LIBERT3WORKS - General Troop Rebalance

[riknap]

So earlier (read: bloody 2 months and 2 days ago), I mentioned that I was planning to utilize the special troops system to its limit for LI by converting all the special cultural troop designations Shaytana made into unique special troops per culture-cluster that use them.

In the process of doing so, I realized that if I were to be thorough with keeping them balanced, I'll first need to start with the very basics - namely balancing the very "base"/standard soldiers first, since special troops are essentially elite versions of said standard soldiers.

And so, I decided first to play some TW games for a while to sort of reimagine how different troop types would function in battle and how to rebalance their stats to fit the Lux Invicta setting. Then, I needed to confirm just how combat damage works, and came to the following thought process:

first, troop deaths are determined by [Total Attack Points Received / Total Defence Points Available]*[Troop Kill Factor]. These determine how many troops you lose per day. vanilla has it set at 0.015, and I decided to set it a 2/3, ie. 0.01 (well, technically it's been freely floating whenever I randomly tweaked it, but I thought I'd use this value as the standard now).

second, morale loss is determined by [Troop Morale Points Lost + Morale Loss Factor]. As each soldier has a base morale point, each troop death reduces the total morale points pool (hence morale drops faster) as well as incurs a morale loss factor too (even more morale loss). This made morale a sort of "HP" system that determined just how long a battle would last. Since it has been agreed upon previously that Lux Invicta battles are all about small scale drawn out conflicts, I've decided to both increase the base morale of all troops as well as reducing the morale loss factor to 3 (vanilla had it pegged at 6).

As such, below is the matrix I've decided on so far:

notes:
attack and defence = skirmish/melee/pursue values
T = "Total Worth" = (Attack + Defence)*Morale

Light Infantry A10, D10, T60
cost = 1
morale = 3
attack = 4/3/3
defence = 3/3/4
notes: I've tweaked LI a bit so that they function more as all-around-skirmishers rather than simply junk troops. They're still the least valuable troop per person, but at least they're less trash now.


Heavy Infantry - A11, D13, T120
cost = 3 2
morale = 5
attack = 1/8/2
defence = 6/4/3
notes: A slight few tweaks from vanilla, heavy infantry are now thrice twice as expensive as light infantry, but are slightly stronger in skirmish/melee as well as a bit more defence. The price difference is to account for better equipment maintenance and whatnot relative to light infantry.

Pikemen - A8, D16, T120
cost = 3 2
morale = 5
attack = 0.1/6.9/1
defence = 6/8/2
notes: Pikemen-type troops had a bit of rebalance to them, having equal morale with heavy infantry as well as also thrice twice as expensive as light infantry. Their melee attack is relatively slighly weaker than heavy infantry, but they have twice the defensive capabilities in melee of heavy infantry (a vanilla ratio anyway) in exchange for weaker performance in pursuit. Also, the decimal-value skirmish is mostly a vanilla safeguard I retained to keep them marginally useful in siege assaults (which apparently use skirmish values).

Light Cavalry - A21.5, D16, T150
cost = 4 2.5
morale = 4
attack = 5/4.5/12
defence = 6/4/6
notes: Light Cavalry had been rebalanced by a fair amount. They have strengthened a significantly strengthened skirmish offensive and boosted pursuit value and thus more valuable as a combat unit, but are comparably more expensive as well (quadruple two and a half times that of light infantry).

Heavy Cavalry A20, D18, T300
cost = 6 5
morale = 8
attack = 0.5/10/9
defence = 6/7/5
notes: Heavy Cavalry underwent a fair amount of rebalance as well. I've reduced their base morale relative to vanilla, as well as slightly boosting their skirmish/pursuit offensive in exchange for reducing their skirmish/melee defences. They're also more expensive as well, now six five times as expensive as a light infantryman, making them less overpowered as it used to be in vanilla (especially since heavy cavalry are fairly numerous in lux invicta relative to vanilla too).

Archers A13, D7, T60
cost = 1
morale = 3
attack = 8/3/2
defence = 1/3/3
notes: Archers also were rebalanced by much as well. Aside from having the same base morale of light infantry (3), they now have a stronger skirmish value (twice of light infantry) and a bit more melee in exchange for reduced skirmish defences. This essentially makes them a glass-cannon variant of light infantry then, but significantly more stronger than the vanilla version.

To summarize their relative "worth" with each other:
1 light infantry, archer = 2 heavy infantry, pikeman = 2.5 light cavalry = 5 heavy cavalry.
Horse Archers, Camels, Elephants, and other vanilla special troops are not yet part of the balance yet, and I'll be making them up as I go ahead with making the special troops matrix and find relevant cultures that would use them (and tweaking accordingly). But a tentative strength ratio I have is 1 LI = 3 HA = 3 Cm, 10 El.


I'll be implementing these in the SVN shortly, so tell me how they feel.

edit:
as I noted, I'll be instead just using the net relative "worth" (computed by adding their offensive/defensive points and multiplying by their morale, then taking their ratios) as the maintenance values, so I decided to go with 1/1/2/2/3/5 for AR/LI/HI/PI/LC/HC, and it seems balanced enough... so far. (the previous cost ratios lay havoc to economies as DarkReborn noted)

 

[DarkReborn]

I agree with making Light Infantry more important: many cultures have LI as part of their troop composition and this will do them good. Otherwise everything else looks OK in paper. Problem with making HC as expensive is that economies might resent, while this might make sense, remember that you can't choose which levy type to raise (that would be a cool free patch feature though)... and as you said a lot of cultural compositions have HC as their centerpiece (at least 3 from the top of my head).

 

[riknap]

aye, I'm well aware I need to fine tune the economic component a bit (namely making HI/PI cost 2 instead of 3, and similar reductions), but I'm really having trouble getting LI to run more than an hour without tripping in its face.

That said, I'm experimenting with making a stand-alone module that only has LIBERT3-related components in it - ie. testing out LIBERT3 mechanics in vanilla. That just might help me balance-test without losing the mood to due to constant need to restart the game :laugh: (So far, my current build though causes some feudal realms to become republics since the capital barony becomes a city, and I have no idea why :rofl: )
on an unrelated side note, I've currently indiscriminately copied all events and decisions (read: I got lazy), and so far aside from faulty events triggering (because I didn't properly copy relevant modifiers heh), there isn't any signficant lag, which makes me suspect there are other things at work that's causing the general lag and instability. I mean, aside from the integrated BLC ofc eating memory ofc :rofl:

 

[DarkReborn]

The BLC RAM usage problem should be less gregarious come 2.1.5, since apparently the game doesn't preload all the graphics now (in the beta I mean) and instead loads on demand.

And yeah, I very much know that the problem for the crashes is somewhere in the complex web of integrated mods and mechanics (80% of my leisure time is actually used around looking for the damn issues these days). One of the causes I've noticed is that LI has many more CBs and they're much more easily available than vanilla's, thus more wars which with the large number of independents in LI means more lag.

 

[riknap]

that might be part of it. more wars mean more CPU strain means more lag.

on a different note, I'm redacting my statement about events not causing lag. They still are, just less noticeably than before (which means they're not the sole/primary contributor anyway).
and yeah, I can feel the pain of it. Right now, I'm already having a hard time getting my experimental setup to not crash everytime a SELIN/LI-specific event/thingy is accidentally triggered (I know I can just handpick files to add, but I also want to test and see how the other events are potentially causing lag as well anyway - or CTD's for that matter).

that said, could I ask where the tings-and-related events are located? I forgot the event file name, and we don't exactly have an index :laugh:
edit: I found them. clancouncil.txt. 'twould have been nice if there were a prefix for the filename :laugh:


on-topic: It's as you say, that costs are a bit too expensive right now. Playing as the Magyars, the ruler's levy alone (nevermind their upkeep-free adventurer companies) would bankrupt you in a span of a few months (I was paying around 154 gold yearly due to the heavy LC/HI emphasis, while only had 33 gold from the probably-not-random prebuilt LI-buildings in my demesne). Then again, the LI economy state *might* be different in LI proper, but it's still not a good indicator.

I'll probably use the relative "worth" as the upkeep instead (1/2/2.5/5 for LI,AR/HI,PI/LC/HC) and see where that takes us :laugh:

 

[ringhloth]

Something I've noticed is that HC favoring tactics are very rare, even for someone like the ERE. Instead, I see my generals using the Advance tactic way too often, crippling the key parts of my army. Rebalancing the troops needs a lot of rebalancing of tactics, too.

 

[DarkReborn]

The cultural tactics aren't set to trigger too often ( I don't know why though). Anyway, Riknap mentioned rebalancing the tactics as part of his special units overhaul, so he'll eventually do it.

 

[riknap]

more accurately, I've been planning (for close to/more than a year now) to overhaul tactics in general actually.
once I've implemented the special units first, I'll start work on a semi-independent module that creates tactics from scratch. you can see the rough details in the red ABCDE link in my signature actually - I've thought about them long ago, but never got around to coding them yet :laugh:

 

[ringhloth]

Right now, the biggest balance problem is tactics, undoubtedly. The advance tactic is incredibly common, and makes HC useless. I've got a save of a 8500 vs. 8500 battle, with two differences between the armies: in one, LI, HI, and archers are about half that of the other army, the rest going into LC, HC, and HA (Specifically, 1400LI, 500A, 2700HI, 1800LC, 300HA, 1500HC vs. 2500LI, 700A, 4700HI, 200LC, 50HA, 500HC). With a lot of regularity, the cavalry heavy army take full casualties, while the infantry heavy army takes barely 2000. It generally goes even losses in the skirmish phase, and then, in the melee phase, it starts out mostly even, with the infantry army doing very slightly better when skirmish tactics are still being used, and then one flank of the cavalry army uses the Advance tactic (it's almost always my 30 martial monarch, too), and just completely collapses. 100, 150 casualties a day in one flank is absolutely devastating. Then, of course, when one flank falls, the rest go with it.

 

[riknap]

that certainly is a useful anecdote

( on a random note, one thing that annoys is me is that absolutely remember reducing the number of days each tactic has to 4 days at most in previous LI versions, yet it got accidentally overwritten sometime during the post-RoI update. )

still, it is indeed a problem that there are only a limited number of tactics available, and the relative homogeneity of vanilla armies makes vanilla-style tactics pretty much a dice roll you have no control with. there are two ways to solve this really: a) making a laundry list of modifier checks that take your composition heavily into account to make inappropriate tactics far less likely to trigger, or b) making a laundry list of independent tactics whose trigger conditions are highly varied to account for a large spectrum of potential troop compositions, both for the same purpose - so an army with 18% heavy cavalry using your example is more likely to trigger tactics that take advantage of that HC composition rather than tactics that ignore it.


that said, since I currently hit a brick wall with regards to special troops (why the hell are heavy infantry -based special troops showing up as horse archers, for example?), I'm working on my ABCDE project instead from here on.