The advantage of science isn't that it can proof stuff (which it can't) but that it is open for change instead of being dogmatic.
Ergo: I know x will happen because when all the variables are the same in the past, x has always happened.
The accuracy of the prediction is what changes in science, which is why it is a better system to use for predicting future events than other dogma based systems.
Ergo: Recently discovered factor y can be changed as a variable, and this accounts for x sometimes not happening the same way. We must account for this as a variable too.