The improved prompt is now more structured and provides clearer guidance to the model. It defines the evaluation criteria, specifies how to handle ties, and includes a chain-of-thought trigger for more systematic reasoning.

Would you like to proceed with any further modifications or tests on this prompt?