We do, but of course there are a lot of convenient choices, so there's no good reason for everyone to choose one. Furthermore, since (as far as we can tell) space is isotropic and homogeneous on a large scale, it's nearly always more convenient to use relative coordinates.

Say we choose a "universal" origin roughly 1,000,000 light years from earth. It causes a few practical issues if (for example) we want to talk about our solar system, which is less than 1/1000th of a light year across:

• We have to carry around more than 9 significant figures to give any meaningful coordinates
• We have to deal with floating point round off error when we make any calculations
• Any numbers we use constantly change as the point drifts relative to the sun, or as the planets move around the sun.
• Since this point is in a void, there's going to be some ambiguity in measuring it anyway. I suppose you'd have to define it relative to the rest of the observed universe, then extrapolate backward / forward in time assuming it remains in an inertial reference frame. This introduces significant uncertainty in your "origin" which have to add to whatever coordinates you measure.

Instead, we could just choose the sun (in this example) as the origin, and these issues won't exits.

We do, but of course there are a lot of convenient choices, so there's no good reason for everyone to choose one. Furthermore, since (as far as we can tell) space is isotropic and homogeneous on a large scale, it's nearly always more convenient to use relative coordinates.

Say we choose a "universal" origin roughly 1,000,000 light years from earth. It causes a few practical issues if (for example) we want to talk about our solar system, which is less than 1/1000th of a light year across:

• We have to carry around more than 9 significant digits to give any meaningful coordinates
• We have to deal with floating point round off error when we make any calculations
• Any numbers we use constantly change as the point drifts relative to the sun, or as the planets move around the sun.
• Since this point is in a void, there's going to be some ambiguity in measuring it anyway. I suppose you'd have to define it relative to the rest of the observed universe, then extrapolate backward / forward in time assuming it remains in an inertial reference frame. This introduces significant uncertainty in your origin at any time other than the present, which in turn means you have to add a large uncertainty to any coordinates you quote.

Instead, we could just choose the sun (in this example) as the origin, and these issues cease to be such a problem. Of course for other problems there will be more appropriate coordinate frames, but the point is that there's no perfect one for every occasion.

We do, but of course there are a lot of convenient choices, so there's no good reason for everyone to choose one. Furthermore, since (as far as we can tell) space is isotropic and homogeneous on a large scale, it's nearly always more convenient to use relative coordinates.

We do, but of course there are a lot of convenient choices, so there's no good reason for everyone to choose one. Furthermore, since (as far as we can tell) space is isotropic and homogeneous on a large scale, it's nearly always more convenient to use relative coordinates.

Say we choose a "universal" origin roughly 1,000,000 light years from earth. It causes a few practical issues if (for example) we want to talk about our solar system, which is less than 1/1000th of a light year across:

• We have to carry around more than 9 significant figures to give any meaningful coordinates
• We have to deal with floating point round off error when we make any calculations
• Any numbers we use constantly change as the point drifts relative to the sun, or as the planets move around the sun.
• Since this point is in a void, there's going to be some ambiguity in measuring it anyway. I suppose you'd have to define it relative to the rest of the observed universe, then extrapolate backward / forward in time assuming it remains in an inertial reference frame. This introduces significant uncertainty in your "origin" which have to add to whatever coordinates you measure.

Instead, we could just choose the sun (in this example) as the origin, and these issues won't exits.