I get the idea of confinement, and how it is impossible to separate a lone quark from a baryon: it needs more energy than creation of two more quarks, so the latter happens first, and you end up with having created a (color-neutral) meson. However, I don't see what prevents free quarks from appearing out of primordial quark-gluon plasma: Whereas quark-gluon plasma is color-neutral on average, when it cools and "quenches" into baryons, the quarks group into color triplets *randomly*. Even if a volume of cubic meter (or a cubic light year) of q-g plasma is strictly color neutral (it is possible to pair up (or is it triple-up?) all quarks into baryons with no leftovers), it is extremely unlikely quarks would manage to do that *randomly*. Imagine that all of quarks successfully combined into baryons except three quarks (one red, one green, one blue) because there is small problem: they are on the order of 100 light days apart from each other. Why? Because quarks aren't sentient, they can't "plan" how to carefully pair up to avoid such a fk-up. The cubic light year is still perfectly color neutral as a whole. However, it contains three quarks which for all practical purposes are lone quarks. What am I missing?