>>13031587The second seems far more plausible, with pokeballs being a simple matter-antimatter reactor.
Let's consider a space somewhere where the pokemon "go" while it's "inside" the pokeball. The exact placement isn't exactly important, but pokemon are affected by passing of time in the form of poison damage (until gen V where they removed it so babbys had it easier).
The process of catching a pokemon is merely scanning its molecular structure and binding it to the pokeball.
The process of recalling pokemon back to the ball is a simple matter-antimatter reaction. The antimatter corresponding to molecular structure is generated in a very small amount, causing a disintegration that releases the exact same amount of energy. Since you can only create matter-antimatter as a pair, the pokeball creates an exact same molecular structure made of matter in the aforementioned space. The total energy quota is exactly zero, because, counting what we started and ended with we come with "matter poke" on both sides, and with it just being moved, which in itself doesn't required energy usage (not counting potential energy difference and fricton that doesn't happen anyway). From a purely scientific standpoint it's viable, but actual real-life implementation would require as-of-yet impossible energy-to-matter conversion that happens on 100% efficiency and accounting for xboxhueg amount of molecules.