Well, this is the kind of calculation that goes on all the time - it's just not explicitly spelled out. When you consider whether you should go for a walk, you are doing a probability evaluation of things like whether it will rain.Is this simply your opinion, or do you have any evidence that any responsible Confederate leader ever actually tried to make the evaluations you are describing?
Similarly when someone is making any kind of decision which is not an automatic one, they are (perhaps not even consciously) doing a probability evaluation. I don't claim that these evaluations were done rigorously, per se, but if someone makes a decision as momentous as revolting against the government then it is my presumption that they have some idea of what they think would happen without a successful rebellion, versus some idea of what they think would happen with a successful rebellion, and that they have decided to go for whichever option seems to them to be better; this includes probability of success.
It's like how when someone joins the army, they do so with pros and cons attached.
Basically I don't think we should necessarily assume that Confederate soldiers all felt certain of victory, and that instead they felt that victory was likely enough to be worth trying for (given their own opinions on what victory would mean, versus the alternatives).