>>12516980When we find some inaccuracy in our measurement data which explains the anomalously high rotational rates of all galaxies.
Dark matter should really be called mass-gravitational discrepancy, because that's what it is. Galaxies simply spin too fast for how much matter they contain; it'd be like if all the planets went around the Sun as fast as they do, but the Sun were 70% less massive than it would need to be to cause that to happen (or rather, if the Sun were still 1 solar mass, but was physically smaller and less bright and had all of the features of a star 30% the mass of the actual Sun).
Another analogy; Say you have a glass box with an object inside, and you measure the attraction of a mass outside the box to a mass inside the box. Based on this attractive force you estimate the mass of the object in the box. Now you remove the object and weigh it, and you find that it only has a mass 30% as big as the one you calculated. You have either miscalculated, your basic assumptions about the content of the box are incorrect, or your understanding of gravitational attraction is incorrect. So you do the math again a million times, and you test your understanding of gravity by doing experiments a thousand different ways, and both of those processes find no errors. You're left with the options that either your theories are incorrect in some extremely subtle way you haven't thought of yet, or there's something else inside the box that you can't see but is exerting the remaining 70% of the gravitational force.
That's all dark matter is. It's the best way to explain what we see when we observe galaxies. The problem of figuring out what dark matter actually IS is that there are many things it COULD be, from heavy particles that don't interact through the other three fundamental forces, to quintillion-ton microscopic black holes formed during the first few milliseconds of the universe's beginning, to cosmic strings, to a bunch of other shit.