Develop a strategy to maximize your average reward per move (equivalent to maximizing total reward over n moves). express this as a function of k, using θ-notation. in other words your maximization doesn't have to be entirely precise; you may assume that k is any convenient number that will make the math easier for your strategy, but you cannot assume that k = o(1). notice that any strategy that you come up with provides a lower bound on reward optimality. the better the strategy, the better (higher) the lower bound. it’s trivial to get ω(1) per move, so you must get ω(1).