Hi,
you always play one full game (e.g. 100 iterations) of the iterated prisoner's dilemma against a fixed strategy, after the game finishes, the reset method is called before you play against another strategy. You can play against the same strategy multiple times, but in principle you go not know it is the same strategy, just that you play a new game.
For example, assume that one game of iterated prisoner's dilemma is 100 iterations. You play 100 iterations against strategy A, reset is called, you play 100 iterations against a strategy B, reset is called, and so on. B can be in principle the same strategy as A.
Does it make sense now?
Have a nice day,
Martin