend toward open-minded behavior?
And, as has been explained to you several times, this isn't even close to the correct calculation. Mutation and selection *greatly* reduces the number of trials required to find (close to) optimal solutions. So the exponentials you use are *completely* inappropriate. I gave a specific scenario where I compared the one-shot, everything must fit perfectly, otherwise discard and start over', which is the basis of your calculation, and the evolutionary 'build up in pieces', adjusting what you already have to fit what promotes survival' scenario. The rates of finding even long sequences are dramatically different. The point is the the evolutionary scenarios tend to produce 'exponents' not of 32 or 16, but of 2 or 3.
We can, and many do, model evolutionary scenarios as methods for solving complicated problems.
So, let's be clear what our starting point is and what you want to be the ending point.
We know of species that only have *two* proteins. One is a serine protease, a digestive enzyme, and the other is a protein that when split increases its solubility. There is nothing magical about either: all that is required of the protease is that it divide proteins. All that is required of the other is that it undergo a conformation change upon cleavage that reduces its solubility. Both general types are in abundance in ALL organisms. And yes, their 'purpose' may have NOTHING to do with blood clotting in the ancestors.
This is another aspect of evolution that makes calculations very difficult to impossible: it is *common* for proteins and structures that evolved for one 'purpose' to be adapted to another just because they are there and work well enough. What this means, in this scenario, is that the precursors to blood clotting proteins may have nothing to do with the blood at all.
And, for example, what happens in lobsters is that one protein is inside of cells and the other is generally outside of them. When the lobster is injured, cells are split open and the two proteins come together, the protease acts on the soluble protein and makes it insoluble, which clots the area.
Because of the way evolution works, it is NOT required that all 32 proteins be produced simultaneously. ALL that is required initially is a two protein system that works well enough in an animal with low blood pressure. After *something* works, other systems can be co-opted to help out and make for a more complicated system that amplifies the signal, works in animals with higher blood pressure, etc.