Different rewards for O and X?

[drozzy]

Sorry if this is a dumb question but here you set different rewards based on whether X wins or O wins:
https://github.com/haje01/gym-tictactoe/blob/master/gym_tictactoe/env.py#L8

Why is that?

If X's reward for winning is -1, wouldn't that encourage an agent that is playing for X to always loose?

Thanks.

[haje01]

Please consider the code below:

gym-tictactoe/examples/td_agent.py

Lines 111 to 115 in 84e22fc

	# select most right action for 'O' or 'X'
	if self.mark == 'O':
	indices = best_val_indices(ava_values, max)
	else:
	indices = best_val_indices(ava_values, min)

[drozzy]

Ah, I see.
But if I wanted to apply a ready-made algorithm, I suppose I should make the rewards positive, and just assign the "-ve" reward to the looser.