, I note with great dismay how NVa and MD turned into a blizzard fest in the first half of November then a sheet of ice from then on. (note: I continued playing the scenario past the end-point) Hmmm, how shall I say this? Having lived in central MD from '71 until '94, then Richmond from '95 until the present, I say - no way.
The problem lies here - Virginia is included, AFAIK from looking at the probabilities in the various weather region files, in the Mid-Atlantic region for weather. This just doesn't get the proper weather pattern down whatsoever. Snow in MD in November is not unheard of, neither is it in VA (but pretty rare). Even so, a smattering is only possible statistically, and a blizzard is a no-no. As well as frozen conditions for weeks on end with snow on the ground, in early November.
The issue here is how states are categorized in the various regions, which can be confusing all by themselves (we have Upland_South and Upper_South - ???).
My suggestion is if you have MD in Mid-Atlantic (which is correct) and VA there, too, you remove VA and drop it south a notch. If you have it by regions and not just states, then you leave NVa in Mid-Atlantic and central VA and the rest of VA in a deeper south region. This would give the ability to replicate better weather patterns. Because I don't want to model Virginia weather for the Mid-Atlantic and boot Pennslyvania weather in the noggin. Some of the tweaks here need to be code.
I can't speak for the other areas of the States, but I sure know VA, MD, and PA weather extremely well and would like to do some work on this whole area of the game. In particular, winter weather. For example, the Mid-Atlantic file has a 5% chance of Blizzard in October (swamp, marsh and other - read: clear). That's just too off for me.
So while there are tweaks that can and should be made, until the states or regions are defined (by states), it becomes very difficult to properly tweak the regions for weather. And I think weather is a HUGE factor for war. <--- understatement of the year