After reading Moscovian's post I am reminded of a conversation between Heinz Guderian and Adolf Hitler just after the Kursk debacle of 1943.
Hitler: "I don't understand, our Panzers used to run right through the enemy now they fail to even come close to matching those former successes."
Guderian: "Change your methods."
Changing your methods doesn't require a massive change in the army either, one can accomplish it with minor changes in force composition, while using a radical change in tactics. I have the fortunate luck of having numerous opponents, and it causes one to constantly think out of the box just to keep up with them
Of the group Jack and myself are probably the most unpredictable on the battlefield, yet both of us only tweek our core armies for each battle. Jack's basic battle plan is to have no plan, no kidding, he pretty much lines up his troops the same way all the time. But once the first moves go in he is off and running winging it as he goes. Don't ask me how but he wins, and no battle he fights is ever fought the same way, by him. I do go into a battle with a plan of action, and an idea of what I want to accomplish, but like Jack I like to fight a wild battle of maneuver and do everything I can to force just that situation (Even with the Ork Horde). To accomplish this I have to constantly re-think the tactical situation on the battlefield so, by default and like Jack, I have no real set tactics.
The rest of the group fall into the 'build the army with a plan' type players, and there is nothing wrong with this, as it serves well their styles of play. They often build the army around a very small group of core formations (That they will modify from time to time), then build the rest of the army to fit the battle plan they have in mind. Their core formations tactics often change very little, and the rest of the army is built to enhance or modify those core tactics. So even though they are more rigid in their thinking they accomplish the same tactical flexibility that Jack and I achieve.
Basically different methods same results at the end of the pipe.
Jaldon