Optimization game

There exists a idle clicker game, that involves actions and rewards. There are 100s of actions to choose from. You select an action you either get a reward or you do not. And for every action the rewards slowly get bigger but further apart separated by more and more non-reward or at times penalty actions. You only know the outcome once you perform a type of action and you know of no future actions or rewards.

visual representation of the optimization game. keep in mind all states above the present one from bottom are not known to you

How would you optimize your reward?

take some time and try out the math for yourself. yeah, anyway I’m not going to cover that here however much I want to tell you about the Multi-Armed Bandit problem, gittins indices and how it is involved in what news do the media chose to show you. I won’t tell you all that.

What I want to focus on is what you would do irrespective of the ideal solution? with enough variables there is actually no ideal solution, because you don’t even know what reward you gonna get. That is life then. or more accurately any action you take during your lifetime. Most actions if done enough times get you to a big reward. But there is no knowing to how many times. Some people might prefer to keep getting atleast some reward by doing a lot of things for few times each. but there is only a finite things one would want to do, or love to do. The only way towards success is in depth, but there is no stopping a person who is floodfilling all rows together, while focusing on one of them the most.