It's not dumb it's just a different take on reality (bayesian thinking). I have a model and accordingly a belief belonging to it: "given the data, how likely is it, that my current model is true?".

So maybe you're starting of believing an orbit has a spherical shape because it describes best your observations. But as you're collecting data you're noticing that an elliptic shape is more probable. Therefore you now believe the shape is elliptical.

In physics it's often different. People start of writing down a law from first principles and see if it agrees with the data. E.g. Kepler writing down his three laws of planetary motion in an act of epiphany. Then he sees this fits the data well and is happy.

The question is philosophical: Do you believe there is some fundamental laws nature obeys or do you say I just take the model which is the most probable given the data.

But I agree that only very few people are belonging to the latter group and even fewer people in theoretical physics, where people are obsessed with "beauty" - e.g. believing orbits are being described by ellipses, not just some shape my data suggests me.