The simplest path to realize the "proof of game" concept is to make a learning game. There is no need of servers. Gamers just solve the tasks. The correct answer is already known in advance (its hash is in the application' memory). The hash of the gamer's response must match the declared hash. The more difficult becomes the task in relation to the gamer's level (it increases with time), the greater becomes the probability of accepting the response of this participant as a confirmation in blockchain system. The more difficult tasks gamer solves, the faster his level grows, and accordingly decreases the probability of getting a reward (why is needed to reward a person for gaining knowledge that he already has?). On the other hand, the size of the reward should increase depending of the gamer's level. This will help to avoid problems with multiple registrations of the same users.