How does a good hypothesis look like?
Predictive power is the name of the game
What makes Einstein’s theory of relativity so amazing? Einstein formulated his theory without experimental evidence. Validation only came years later, when experimentalists tested a prediction Einstein's theory made.
A hypothesis is about generalization: it needs to do more than explain past data, it needs to be predictive.
Making such a hypothesis is risky. Because you can be proven wrong. But it is this risk that adds value.
Importantly, you make the hypothesis before the experiment that is testing the hypothesis. This does not mean that it predates all experiments. But one needs to be careful not to cherry-pick the data:
“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” ― Sir Arthur Conan Doyle, Sherlock Holmes
It is OK to start with a strong hypothesis, and it is also OK to let the data shape it. The strength of the hypothesis probably decreases over time, as you realize that the original formulation is too strong. At this point, and only at this point are you allowed to hedge and say that “X holds under Y assumptions with high probability.1” But be careful that what you say is conclusive and meaningful.
If you form a weak hypothesis such as “The sun will probably come up tomorrow morning,” then it does not predict anything new.