Adaptive dynamics with vector-valued strategies