Fixed. That was easy. Still working on the armor stuff.
Ok, well tried the 1 Tk = 2 Tks thing and it didn’t work too well. Nerfed the assault value of the Armored units too much; so, back to the drawing board.
Issue is that, originally, game Load Costs for AFV/Vehicles was their combat weight. This is problematic in terms of unit troop count for places like Atolls. Your average M4 Tank Bn will have about 60 AFVs, so troop count is about 1380 against an actual of 750. But the unit also needs Sup, which adds another 810 troops, so the unit actually counts out as 2190, which is 3x actuals. Makes it difficult to put a Combat Regt, a Tk Bn, and an Arty Bn onto an Atoll, without over-stacking.
There is no real reason to set AFV LCs to their combat weight. LC is a game abstraction that has implications for both loading and troop count – occasionally, mutually exclusive. Ship load factors automatically multiply all AFV and Vehicle LCs by 3 to calculate their “effective” load costs, so a useful algorithm can be devised that gives a more reasonable troop count, but results in an “effective” load cost equal to or greater than the device’s combat weight. Weight is a consideration, of course, but in loading ships, footprint is of concern (a GMC truck takes up as much hold space as a Sherman, and more than a Stuart).
So .. we are looking at a mix of footprint and crew size to give things more of a central tendency. Looking at Sqrt(Wt) + crew size to model both the footprint and troop count parameters. So an M4 Sherman (original LC of 23) goes to 10, which gives a much better troop count, but still “effectively” loads at 30. An M3 Stuart (original LC of 11) goes to 8, which gives a much better troop count, but still “effectively” loads at 24.
Given all this, the average M4 Tk Bn, will now have troop count of 680, plus roughly 700 Sup, for about half the original, and “reasonably” close to the actuals. It will load at about 4150 cargo total, so will fit on one big cargo ship – as it did.
So far, looks reasonable.