stockwellpete wrote: Mon Apr 29, 2024 5:38 pm
One notion that has come to me is to introduce the idea of "Army Reorganisation". I am reading John Keegan's book on the ACW and, according to his narrative anyway, such reorganisations were quite commonplace. So, linking such army reorganisations to the "Corps Organisation" research option, maybe a player could merge existing Brigades and Divisions into new Corps once a certain level of Corps research has been achieved (maybe level 2)? Maybe it would take 2-3 turns. This would reduce the clutter somewhat. Whether it would be enough to solve the problem that some of us are concerned about, I am not sure. And I don't know either whether the game engine could actually do this.
I think players would mostly avail themselves of this option in the Virginia area, whereas further west you would probably not always want to concentrate your forces in this way as there are relatively few units operating in a much bigger geographical area.
Any thoughts on this?
It's funny that you mention this, because this was actually the original idea I had for how this aspect of the game would play out. Corps Organisation was one of the first techs I coded into the game, originally planned to be "+4 Corps build limit, -4 Division build limit" and brigades wouldn't be buildable at all - the idea being that the smaller, weaker units would get killed off over time and not be able to be replaced, which would keep total unit counts to about where they normally are in 1862 in the final game. Setting aside the fact that some of this either wasn't possible in the engine then, or still isn't (remembering that I'm talking about a game very early in development at this point - if the engine needs to be changed, it was much more a possibility at that point than it would be now), there's quite a few pitfalls with this approach.
Probably the most glaring issue I found comes from the idea of killing off the weaker units. In practice, this just doesn't work, at least not to the extent it would need to to build a game mechanic around.
The game measures two stats for counting losses, MPPs and units killed. In a game between two evenly matched opponents, or between the AI playing against itself, the total losses under both measures will tend to be about equal in the absence of complicating factors (eg units starting understrength). No problems there. But in a game where the opponents are not evenly matched, the losses as a ratio of units killed will tend to increase at a far greater rate than the MPP losses will. It's not too hard for an expert player against the base AI (with no experience boni &c) to rack up a 10:1 kill count, but still only achieve perhaps 2:1 or even less in MPP advantage.
What this usually translates to is the AI damaging some units, so our expert player has to go back and reinforce them, which still slows his offensive down, while the AI's units get killed - the player then can advance into the gap left behind, until those units get rebuilt.
However, if we now suppose that the AI can't rebuild those units, that gap is going to widen a lot more quickly. If both sides have 20 Corps and 20 (old) Divisions, and the player's Divisions get repaired while the AI's get killed, very quickly you're looking at a situation where the player has 20 Corps and 18 Divisions, while the AI has just the 20 Corps (of which half are still in the build queue). Even a small difference in skill is going to skew the result of the game much more quickly, and once a player has the advantage, it becomes a lot harder to reverse that advantage later on. I also note that while I've used "AI" in this example, the game goes for MP games too.
I also found very early on that the SC engine has something of a minimum density of units, below which the game just doesn't function very well on an open map. I say "open map" because of places like the Ozarks, or still-neutral Kentucky, which are difficult to move through and thus effectively restrict the number of units needed in an area to it when said area is not being made a priority - 1846 works as a campaign despite having fewer units on the map than even early-game ACW, because there are a finite number of places where those units can go - primarily, the mountain passes - and that's only after I lock off significant parts of the map as "Indian country" and impassable.
That minimum density comes into play right from the beginning in the East. I've tried several versions of the game with fewer units on the map (considering just 1861 for the moment), what happens is either neither side is strong enough to do anything (which is boring), or one side has an advantage and can encircle enemy units very easily (obviously not ideal either). Historically that sort of thing didn't matter, because armies could react to each others' movements in real time - but that's not possible in a turn based game (yes, I could have made ACW as a RTS, but at that point we're not talking about SC at all, but rather a totally different game).
Why the map appears to crowd up in the East is very simple from here: if you're playing the game right, you will have more units in the late game than the early game (industrial tech and captured towns giving more MPPs, and hopefully not losing more units than you can afford to rebuild), and because the East is the most important part of the map, units are only going to be moved away from it on a very rare occasion.
Whether a crowded East is necessarily a bad thing, I definitely understand the argument there, but I'm not exactly convinced by it either. Except for the weeks leading up to Gettysburg, the AOTP and AONV basically paced around the same few dozen square kms trying to force a favourable battle and avoid being caught in an unfavourable one, until the Wilderness battles in 1864. The actual territory "occupied" by each side didn't change all that much (similar to WWI), and any time either army moved more than a couple hexes' distance (a hex being ~15km), the other would respond with a major battle usually the result. If you think about how often units get swapped around, destroyed and replaced in the East, I think it's fair to say there's a lot of movement back and forth over that small area, with lots of battles of varying importance being the result.
stockwellpete wrote: Tue Apr 30, 2024 7:34 am
There is a lot more to say about all this and my ideas are very provisional, but it would be interesting to know from the developers if the merging of units is a possibility with the current engine. If it is, maybe someone who can mod might like to have a look at the idea in the near future?
Stacking is something else we put a lot of thought into. The short version of that story is that we never came up with a system that we thought would work (keep in mind, suppose you stack two units, how do you combine their stats in the stacked unit? What if their base stats are different? How do upgrades factor into it? &c &c), and when we built the game without it, we found it wasn't really needed either.
Merging units is not currently possible with the engine and probably won't ever be. Unit deletions are possible via scripts now, so with some elaborate code it might be possible to say have a hex where you place a unit, have it automatically delete and be replaced with another (better) unit.
stockwellpete wrote: Tue Apr 30, 2024 7:34 am
I am not sure of the relative sizes of "Regiments" - "Brigades" - "Divisions" - "Corps" in the game at the moment. Maybe 3 divisions makes up a Corps, does it? Not sure how many brigades make up a division.
During the war, this wasn't fixed. Usually 2 or 3 of one level made up the next, but on some occasions that number could be as high as 5. The game is vague about this intentionally as a result, I've just used what I felt played out the best in each given circumstance.
- BNC
