On the swarming behavior of Dratinis (or other rare Pokemon) in the wild

There are places in Pokemon Go called "nests" where many Pokemon of the same, often rare, type spawn. This is why there are (or were, depending on when you're reading this) many Electrabuzz in Central Park and many Bulbasaurs in a park near me.

My theory is that the pokemons move and thus appear to spawn, so there may be cycles but not fixed spawn points - people used to fixed loc spawns in mmporgs are trying to crack a spawn location that simply doesn't exist. We already know that any player present for a pokemon can get it; they aren't zero sum. If P is at loc Y, there will be P(x) copies in that cohort of gamers in which x is the # present to pick it up. Prosocial etc.

Knowing that a given P can make an unlimited number of itself for gamers who intercept it, I think it's more likely that the Ps are modeled on something similar to a celestial sphere; all those stars in constellations and dispersals that aren't constantly being RNGd (can you imagine the server load for that kind of scale) based on terrain type. I think that's a myth. I'd bet solid cash that you can get a water type in the desert. I live in Kansas. I've seen them all, and we do NOT have all the terrain types associated with the Ps. If it was truly gsm like that, the game would be boring and maybe unplayable (depending on player goals) for people in monothemic locations. I don't buy it. The game is intended to play globally. No doubt Google has the computing power to make an inelegant and clunky design work anyway, but I think the designers would be smarter than that. I think they built a sphere. As it moves, the Ps appear to move. If you count spin, axis rotation, time/velocity, etc. - your basic variables of planetary motion - it would take hundreds of years to see repetition. And they could mess with the axis angle etc. at will, meaning that cycles and patterns could be eliminated just by setting those three variables to be random, rather than millions and millions of pokemon.

Nocturnal vs diurnal Ps are definitely a thing, and that aspect of the mirror reality got wicked simplified. Sun and moon (virtual, in game reality) 12 hours apart on each side. My game goes night dark at 6. Sun doesn't set here til 9ish in the summer. That's also a pretty good giveaway about the nature of the model.

It's the ultimate illusion of infinite random number generation, using a sphere that has variable movement in all three dimensions.

I'm pretty sure I'm right. Mostly because of how tracking works. Especially once you start really watching and following tracks using the orientation beacon (green pulse box). I think it's hypothetically possible - a daemon hypothesis - that a Pokemon trainer demon unlimited by human frailty and physics could follow one Pokemon forever. 360 degrees. All around the world.

There are certainly RNG elements in the game. I've seen some weird stuff hapoen, but we have to remember that the servers are jacked right now. The only way to form stronger hypotheses about the game's design would be assurance that it's even perceptible, given the randomness of server load errata. Crashes, weird data push timing and lack thereof due to server lag, etc.

The Pokeverse we see probably isn't always behaving like it should. Which makes discerning its natural laws problematic.

And I think the bog standards (pigeons, rats) are basically just diffuse atmosphere, not stars. There's just a layer of that crud.