However, there is plenty of evidence to suggest that it does exist.
The distribution of dark matter being uneven is evidence that something functioning like dark matter has to exist as MOND doesn't easily replicate that. Colliding galaxy clusters show a distribution of matter that differs from the distribution of visible matter, so the simplest explanation is that there is invisible matter that reacts gravitationally.
The simplest explanation is that there is regular matter which we fail to detect. Not weird, given the huge distances involved - there almost assuredly is mass that we didn't detect (e.g. planets).
The point where fairy matter comes in, is that we can guess how bad we could be at not detecting matter. Basically, regular matter that we didn't observe can't explain everything - we would've observed it otherwise.
Sorry, There are currently hard constraints on the lumpiness and size on that matter which basically exclude all reasonable possibilities for "normal matter we fail to detect". It's certainly possible something strange is happening though, or our statistics could be wrong.
That's not the simplest explanation as our current theories limit the amount of normal matter that can exist and galaxies behave gravitationally as though there must be a lot more than normal matter. Normal matter that is invisible to us is not a viable solution to the evidence.
> The simplest explanation is that there is regular matter which we fail to detect. Not weird, given the huge distances involved - there almost assuredly is mass that we didn't detect (e.g. planets).
What is the difference between that and dark matter?
> there almost assuredly is mass that we didn't detect (e.g. planets).
Don’t planets give off light and thus contribute to the distribution of light observed?
The GP is talking about MACHOs; MAssive Compact Halo Objects. They used to be a valid DM candidate but have been ruled out 20-30 years ago. Planets don't give off any non-negligible amount of light. However, they do consist of baryons (read: atoms) which would contribute to the baryonic accoustic oscillations in the CMB power spectrum, so that's a good probe to set upper limits on baryonic matter in general.
Primordial black holes are a somewhat related candidate; but those are mostly considered ruled out as well. They are not considered baryonic matter.
The evidence that it exists boils down to: gravitational anomalies. It's not multimodal, which is a problem because you're saying "it must exist because of x therefore it must exist" versus "it must exist because of x, which has a side effect of y, which aha also we see". That's why it's weak.
> the simplest explanation is that there is invisible matter that reacts gravitationally.
That's the whole point of the poltergeist example. If someone said: "Oh, that urn on your mantle just fell over? Must have gotten knocked over by a rogue ball of dark matter. The simplest explanation is that dark matter must exist". You would think they are insane.
You can think of dark matter as gravitational anomalies that aren't distributed uniformly. If we assume that physical laws are constant throughout the universe, then it becomes difficult to account for those anomalies without saying that there is "something" there that is influencing gravity. Obviously, we've labelled that "something" as dark matter and figured out what properties it can and can't have.
With your poltergeist example - if it was relatively common for urns on mantle to be knocked over, then it would make sense to come up with a hypothesis. If objects around the urn were affected gravitationally just before the urn falls over, then dark matter would appear to be a rational hypothesis.
The distribution of dark matter being uneven is evidence that something functioning like dark matter has to exist as MOND doesn't easily replicate that. Colliding galaxy clusters show a distribution of matter that differs from the distribution of visible matter, so the simplest explanation is that there is invisible matter that reacts gravitationally.