I'm going to walk back some of what I said about tactical, operational and strategic games last night.
Tactical - Yes, there are some historical scenarios where either side could win. But there are just as many (or more) where one side neve really had a chance. Designers shouldn't try to nerf one side so they can try and get a different outcome than what really happened. What they should do, it challenge the player of the losing side to do better than what was historically done and show plausible effects if those goals are met. To me, it seems like laziness on the part of designers to understand the effects of changing the outcome of a battle could of had on the larger picture.
Operational/Strategic - I did say that it's probably more realistic for designers to setup possibilities that the historical losing side could win. I stand by that, but designers cannot get too wild with it either. As an example, Poland in 1939 had absolutely no chance to stop either Germany or Russia, or both at the same time, as what really happened.