>>13795720We can actually create a simple system to do magical work using two games running side-by-side for anonymous comparison of population size or other relevant traits.
The main requirement is just that the two game worlds have no common contents. By forcing an aesthetic difference between games, players have to select which world matches their own on a continual basis. It doesn't matter if a player regulars in both worlds, since any in-game currency would be restricted to the game it comes from. Open trade would violate the impartiality of the system and lead to biases somewhere in the balance between worlds.
Each game gives rise to a different theme of gameplay as a result of these distinct content-based aesthetics, and the inherent freedom to choose between digital worlds guarantees voluntary (and therefore honest) participation. Once the relative population sizes between worlds are close enough to be considered fair, it becomes possible to mine an equilibrium from which we can derive actual magical work. There will obviously be players who trade using foreign currencies, but this is a violation of game construct and would pose an attack on the world. Such activities are liable to break immersion, giving some players an incentive to fight against these "theme" corrupters. What's important here is not the rate of external aggression, but the subtle flow in population levels between games. It is critical to the system for population levels to be balanced in advance of measuring fairness.
A proper construction lets us show that the magical work is provably agnostic; it is neither a function of the player population nor the content in either game, but is a function of the relation between worlds, requiring distinct content. (Note the difficulty in retrofitting uniqueness for a game that wasn't designed this way.)
Our goal is to show:
>magical work can be stabilized>such work is useful beyond games>a work of sufficient complexity has scientific value 
