Enhancing Video Games Policy Based on Least-Squares Continuous Action Policy Iteration: Case Study on StarCraft Brood War and Glest RTS Games and the 8 Queens Board Game

Table 6

Three towers case.

LSCAPI
Evaluation metrics	State 1 (old)		State 2 (new)		Reward
	Game (agent)	Player (enemy)	Game (agent)	Player (enemy)
Kills	250	150	900	300	20,000
Units	200	200	850	300	7,500
Resources	1100	400	3000	800	70
Total reward					Reward = 27,570
LSPI
Evaluation metrics	State 1 (Old)		State 2 (New)		Reward
	Game (agent)	Player (enemy)	Game (agent)	Player (enemy)
Kills	300	100	700	120	18,000
Units	200	152	660	240	3,400
Resources	2500	200	4000	300	100
Total reward					Reward = 21,500