Trees, regardless of your probability, the average is 50% either way. So, I will ask again: is 50% good enough? If we take Savage's example: such as 100 hostage situations all over the country, would using a bomb expert, giving 250 saved on average per case, be good enough? If there were no other alternatives I would agree, but would it be worth the time to look for other alternatives as well?
Also: where do you draw the line? In a case like this, I can't say this would be a good example to show what I am asking about.
Different example: Person A wants to save person B. If person B is saved, person C lives, but if person B dies, person C has a 10% chance of living. If person A doesn't try at all, person B will have a 50/50 chance of living.
So, in this case for example: if person A cannot find a way better than 50% chance to save person B, it would be in person A's interest to not do anything at all.
Now, if we take out the knowledge of person B's 50/50 chance--->Where would Person A draw the line? I am sure if a possibility of 40% came up, it will be considered, but something like 20% doesn't seem likely.
Does this make sense?