Supply System Brainstorm

[Diomedes_]

So since I posted a thread a couple of weeks ago about my gripes with late-game combat and then reading Axatin's congress proposal, I've been thinking about the supply system a bit more. I have some of my own ideas but would also like to hear what others are thinking as well. I didn't make a congress proposal because I don't think my ideas are fully fleshed out yet, and I want to see how Axatin's proposal affects gameplay if it is passed, because I think there's a good chance it accomplishes a lot of what I personally want to change.

My biggest idea, and the one that I think might be the most controversial, is that I think a Military Supply Cap, which only penalizes a player that goes above it and scales with both city number and population, is not a very good system. I see two main problems with relying mainly on a Military Supply Cap system: 1) penalties incurred for army size are not linear and there are few economic incentives for keeping a small army. This makes player decision-making regarding their military pretty one-dimensional, as large standing armies are almost always better, but players incur huge penalties for exceeding their supply cap; 2) total city has a much larger impact on one's ability to maintain a large army than total economic strength/output, which both makes little intuitive sense and unduly punishes tall players. Currently on VP, wide and "thick" playstyles are essentially required because tall civs get bodied militarily by tall civs essentially every time. Thus, even if we changed the supply cap system and unit strength/cost scaling to make late game armies smaller and more manageable, we would still be relying on an undynamic, over-complicated system that makes combat and the game as a whole less interesting.

Spoiler Explanation :

1. While units still have gold maintenance, and thus players incur penalties even before supply cap is reached, for most of the game the difference in maintenance costs, both on a per-unit and whole-army basis, between being at 50% of your supply cap and 90% of your supply cap is pretty small and generally has little bearing on the size of your army. Sure, there are periods of the game, such as most of Industrial, where you feel a lot of pressure from unit maintenance, but compared to building maintenance and even tile maintenance it is still quite small. Sure, keeping a small standing army does allow you to build more buildings and thus increase your city yields, but for one I think that after around mid-medieval training units becomes much less of a burden than it is in Ancient and Classical, and for two I think that we as players tend to overestimate the strength of a lot of buildings and underestimate the strength of a large standing army, especially for later game buildings with high maintenance costs.

2. Wide empires have a massive leg-up in terms of supply, even when the gap between a wide empire and tall empire in terms of production and gold income is comparatively much smaller. Different yields scale differently with regard to empire size: gold doesn't scale very linearly with city number because trade route number is pretty flat and taller cities can work more gold-heavy tiles; production and excess food don't scale very linearly with city number as smaller empires have less unhappiness and can grow larger cities that have better tiles; yet unit supply scales essentially linearly with both city number and total empire population, which combine together to create non-linear supply cap growth. Say you have two empires in Industrial, both with 100 total population, but one has 5 cities and the other has 10. The empire with 10 cities is going to have a higher military unit supply cap, sometimes greatly so, like 90% of the time. Only if the tall empire picked up Terracotta, Brandenburg, Himeji, and/or Great Wall do they have any shot at being able to support a larger army, even though the taller empire could have higher empire-wide yields of every kind and thus support greater unit production and maintenance. If that overextended wide empire attacks the tall empire, the tall empire might be able to out-produce the wide empire and thus possibly win in a protracted war, but because of production penalties for exceeding their supply cap the tall empire is much less able to actualize their economic superiority. Wide play should have its benefits, of course, but currently it is just too much of a benefit for warring.

So what do I think might be better? I think @axatin is on to something by trying to solve late-game army bloat at least partially by increasing unit gold maintenance costs and not through supply cap changes. Increasing gold maintenance while keeping unit supply unchanged does shift balance more in favor of taller empires, as as mentioned previously gold does not scale linearly with city number, unlike unit supply. However, this still doesn't address something that at least I consider a problem: the fact that the most impactful penalties for maintaining a large army are only incurred after your army reaches a particular size, and they are then pretty massive penalties. Here is one alternate method for regulating army size, though it would entail some fairly large changes:

1. Supply Cap is removed.

• Military/defensive buildings are compensated in some way, maybe by giving them slight negative maintenance costs modifiers, stronger happiness bonuses, adding production modifiers for units, or providing flat gold yields.
• By relying on yield maintenance costs, supply burdens could be made more proportional to unit cost and strength, allowing us to better balance and differentiate both cheap/weak units and expensive/strong units like Pikemen and Tanks/Planes, respectively.

2. Military units cost both gold and food maintenance.

• This means that armies affect growth not just after an arbitrary size is exceeded, but always, making army maintenance much more dynamic.
• This would mean that there would have to be buffs to food tiles/buildings to compensate for the extra food eaten and maintain amounts of net food. Pillaging food-bearing tiles could also give food yields to your closest city, or instead of giving yields, give the pillaging unit a temporary promotion that removes its supply cost for x number of turns.
• I don't think food maintenance should be as big of a burden as gold maintenance, as large food maintenance costs would likely be too hard on wide play. Maybe a ~2:1 ratio of gold:food maintenance would be good.

3. Gold and food maintenance for military units is extracted on a per-city basis, not on gross total empire gold as it is now.

• This means that army size also affects city happiness, as city yields are penalized based on army size.
• The amount of gold/food taken from each city would be spread over each city in proportion to that city's output.
• Unhappiness modifiers for gold and possibly also distress would need to be adjusted, the degree to which depending on to what extent extra food and gold yields are added to compensate.
• War Weariness would have to be adjusted to compensate for the lack of supply reductions. I think simply having War Weariness also increase unit supply costs (in addition to existing increases to training/upgrade costs) would likely work well, but fundamentally changing unhappiness from War Weariness could also work, i.e. having War Weariness add a scaling modifier to some/all needs instead of a simple flat amount.

Even if my idea for food maintenance is unpopular, I still think getting rid of the supply cap and instead having higher gold maintenance that is extracted at the city level, not at the empire level (and thus affect happiness) would be a very positive change and would simultaneously make the supply system more dynamic and more balanced between tall and wide empires. It would also pave the way for future changes to unit and army size scaling that would address army bloat and make late game combat more interesting. I have ideas for those changes, but that is a whole other discussion. For now, let me know what you think, and I'd love to discuss ideas related to supply further!

Edit: Here's how the current supply/maintenance system works, as far as I understand it. If I missed anything or got anything wrong, please let me know:

Spoiler Current Supply/Maintenance System :

Currently, supply sets a limit of how many units your empire can support before incurring growth and production penalties, and is provided/modified in a number of ways:

1. flat supply from buildings, policies, and expending Great Generals and Great Admirals

• buildings like the Barracks give +1 supply per city

2. flat supply from tech level

• actually a penalty that decreases as you research more techs

3. scaling supply based on city population

• buildings like walls give +x% supply from population
• supply from population starts at +0%, and rounds down. This means a city that gets +10% supply from population gives 0 supply if its population is less than 10, and gives 1 it is between 10 and 19, etc.
• early game buildings tend to give more flat supply, later game buildings tend to give more population supply

4. flat supply based on difficulty

• like tech level, its actually a penalty that decreases with lower difficulty.

5. percentage supply decreases based on War Weariness

• not exactly sure how it is calculated, but seems to scale linearly with War Weariness, and maxes out at -75% supply

6. Some units, either as a feature or due to policies, do not cost supply

• I'm unsure if Grand Canal also makes some of your units not cost supply or if it just makes them not cost maintenance

These sources of supply work together to increase total supply as the game goes on in a non-linear and relatively inconsistent way.

If you exceed your supply cap, every city in your empire receives a -5% food and production penalty for each unit over the cap. This is a pretty serious penalty, as it modifies base food, not excess food, and I believe modifies base production and not modified production. Thus in practice it can actually be more like a 8-10% penalty per unit depending on how many food and production modifiers you have in a city.

Let's say you have 100 supply, but you just created some units and now have 105 units. These 5 extra units only represent a 5% increase in the strength of your army (10% if you want to use Lanchester's Laws), but will incur a 25% percent growth and production penalty on your cities, which is likely more like 30-50% in practice. That is huge!

Maintenance is determined by the following formula:
UNIT_MAINTENANCE_GAME_MULTIPLIER = 7
UNIT_MAINTENANCE_GAME_EXPONENT_DIVISOR = 6

• The multiplier is determined by tech level, while the divisor is determined by army size.
• Thus, per-unit maintenance increases as the game goes on and as your army gets larger. However, the affect of the latter is not super substantial (though Axatin's congress proposal will make it more impactful).
• As far as I know, unit maintenance does not differ amongst units. Thus a Pikeman and a Knight, two units available at the same point in the tech tree but with massively different strengths and costs, not only both cost 1 supply, but also the same amount of gold maintenance, though the Knight does use a strategic resource.
• Maintenance costs are deducted from your gross empire income, and thus do not affect happiness at the city-scale.

 

[CppMaster]

Yeah, I agree. Supply cap is very artificial. I'd prefer if the penalty was more linear instead of the wall.

 

[ma_kuh]

I like the idea of flexing the maintenance tax into another yield.

I don't know that you need to throw out supply entirely though, using the happiness system as a model you could have penalties and cost-increases tied to a more gradual system, which solves your concern regarding sudden changes in penalty/no penalty. Just spitballing, but something like:

• Overall maintenance costs scaled by used supply %, e.g.:
  • You have 10 units, costing a base of 10gpt maintenance.
  • If you are 10/10, you have a cost scale of 100%, so 20gpt total.
  • If you are 10/20, you have a cost scale of 50%, so 15gpt.
  • If you are 10/100, you have a cost scale of 10%, so 11gpt.
• You could tune the cost scale as well, so that 100% supply = 50% cost scale, as desired. Numbers are placeholder.
• Above 100% supply, you would have morale penalties set in, as they do now. Again, an easy option might be excess supply %/total supply is the penalty.
  • Following the above example, if you had 11/10, you would be 10% over, and receive a morale penalty of -10%.

Finally, I wanted you call out that dynamic supply costs for different unit types/classes seems great, it would be an additional tool for keeping things like tank:infantry balance under control.


It would also be good for this discussion if you could share the details of the current system. Ideas a good, but talking about a bunch of things that might already be happening behind the scenes will just make us talk in circles. If you have the time, it would be appreciated.

 

[Diomedes_]

@ma_kuh I just added an overview of the current supply/maintenance system at the end of the OP!

 

[Tekamthi]

I like the supply cap in that its pretty clear to the player what the limit is, I don't have to have a calculator or spreadsheet out next to the game to understand where I am on the invisible supply curve etc -- there's a limit, I'm under it? Then I'm good.

I infer from this round's proposal that there is already some invisible cost scaling taking place outside the supply system -- not such a fan of these invisible things. Give me some UI method to see at-a-glance what penalties I'm facing for having too many units, then I don't mind these sort of features.

There are some interesting ideas in OP, the per city supply system feels familiar, I'm sure one of the past civ games did things this way. Food makes sense as well.

Ultimately though I'd find this supply question most interesting if it were solved somewhat dynamically, and thematically. Right now each unit costs about the same supply and maintenance as far as I'm aware, no matter the unit, no matter what its doing. Being far from home territory should cost more. Having an elite, highly promoted unit should cost more than a level 1 rookie. A battleship should cost a lot more maintenance than an infantry division, etc.

I don't like having a unit cost more just on the basis that there are a lot of other units (the supply system should address this) -- things don't work that way irl: small army is more expensive on a marginal, per unit basis than large, all else being equal, and assuming they're not so large to disrupt the entire economy (ie over the supply limit). But if a unit costs more cuz it is fighting every turn and repairing etc, this makes perfect sense in my mind... Also, it doesn't have to be ONE system for the entire game, systemic changes could represent progress and advancement: maybe ancient era we start out on a per city supply basis, but upon discovery of civil service or banking we then get the national system we have now...

 

[CppMaster]

I like the supply cap in that its pretty clear to the player what the limit is, I don't have to have a calculator or spreadsheet out next to the game to understand where I am on the invisible supply curve etc -- there's a limit, I'm under it? Then I'm good.

I infer from this round's proposal that there is already some invisible cost scaling taking place outside the supply system -- not such a fan of these invisible things. Give me some UI method to see at-a-glance what penalties I'm facing for having too many units, then I don't mind these sort of features.

Sure, I assumed that it would be shown in UI, because why not? Clarity is important.

Ultimately though I'd find this supply question most interesting if it were solved somewhat dynamically, and thematically. Right now each unit costs about the same supply and maintenance as far as I'm aware, no matter the unit, no matter what its doing. Being far from home territory should cost more. Having an elite, highly promoted unit should cost more than a level 1 rookie. A battleship should cost a lot more maintenance than an infantry division, etc.

That would be cool!

 

[Stalker0]

1. While units still have gold maintenance, and thus players incur penalties even before supply cap is reached, for most of the game the difference in maintenance costs, both on a per-unit and whole-army basis, between being at 50% of your supply cap and 90% of your supply cap is pretty small and generally has little bearing on the size of your army.

I disagree with this. While there are games where my coffers are full of cash, there are plenty of others where money is very tight, and I literally don't have enough gold to full my unit supply....let alone than have gold for upgrades. Maintenace already is a solid factor when your army gets very large.

 

[Tekamthi]

there are two existing fields in 'Units' table that should be examined before building new supply mechanisms: <SupplyCapBoost> and <ExtraMaintenanceCost>

while it does not appear that modder can affect the supply cost directly (ie have some units cost 2 supply vs 1 for others), I infer that you could just decrement the SupplyCapBoost by the amount corresponding to the extra supply desired. ExtraMaintenanceCost should allow costs to be increased on a per unit type basis.

 

[ilteroi]

nice idea but also has a big UI impact ... you need to be able to see how many units a city is supporting ... can you transfer ownership of a unit etc etc

 

[azum4roll]

Supply from difficulty:

Base: 8 for AI, 6-10 for humans
Flat supply DECREASE per era: -1
Minimum 0

---

Supply from cities:

Base flat supply per city: 1

Sources of buildings that increase flat supply:
Barracks: +1 (Ikanda: +2)
Armory: +1 (Dojo: +2)
Military Academy: +1
Lighthouse: +1
Harbor: +2
Teocalli: +1
Terracotta Army: +5
Great Wall: +3 (obsoletes at Gunpowder)
Himeji Castle: +5
Brandenburg Gate: +10
Pentagon: +1 per city
Piazza San Marco: +5

Supply DIVISOR from tech: 1 + tech progress% * 5/6

e.g. If you have researched 60% of all techs, your supply from cities is divided by 150%.

---

Supply from population:

Base %supply from population: 25% for AI, 15-35% for humans

Sources of buildings that increase %supply from population:
Walls: +10%
Castle: +10%
Bastion Fort: +10%
Arsenal: +10%
Military Base: +20%
Stable: +10%
Seaport: +20%
Heroic Epic: +10%
Parthenon: +10%
Red Fort: +5% to all cities
The Motherland Calls: +100% (?!)
Arsenale di Venezia: +5% local and +10% to all cities
Royal Guardhouse: +20%

Sources of policies that increase %supply from population globally:
Dominance: +10%
United Front: +50% (!!)

Sources of traits that increase %supply from population globally:
La Grande Armée: +25% (I think they need a maintenance reduction more...)

Non-Venice puppets: halved (multiplicative)

Supply DIVISOR from tech: 1 + tech progress% * 7

e.g. If you have researched 60% of all techs, your supply from population is divided by 520%.

That sounds like a lot.

---

Supply from expending Great People:

Expending Great General: +1
Expending Great Admiral: +1

---

Final supply is adding the above 4, then reduced by War Weariness (max -75%).

 

[azum4roll]

I think the reason why wide seems to have much more supply than tall is that the population supply reduction from tech is enormous.
Even for a 30-pop city with every defensive building + stable + seaport built, assuming you beeline Military Base, the divisor is 587% and the supply from population of that city is only 5.88.
Meanwhile the flat supply from the city is still 4.43.

If we reduce the divisor from population (or maybe it's not even needed if we only want to nerf wide supply) and increase the divisor from cities, I think this can be somewhat fixed.

 

[Amask]

Agree with every word in the OP.

 

[AndreyK]

I think that we definitely should reduce divisor from population and increase divisor from cities.

 

[Diomedes_]

@azum4roll thanks for giving a much more thorough and accurate description of the system, and yes I think you're totally right that reducing the divisor from population would be a good place to start (in addition to the maintenance proposal in the current congress).

 

[azum4roll]

Top panel supply tooltip for @Enginseer since it's a pain to type on Discord:

Military Supply from Cities: 30
Divisor from Tech Progress: 150%
---------------------------------------------
Sub-total: 20

Military Supply from Population: 52
Divisor from Tech Progress: 520%
---------------------------------------------
Sub-total: 10

Military Supply from Difficulty: 6
Reduction from Era: -4
---------------------------------------------
Sub-total: 2

Military Supply from Great People: 3
---------------------------------------------
Sub-total: 3

Total Military Supply: 35
Modifier from War Weariness: -20%
---------------------------------------------
Net Military Supply: 28

(Add icons where appropriate)
What do you think of this?

 

[Enginseer]

...

Possible UI display
Total Military Units: # (# Supply has been used)
Military Supply Values:

• +Net from owned Cities ( Gross - Divisor from Tech Progress)
• +Net from population ( Gross - Divisor from Tech Progress - from Puppets / Occupied Cities)
• +Net from Difficulty (Base - Era)
• +Total from Great General/Admiral

Military Supply Reductions:

• -Whatever from War Weariness (War Weariness Supply Reduction cap insert here)
• State the line for Divisor for Tech Progress

---------
# Total Military Supply

Military Unit Gold/Faith Cost going up from War Weariness.

Possible UI display for this.

Going to post my thoughts on the Supply System.
Tech progress is defined as tech researched / available techs to research

	Current	Proposed	Justification
Flat Supply Bonuses Penalty	2-46%	2-50%	Increase the penalty, and keep the supply from population penalty consistent so they are easier to balance.
Supply from Population Penalty	8-88%	2-50%	Nerf sources of Supply from Population also, so the penalty doesn't make the effect meaningless.
Occupied Penalty on Population (No Courthouse)	0%	-50%	A city with an oppressed population should not provide more population than a more tame puppet governor.
Difficulty Flat Supply Bonus	Whatever it is currently	+1 the current number (to Human, AI, but not City-State)
Difficulty-based Supply from Population	35%-15%	20% (to Human, Ai, and City-State)	Difficulty can really skew the amount of "army" a player needs to have.
Base Supply per City	+1 Supply	0 Supply	Reduces Wide/Tall Deficit.
Lighthouse (and its UB)	+1 Supply	0 Supply	I don't think of getting +1 Supply when I build a Lighthouse. Lighthouse's growth bonuses should complement the supply from the population instead.
Harbor (and its UB)	+2 Supply	+1 Supply	Part of reducing coastal city/landlocked deficit. One got moved to Bastion Fort.
Seaport	20% Supply from Population	+1 and 10% Supply from Population	Since we nerfed supply from the population penalty, we nerf the sources as well.
Walls, Castle, Stable, Bastion Fort, Arsenal, Military Base's Supply from Population	10%/10%/10%/10%/10%/20%	All 5%	Since we nerfed supply from the population penalty, we nerf the sources as well.
Bastion Fort's New Change	0 Supply	+1 Flat Supply	Part of reducing coastal city/landlocked deficit. Stolen from Harbor.
Parthenon	+10% Supply from local Population	0%	Gone! Parthenon is nerfed!
Justice's Royal Guardhouse	20% Supply from Local Population	+10% Supply from Population in all Cities	Stolen from Dominance.
Dominance	+10% Supply from local Population	+5 Flat Supply	No questions asked, get 5 flat supply. Way better to see than waiting for 50 Population.
Autocracy's United Front	+50% Supply from Population in all Cities	+25% Supply from Population in all Cities	Since we nerfed supply from the population penalty, we nerf the sources as well.
France UA	25% from Population in all Cities	15% from Population in all Cities	Since we nerfed supply from the population penalty, we nerf the sources as well.
Intent (mathematical calculation below)
Avg Landlocked City's Max Flat Supply+Supply from Population	+4/+60% (+10% from Stable)	+4/+25% (+5% from Stable)
Avg Coastal City's Max Flat Supply+Supply from Population	+4/+60% (+1 from Lighthouse, +2 from Harbors, +20% from Seaport, +10% from Stable)	+4/+25% (+1 from Harbor, +1 and +10% from Seaport, +5% from Stable)

 

[Diomedes_]

Love this!

 

[Stalker0]

Asking teh question. If we mostly did away with pop based supply, and focused mainly on flat bonuses....would we even need supply reducers from tech?

Its strange that we balloon up supply but then beat it down with a stick due to tech. If we are truly doing a cleanup, maybe we just start there.

 

[alchx]

Asking teh question. If we mostly did away with pop based supply, and focused mainly on flat bonuses....would we even need supply reducers from tech?

Its strange that we balloon up supply but then beat it down with a stick due to tech. If we are truly doing a cleanup, maybe we just start there.

Very good question.

Probably, the reduction from technology should have pushed the player to change the type of his army - new units are stronger and, conditionally, 2 new ones are as powerful as 3 obsolete ones. There is less gold in vanilla than in VP and it is not always possible to modernize units due to a lack of funds in the treasury. It's easier to send these obsolete units as gifts to CSs and produce new ones. But in VP, unit experience is much more valuable to humans, since the AI initially has bonus XP for recruits.

 

[Enginseer]

Asking teh question. If we mostly did away with pop based supply, and focused mainly on flat bonuses....would we even need supply reducers from tech?

Its strange that we balloon up supply but then beat it down with a stick due to tech. If we are truly doing a cleanup, maybe we just start there.

In the current model:

Supply is limited by the amount of population, cities, tech, and flat supply bonuses.

If we threw away the population-based supply bonus and focused on flat bonuses then supply would have no limit. It would only be increased as you went along the technology line. The supply reduction from technology is there to ensure you have a proper infrastructure to maintain an army rather than beelining technology asap to get supply up.