[Economics] Will the King's Purse mechanics be revisited/revised?

[QuintusSertorius]

People have reported issues with some factions being able to recruit stack after stack, and I suspect the culprit is the King's Purse mechanic. This gives the AI money to ensure they don't go into the red, in other words they have virtually unlimited money. This is why one-province Pergamon or Pontos, for example, can maintain several full stacks without going bankrupt. Problem is, even aside from it being ahistorical, if set too high, it makes the economic game meaningless. You can't force the decline of a faction through economic warfare, blockading ports and disrupting trade. Indeed it means the only faction who can be affected by that sort of thing is the human player's faction. I don't mind the task of the player being made harder, but it removes a whole tier of interesting gameplay when economics don't matter.

Is this likely to change in later releases?

[Shadowwalker]

The Kings Purse is not the main part of the problem, Quintus.
It's just a couple thousand mnai at best (5000 is max at the moment, iirc).
The main part is indeed the money script, which adds 1000 mnai to the purse whenever an AI faction is losing money.

I am experimenting with a reduced money script (adding 300 instead of 1000, and substracting 400 instead of 1000 if the factions is making money and has more than 20k.

The results so far are promising - there are a lot less stacks around, especially factions like Pergamon, the Areuakoi and the Nabataeans don't field 5 or 6 stacks anymore, not even after 200+ turns.

Although there are disadvantages, too - for example the fact that Qart-Hadasht is always eaten up by the Massylians. :wall:

I just wish I'd make any progress with the planned script to have a King's Purse that is dependent on the factions' size. If that would work as planned, we could completely skip the current flat "if the AI loses money, throw more money their way" Money Script. . .

[V.T. Marvin]

It is very likely to change! :yes:

Consider the current money script as a placeholder intended primarily as way to make sure that no faction becomes crippled by being in debt. It does the job, yet at a cost of way too many armies (https://forums.totalwar.org/vb/showthread.php?148267-Way-To-Many-Armies-%21%21%21%21)...:shame:

It was based on the assumption that using a mechanism that the AI is actually aware of will make it more likely to take proper stock of its finances and manage its resources in a reasonable manner. That assumption proved false - no matter how well of a faction is it always arms itself to death with unsustainable recruitment which pulls it into crippling debt. :wall:

Clever ideas for a more appropriate money script are much needed and welcome. :bow:

[mogami]

Hi, EBI used the Rome Total War-Alexander engine where when you recruited a unit the population decreased (if you disbanded a unit in town it increased population) EBII is using MedII where population is not effected by recruitment. This allows factions an endless pool to recruit from. Combined with endless money.... I don't think MedII gave the free money (I might be wrong but I don't recall so many stacks)
The vanilla Rome also impacted population
Does sending out colonists effect population? )decrease sending increase receiving)

[QuintusSertorius]

Hi, EBI used the Rome Total War-Alexander engine where when you recruited a unit the population decreased (if you disbanded a unit in town it increased population) EBII is using MedII where population is not effected by recruitment. This allows factions an endless pool to recruit from. Combined with endless money.... I don't think MedII gave the free money (I might be wrong but I don't recall so many stacks)
The vanilla Rome also impacted population
Does sending out colonists effect population? )decrease sending increase receiving)

Units have a refresh pool which controls how frequently they can be enlisted, so depleting population isn't necessary. However, reducing the rate of refresh might be necessary.

[mogami]

Hi, How are refresh rates determined? What does city population actually represent? (what is scale) At start the city of Rome has a population of 9.5k (compared to 7.5k in vanilla RTW)
With population reduction the city could build fewer then 40 units (2 stacks with unit scale set to huge) The cities tax base would collapse long before ever actually recruiting 40 units. I'm going to test (since we are all really beta testers) how long is needed to recruit 40 units just from city of Rome.

[QuintusSertorius]

Hi, How are refresh rates determined? What does city population actually represent? (what is scale) At start the city of Rome has a population of 9.5k (compared to 7.5k in vanilla RTW)
With population reduction the city could build fewer then 40 units (2 stacks with unit scale set to huge) The cities tax base would collapse long before ever actually recruiting 40 units. I'm going to test (since we are all really beta testers) how long is needed to recruit 40 units just from city of Rome.

Refresh rates are determined by the unit - each entry in the EDU has a number for how quickly the pool fills up once it's available to recruit. I'm guessing there's probably a multiplier for the size of settlement, and maybe some hidden resources or other building triggers too. City population represents the number of adult male citizens or families, but has nothing to do with recruitment at all. I've no idea if that's hardcoded in the M2TW engine.

[mogami]

Hi, Thanks for the quick response. How historically technical do we want to get? I'm using Rome as my example mainly because that faction pretty much drives the game. In our starting time frame Rome raised 4 legions annually.(by law) Two per consul. But the soldiers were not full time professionals they went home at end of campaigning season and new legions were raised the following year. Rome could and did raise "emergency" legions. What is the scale in units? Since players can choose unit size if they select huge a Roman unit has between 100-200 men for infantry. A legion is this period would have around 4800 infantry (in 3 classes) and 200-300 cavalry. It appears the scale is not the same for infantry and cavalry. They raised all 4 legions together every year. (when the recruits arrived at muster they were divided by age and property into the various classes of infantry. Each tribe contributed a set number of men per year.
The later reforms changed all this but I'm sticking to our starting time frame. Rome should be able to field 20k Roman citizen soldiers per year pretty much for free. (Since the men brought their own equipment) Rome would be one of the larger factions at 20k
They would add a like number of allies as auxiliaries. By the end of the second Punic war the overall population had declined to where the state had to recruit poorer males and pay for their equipment. Now the organization does not change but the units get more expensive.

[V.T. Marvin]

City population represents the number of adult male citizens or families, but has nothing to do with recruitment at all.
That's correct. :yes:

Refresh rates are determined by the unit - each entry in the EDU has a number for how quickly the pool fills up once it's available to recruit. I'm guessing there's probably a multiplier for the size of settlement, and maybe some hidden resources or other building triggers too.
Not at all. :no:

Replenishment rates and recruitment pools are defined in EDB and lots of factors were taken into account, however population size is not among them directly because there is surprisingly no way to check for populations size other than checking for the core building's level and that causes CTD if combined with other boni.

Therefore look at it this way: population size represents the tax and other revenue base of the region, whereas the capacity to recruit armies depends on how the region is governed and that in turn depends on WHO is doing the governing (faction) over WHOM (region) and how is the region developed (infrastructure buildings as pre-requisites for a particular government type/level).

The recruitment system is fairly complex, yet we aimed for a certain uniformity and consistency which might be inferred from the comments in the code. ~;)

[mogami]

Hi, Wow you know this is a complex question. I am on your side. I don't have an answer for how to recreate the way factions create their armies where history and computer engines meet in harmony. Most of the armies in our starting period did not exist year out.
In reality if city "a" fields an army then it should lose it's industry and farming while the army is in the field. But also when at peace there should be no upkeep for units because the soldiers are out in the fields/mines/shops Can you connect "at war" or ""at peace"?
There are no standing armies at start of our game. Classical Sparta is gone. The Roman Republic changed exactly because of the need to have a full time army. They were one of the few factions that could pay for the change.
Personally I favor no upkeep costs at all for units in their home. But the moment they step outside..... If a faction deploys their army then their farming and industry shut down. Now 300+ turns into game this will begin to change. Now units cost upkeep in town. When you change from citizen soldier to full time the economics change. How you do this with the mechanics of the engine I offer no help. Sometimes I think it would be easier to start from scratch. Having experience with large projects scares me because they grow beyond initial goals.
I am a grognard and sometimes details do distract me from the otherwise great and difficult work people do. I know and you should too that you can't make everyone happy. But never say you can't redo or undo work because of how much you've already done. Don't waste hard work with less then accurate results. Personally I could care less what color uniforms a faction wears. I don't really care about much except does it work. I can not stand the original versions of most of the games I play. Recall Egypt in the original Rome total War? That's what started EBI. Take as much time as you need and we will be here to play test your efforts. I have played over 1000 turns of EBII and while I still point out things I still also say it's better then I could have dreamed of 10 years ago. If I sound nit picky it's only because I 'm a nit picky SOB
Oh I remember what that subject was. The whole model for early game recruitment is funky. First we have to agree on what would be historical. I played a game as Britain where I marched an army for 4 years across Gaul to Rome. They would have assassinated me and went home the first winter. The logistics are too modern where armies can stay out multiple seasons. When people say there are too many armies what they are really saying is there are too many armies year round. It should be possible to march an army in winter without a movement restriction. The restriction should be on the size and cost of that army.
This brings me back to the population question. The population can't be soldiers in a far off land and farmers and tradesmen at home paying their upkeep

[QuintusSertorius]

Not at all. :no:

Replenishment rates and recruitment pools are defined in EDB and lots of factors were taken into account, however population size is not among them directly because there is surprisingly no way to check for populations size other than checking for the core building's level and that causes CTD if combined with other boni.

Therefore look at it this way: population size represents the tax and other revenue base of the region, whereas the capacity to recruit armies depends on how the region is governed and that in turn depends on WHO is doing the governing (faction) over WHOM (region) and how is the region developed (infrastructure buildings as pre-requisites for a particular government type/level).

The recruitment system is fairly complex, yet we aimed for a certain uniformity and consistency which might be inferred from the comments in the code. ~;)

Ah, my mistake. I'm looking at the EDU and I've only just realised the line with the refresh rate/pool cap on is commented out - it's a flag, I guess, to the value in the EDB for cross-referencing.

[V.T. Marvin]

Ah, my mistake. I'm looking at the EDU and I've only just realised the line with the refresh rate/pool cap on is commented out - it's a flag, I guess, to the value in the EDB for cross-referencing.
Exactly! :yes:

...and an out-dated one actually so there is no direct cross-referencing and the actual EDB code follows a logic of its own - also explained in the commented-out part of the code just in case anybody would like to embark on further tweaking. ~;)

[Bucefalo]

Even in vanilla M2TW there were money scripts to help the AI. I suggest looking at how other mods have tackled this issue.

http://www.twcenter.net/forums/showthread.php?491010-AI-Unit-Upkeep-and-the-Money-Script-Perpetual-Growth

Here is a script that could be used, which freezes the AI recruitment if the AI treasure goes under X money, then activates it again once it goes over Y money.

Also take a look at these scripts which are very interesting
http://www.twcenter.net/forums/showthread.php?382574-Modding-Money-Script-Generator

I hope this is some help. Best of luck.

[Harkon]

The Kings Purse is not the main part of the problem, Quintus.
It's just a couple thousand mnai at best (5000 is max at the moment, iirc).
The main part is indeed the money script, which adds 1000 mnai to the purse whenever an AI faction is losing money.

I am experimenting with a reduced money script (adding 300 instead of 1000, and substracting 400 instead of 1000 if the factions is making money and has more than 20k.

The results so far are promising - there are a lot less stacks around, especially factions like Pergamon, the Areuakoi and the Nabataeans don't field 5 or 6 stacks anymore, not even after 200+ turns.

Although there are disadvantages, too - for example the fact that Qart-Hadasht is always eaten up by the Massylians. :wall:

I just wish I'd make any progress with the planned script to have a King's Purse that is dependent on the factions' size. If that would work as planned, we could completely skip the current flat "if the AI loses money, throw more money their way" Money Script. . .

Mind sharing this altered script please? Seems decent. Shadowwalker