I think militia are ideal for defense. I haven't tried it, but I think militia would be effective in corps with another unit with an engineer brigade. Those divisions would have a +40 dug in bonus in forty days.
Porkman... can you analyze engineering brigades?
I don't know the right equations. Engineers add to defensiveness and toughness but the BWA equation only uses defensiveness and soft attack. (It uses hard attack to but because all the units in the Chinese theater are at 100% softness, it doesn't matter.)
Theokrat who made the equations for BWA, also made an equation to measure the effectiveness of defense and toughness. Unfortunately, it varies by four variables ... i think..., (# of attackers, # of attackers, defense vs soft attack, toughness). One of the things he figured out was that the more divisions you had the more likely it was that two hits would fall on one enemy division, exceeding it's defense value and causing more casualties. That was all for vanilla HOI2, with AOD the square root term makes these equations so hideously complex that I can't begin to figure out how to calculate it. The relevant thread is
here and maybe you'll have more luck than I will.
The basic idea is that the soft attack, (firepower) of a unit rarely exceeds the defense (or toughness) of the opposition on a 1 to 1 basis. However, hits are assigned randomly in combat so it's possible for a division to be targeted by more than 1 division from the opposite side. In those cases, the defensiveness is exceeded and the division takes more damage than the stats would indicate. Using those numbers, theokrat figured out Engineers were not worth it as the small increase in defensiveness was canceled out by the chance of double targeting. (6 infantry with ENG brigades will get targeted more than 8 unbrigaded ones.)
This was all done using the linear model for firepower. With the square model, I think that Engineers become more effective, but I don't know how to make the equation.
Could you not make the same argument about 1918 INF compared to 1936 INF, that they are worthless or overpriced.
So IMO MIL should probably have the same amount of models that INF does. But IMO MIL should never be a better option than INF, the only advantage MIL should have is short production time (and hopefully short re-tooling time too)
You could indeed make that argument. But it's irrelevant. A player will never have to choose between building 1918 infantry vs. building 1936 infantry. They do have to choose between building 1936 militia and building 1918, 1936, 1939, 1941 infantry.
I think removing retooling entirely from militia would be a good bet as well as adding new models (1939, 1941, and 1945.) Another thing I would like to see is a reduced TC load, dropping it to half of that of infantry would make militia a bit more cost effective. (currently China will exceed it's TC at around 80+ divisions). I think that Militia 1943 is about where militia should be in terms of cost effectiveness, once the retooling time is removed.
