Individual Unit Maintenance Costs!

This is not much of an idea as it should obviously have been included into the CIV series as far back as Civ1!

Doesn't it make sense that a Battleship costs more to maintain than Infantry?
Realism aside, by giving all units the same maintenace cost as is the present case with Civ3, there is little incentive to build different units within a similar role (i.e. nothing to stop you from manning all your cities with Tanks instead of Infantry --resources aside). Also, poor civs can maintain expensive units almost as easily as rich civ can --in reality poor countries don't have Carriers not because they don't have the knowledge of modern shipbuilding but because they could never afford such a huge investment.

Example:

Foot unit: 1 gold per turn
Mounted unit: 2 gold per turn
Tank: 4 gold per turn
Battleship: 12 gold per turn

There should be more to building units than just the time needed to build them --resources aside.
This would intertwine economics and units far more than they are now thus giving economics a far more strategic rol; i.e. the richer you are, the more advanced units you can build (makes sense to me).

At one point I thought I saw this as a C3C idea but I don't think it will happen. I do like it though. Even better would be to allow difference between governments. Then civ2's Fanatic could be done more easily.

I agree different units should have different maintenance costs. I like the idea of governments dictating the maintenance costs. A fascist government would have lower maintenace costs than a democracy etc...

Different Civ3 governments already have the capacity to have different unit maintenance costs. Using the Editor, you can make Democracies maintain units at 2 gold each instead of 1, if you like. Without changing the default rules, Civ3 governments just give free maintenance based on the number of cities and some free units regardless of civ size (depending on the govt.).
The problem is of course that it still applies to all the units equally.

I guess that if units had different maintenance costs, the solution would be to have governments change the costs using percentages of the base cost; e.g. units under Communism cost 50% less to maintain than the base cost, so a Tank would cost 2 instead of 4 gold each turn (rounding up) --limiting it to 1 of course, so that the Reds don't get Infantry for free.

In the case of the Fanatic unit, I would just give it a "50% cost for Fundamentalism" flag which would would decrease the unit's cost in addition to the cost modifier given by the Fundamentalist government.

...but I don't think it will happen.

Click to expand...

Yeah. You're probably right. It's too bad though because that's a request that players seem to agree on.

It would be REALLY good.
A must-have I could say.

You saw the link I take it. Thank YOU.

All it would require is a single 'Maintenance' field in the Editor's Únits'window. Can't get any simpler than that folks.

As for the AI, I assume that if it already takes unit maintenace into account when building units, then taking variable maintenace costs into account should not be a problem...if the AI doesn't take maintence into account, then shame on Civ3's designers for overlooking it.

The only two aspects of the interface that would need to be changed are the Government modifier, which would now have to be a percentage (shouldn't be a problem), and the problem of which units get disbanded first if you run out of gold (the program could simply be set to disband the most expensive units first).

As I said above, when I first bought Civ3 I was shocked when I realised that all units had the same maintenance cost --although I am happy that at least they no longer cost shields per city as was the case in Civ2.


Maybe I should open a poll on this? (People seem to respond more to those threads.)

Do a poll, I bet everyone would agree.

Alot of poeple have expresse4d ideas on how to restrict unit building for smaller nations to economics. This means that wealth not size would matter, which is how it is anyways. Could have some civs like Kuwait or Saudi Arabia who just happen to sit on the worlds supply of oil or uraniaum or aluminum or, well, you get the idea.

Originally posted by yoshi
You saw the link I take it. Thank YOU.

All it would require is a single 'Maintenance' field in the Editor's Únits'window. Can't get any simpler than that folks.

Click to expand...

I really doupt it would be easy as that.

Probably atm its simply : number of units=total upkeep cost

Variable unit upkeep would require more 'complex' (well understanding the calculations is easy but it implementing them in code often far more complex) calculation alcorythm.

like

Number of units with [1] upkeepcost * upkeep cost [1] +
Number of units with [2] upkeepcost * upkeep cost [2] +
Number of units with [3] upkeepcost * upkeep cost [3] +...
and so fort ad infinum (or something) plus there's likely other parts in code that are affected by this plus the balansing issues that come with it (unless its just created for modders ofcource). Also number and complexity of the calculations may greatly affect game performance.


Coding something is rarely as easy as many think or easy it seems from the surface (eg in game). More often than not, in order to change some ingame functionality requires massive changes in code moreso due to most code affecting the rest of the code directly or otherwise.

Idea is great tought and I hope it can and will be done if at all possible.

You just need to make the individual maintenance cost as a multiplier, and it's done.
For the AI as well as for the editor.

I would assume that the calculating mechanism used by the AI to determine if it can afford the expense (that's assuming it limits unit production according to amount of gold and not number of units) would be almost identical to what it uses to determine the maintenance costs of City Improvements. Whether the AI actually uses a mechanism or if it just builds stuff until it can no longer afford to do maintain it is anybody's guess. (If anyone knows, feel free to add that information.)

As long as the AI doesn't just build a whole bunch of the cheapest units available (something AI's in other games tend to do; i.e. quantity not quality) then balancing issues should be few --it's just a matter of play-testing.

The main issues that would get in the way of this that I can think of are:
a) Government Unit cost mofifiers: as things are, they only account for general cost (i.e. governments can increase the general amount but not specific amounts). The simple solution to this is to have the government cost modifier represent percentage of unit maintenace cost rather than a specific integer. For example, if Democracy increases maintenace by 2 (1=100%) then a unit that costs 2 gold per turn will then cost 4 gold. The thing with that is that a unit of 3 (Tank?) would now cost 6. Multiply that a few times and even the US will have trouble upkeeping too many Tanks --in addition to all the other units and improvements. granted, Democracies have double Commerce income so it just means that the player will have to watch the budget closely.
The again, you could just get rid of this feature altogether as it's redundant --I'm pretty sure it's not used in the core game (i.e. all govts. are set to 1).
b) Free/disbanded units: which unit are affected first? Under the present system, there is no bias (unless it's the newest units --I don't know about automatically-disbanding units because I've never had my treasury so low as to actually lose units. I assume that under this system, the units with the highest maintence would disband/cost gold first (i.e. Battleships before Infantry). Although players might complain that they put so much work into building the Battleship so why should they lose it first? All I can say is that it's the penalty for building too much stuff when you can't afford it --besides, the last thing you want is to lose your city defenses (Infantry before the powerful offensive units).

Aside from that, I don't think it would present much of a problem to implement.

Note: most games that include some sort of maintenance feature use the 'individual' system.


BTW, I don't know about the poll. The forum moderator would probably delete that thread since it would essentially be duplicating this one. But I'll try it anyway.

Actually there is already a maintenance field. It's just a 0 or 1 field. An integer flag would solve this. I imagine that they have to run through the types of units and check if that type of unit has a cost now.

Getting the AI to know what to do is the only hard part I see for this suggestion. I think it has a hard time balancing this with 1 gold for all units now (except on the really high difficulty levels where it gets so many free units).

Coding something is rarely as easy as many think or easy it seems from the surface (eg in game). More often than not, in order to change some ingame functionality requires massive changes in code moreso due to most code affecting the rest of the code directly or otherwise.

Click to expand...

the coding you suggestion is really over complication. seem like you 1st sort out the unit by cost than add, sorting to me seem pointless, at most we just wanna know the most expensive unit (in the case of disbanding over-expensive military) since the idea is each unit has a cost, all you have to do is to add the cost instead of 1, and add another higher unit cost variable in current turn to their civilisation.

looking up the unit cost is as easy as looking up the movement rate variable, there is no problem fetching the data for processing. at most is you use up a bit more memory, but given the kind of data wastage in CIV3's engine, it is a rather "cheap" feature to add.

i believe the fact that the feature has been in game far earlier than CIV3 has already proven the math to be simple. it seem like a design failure than an implementation failure to me.

Whether the AI actually uses a mechanism or if it just builds stuff until it can no longer afford to do maintain it is anybody's guess. (If anyone knows, feel free to add that information.)

Click to expand...

The AI doesn't look at what he can afford when building stuff. I've seen an AI build a barracks, the barracks is sold later due to running a defecit, then the AI rebuilds it, barracks is sold due to still running a defecit, rebuilt it, etc. Supposedly, the AI is programmed not to run a defecit, but if he suddenly loses several cities.....

This can be witnessed by playing Chieftain or Warlord level with mass regicide option if you make it so the king units cost upkeep. And have the AI start without any river or coastal tiles. They run out of money, build a spearman, lose the spearman because it can't pay the upkeep, build another spearman, lose the spearman because of defecit, build another spearman, etc., etc. till the end of time. 1000 AD and they are still stuck with 1 city.

Another reason the AI falls apart in the late game. It builds tons of stuff in cities it doesn't need (or the city is too corrupt to really benefit from them).

As long as the AI doesn't just build a whole bunch of the cheapest units available (something AI's in other games tend to do; i.e. quantity not quality) then balancing issues should be few --it's just a matter of play-testing.

Click to expand...

I suspect that is what would happen. Currently, the AI just builds the best unit available based on a/d/m (depending on if it needs offense/defense or ships) . Sometimes, if the shield cost is way too high in proportion to the city production (like in a really shield-poor city) they will build the weaker unit.

Sure, to you and me, we know where it is better to build a bunch of cheap units instead of fewer more expensive units. Try telling a computer all of these 'situations' where tons of cheaper units are better than the fewer strong ones (or when the fewer strong ones would be better). Can you list some *exact* situations where they should build lots of one unit instead of fewer of another unit? Remember, you must be PRECISE! (giving exact criteria for when they should build one unit instead of another).

The AI needs to be told strict guidelines of when to build one type of unit instead of the other kind. The AI can't learn to adapt like a human can. If the AI does one thing, the human learns from it, and exploits it. You try changing the tendancies of the AI in a patch then the human learns from it, changes his tactics slightly and can exploit the AI all over again, in every game.

Bamspeedy starts singing the song "If the AI only had a brain"

Well...that clears up the question concerning AI production.

In that case, giving the AI a variety of units with different mainenance costs would be disasterous.
(I would be cool for Multiplayer though.)

I know I frequentely say that the AI is not our problem and pretty much leave it at that. On further consideration, I think it is our problem. Designers can only be expected to put in the time to work on the AI if it's a 'make-or-break' feature of the game. Is it? Most of the complaints I have read concerning Civ3 tend to be about other features, not the AI. I don't think that most players care how the AI conducts its economics but just that it presents a challenge. Unfortunately, if the AI runs out of gold, it won't even have enough units to throw at you --let alone present a real challenge in the sense that it threatens the existence of your civ.

How can this be dealt with? With much difficulty at best.

The problem is that due to Civ3's scope, the number of possible variables would require a Big Blue's AI just to take it all into account.

The part of the program dedicated to the AI would be bigger than everything else put together --assuming it could actually fit onto your hard drive.

As far as I'm concerned, preset actions are the way to go. Granted the human player can adapt and exploit, but then it's up to designers to tweak the exploitable parts of the game in favor of the AI. They already do this in a way as they give the AI certain bonuses to beef up its play.

In terms of the preset program, I was thinking along the lines of: IF x gold then x unit/improvement; IF war then x # of units towards x civ's border; and so on. In other words, the AI doesn't actually 'decide' but rather is set to perform a scripted action --more into the realm of 'Events.'

It's a touchy subject and I don't know enough about how Civ3's AI 'thinks' to make an accurate judgement of how best to implement the AI aspects of IUMC.