So then if the star appears to move against the stars so far away that they didn't appear to move, you can do a spot of geometry and work out how much the subject star has moved in "arcseconds".
A star that is exactly far enough away to "move" 2 arcseconds over this 2AU distance (or, if you prefer, one which has "moved" 1 arcsecond when we move 1AU) is one parsec away. By definition.
When discussing relative distances from earth for stars, it's therefore a useful unit (we can't just compare arcseconds because further stars move less against the background, not more).
As with so many puzzling units the key is to find out why they're used. This one was used by astronomers who didn't have a better way to determine how far away stars were.
The wikipedia article, incidentally, is much better than that old Encyclopaedia Britannica article that I read in high school. I envy kids today.
https://en.wikipedia.org/wiki/Parsec
Oh and the actual distance is kind of irrelevant and I don't even try to remember it :)