Which statistical tests to conduct

Question

I am doing a project in which I am trying to find alternative tactics that could have been used by the British navy in the Battle of Jutland. I have a stochastic simulation which runs multiple simulations of the battle and outputs data such as the number of ships sunk and the amount of hits taken by each side. I have a baseline set of data which represents the simulation run in a way that accurately depicts the historical outcome of the battle and several other simulations with changes made to represent different tactical decisions that could have been made in the battle. I am interested to find which of these alternative tactics produces the best result for the British Navy. Is it appropriate to just conduct tests comparing the means of the data to the baseline sample. If the tests are significant can I then simply draw conclusions based on which sample has the greatest effect on the mean or do I need to conduct further tests to determine this?