The case for explanatory knowledge
As you already know, science is about first and foremost, explanations:
Science (from the Latin word scientia, meaning "knowledge") is a systematic enterprise that builds and organizes knowledge in the form of testable explanations and predictions about the universe.
That is the central theme of science. Even prediction, the supposed utilitarian "purpose" of science, isn't possible without explanations (if you think that it's possible, then you'd be wrong).
So the question is, why do we need to explain things? Why can't we just collect a bunch of data, and summarize them? Or average them out, using statistics? Or apply probabilities and say what will "probably" likely to happen, based on past data?
Well the problem with that approach is that it will depend on what kind of data that we have collected so far, which will always be limited to our local and parochial data. For example, we don't actually go all the way to Mars, perform a physical experiment there, report back the finding all the way back to Earth, and then update the "data" we have about physics.
What we do instead is that we deduce how objects will behave on Mars, by doing calculations with pen and paper, or a computer, right here on Earth, without actually ever being on Mars. And we expect that our calculations will closely match the result of an actual experiment being performed on Mars, once we get there. This is done all without ever "verifying" that it must be true by doing an experiment beforehand or collecting "data" from Mars. It is simply assumed that the known laws of physics are true.
So another question is, how is that even possible? How is it even possible that we can just come up with a theory out of seemingly thin air, which we can use and apply in which the results will closely resemble reality to an astonishing degree?
The simplest answer is that we simply have the correct explanation about how reality actually is. We happen to have the correct explanation about how the universe actually works and behaves (amazingly enough).
The only way to know how objects will behave on Mars, is to explain things. We'd have to explain why the physics of Mars is different from here on Earth. We'd have to explain why objects on Mars will behave in this way and not that way. And the correct explanation will be that of our known laws of physics, which is that it directly has to do with the mass of objects and the relationship between them. That is the causal explanation of why the physics of Mars is the way that it is.
Of course, you can just collect "data" and say that the physics of Earth is this, and the physics of Mars is that, but we don't explain why they're that way. You may endlessly collect more and more data, and keep making more and more classifications about the physics of trillions of different stars, without ever explaining them. And hence, without explaining things, it's not even possible to predict things by deducing results from certain theoretical premises.