When was the first game-theoretically optimal game of chess played, where neither player made moves which altered the theoretical game value of the opening position? We do not require players to put up the strongest resistance (for example, if a position is lost). Has the first perfect game already been played or has it yet to be played?
Of course, barring tremendous improvements in computing, we can never prove whether any given game is a perfect game, though we can disprove some games.
Consider a contest (of hubris), nominating games (the older the better) believed to be perfect games, that stand up to all current human and computer analysis, and that are predicted to stand up to all future analysis. Neither player makes a half-point (or full-point) mistake.
I suspect the game value of the opening position is a draw, so unfortunately we are looking for old insipid draws where both players play very safe moves: "grandmaster draws".
Comically, if we allow game results by mutual agreement, then one of these zero-move games must be the perfect game:
1. Draw agreed 1/2-1/2
1. Black resigns 1-0
1. White resigns 0-1
In a similar vein, if the game value of the opening position is decisive (say, white can force a win), then a game in which black immediately plays very bad moves and quickly gets checkmated may be a perfect game: we do not require black put up the best resistance from the losing opening position.