@ʝօhn 6IX ɮʀɛɛʐy

You can explain the whole of reality without putting God into the equation; but having done so does not mean you have accurately represented reality.

'
I'm confused - when is it that we MUST "put God into the equation?" I'm not sure I actually saw that covered.

Science gets things wrong sometimes. Still no gods are needed, just more knowledge.

Of course, one interesting question that arises in the interest of rigour from all this, not that the OP would probably understand, is this:

How do we distinguish between entities that are merely beyond our current observational reach, from entities that genuinely don't exist?

Wind the clock back to the days of Newton, and no one in his day would have been able to observe electrons. We had to wait another 200 to 250 years before the first physical experiments determining the properties of electrons were conducted. Consequently, proposing the concept of 'electron' in Newton's lifetime (or, for that matter, the atomic theory from which the concept of electrons arose), may well have been met with the sort of vituperative response with which Samuel Johnson responded to the output of Bishop Berkeley. At least, initially.

Likewise, there was a 40 year gap between the first publication of the postulates about the Higgs Boson, and the emergence of experimental data consonant with those postulates.

Now, this tells me that we need to tread somewhat carefully, with respect to negative observational results. I find it difficult to see how one can distinguish a priori between absence arising from lack of observational capability, and absence arising from outright non-existence. At the start, the two appear indistinguishable, and at bottom, the only way to tell them apart, is to acquire through diligent effort the observational capability allowing us to classify the former appropriately.

But, before people start becoming alarmed at the above, let me be explicit, and state that this in no way justifies the insertion of blindly asserted entities, as a failed attempt to fill observational gaps with magic. Instead, we need to be sure that we recognise this issue, before trying to tackle duplicitous apologetics on its own ground. Instead, we need to remind ourselves that in the world of empirical sciences, the first question that is asked, the moment a new entity is postulated to be required in order to explain the available data, is "can we devise an observational test for this new, postulated entity?" An entity that is not observable is indistinguishable from a non-existent entity by definition, and consequently, non-observables need not concern us as active agents in the interactions we observe. An entity that is exerting a detectable effect upon interactions, is by definition observable via that route, and the quest then becomes one of moving from indirect to direct observation.

That's the whole point. If your entity does not admit of reliable and repeatable observation, what's the point of introducing it? If a physicist can't work with it, none.

Oh, and that's one essential point I omitted from the above, in the rush to prepare some food. Namely, that when scientists decide it's time to postulate the existence of a new entity, that decision is driven by observational data telling them that said new entity is required. Without that data and the conclusion it points to, they don't bother.

Item number one: if a given model not only replicates all currently observed entities and phenomena, but also predicts new, as yet to be observed entities and phenomena, which are subsequently found to exist in accord with the description of the model, in what way is this a failure, exactly?

Just because the model doesn't include a merely asserted fantastic entity, and manages to exhibit superlative explanatory and predictive power without one, doesn't mean in the slightest that the model is purportedly "deficient". The deficiency lies instead, in the assertion that said fantastic entity is purportedly "necessary", when no such necessity is manifestly present.

Item number two: if a model achieves the above, whilst only calling upon entities and interactions that demonstrably exist, without any inserted ad hoc assumptions, then the model is superior to one that does require ad hocassertions. The first model is completely self-contained and sufficent; the second model isn't by definition.

Item number three: I'm pretty sure quite a few physicists are well placed to distinguish between "observational artefacts" and genuine phenomena.