As I do more and more puzzles, I've gotten used to being surprised by the way the engine defends the original player's blunder. Nowadays, I've been using the training to learn both what my optimal move is, and try to figure out what the engine will do to best defend the losing position.
Is there any chance we can implement a training where a player attempts to defend the blunder? Does anyone else agree that this would be useful?
Is there any chance we can implement a training where a player attempts to defend the blunder? Does anyone else agree that this would be useful?