No it's not, defining it as: If it's possible for a player to force a win or draw, do it. Just because it's computationally unfeasible to determine, doesn't mean it isn't theoretically possible to decide.

That's the point, while chess is deterministic, it's not computationally feasible to play the game perfectly.

Perfect play as is currently used in the chess world means you don't make moves that result in you losing your advantage or giving the other player an advantage, given current computational resources to evaluate a position. If both sides play perfectly then the game results in a draw.

I think most of the games that are possible (ie games in game space) in chess are draws but this is just my sense of it.