I suggested that "something" within science can be a hypotheses without evidence.
A hypothesis can be based on evidence, a lucky guess, throwing a dart at board, etc.. They key that makes it a hypothesis is explanatory capability with falsifiability.