Realism Invictus

[AllTheLand]

I think since you answered so fast, you might have missed my edit in the previous page. I'm seeing 60 gold upgrade cost for a 50 hammer difference between a 104 hammer unit and a 54 hammer unit. I'd expect a cost of 50. Any idea where the extra 20% might be coming from?

Also a side note about the previously mentioned topic of ancient eras going by too quickly: in my new game (16-civ, large map, rev 5436), Islam got founded in 348 BC.

When I was typing my previous post, I wanted to caveat that I was talking about the pre-gunpowder era. A skirmisher with a gun, especially a rifle, is absolutely a menace in rough terrain, and that is properly reflected in later units, I feel. But pre-gunpowder, a skirmisher is often a peasant with a sling or a javelin; in other words, not much different from the very same warband. While there were elite skirmishers (such as Thracian peltasts in Alexander's army), they were more of an exception than a norm, unlike the gunpowder era, where skirmishers were often better trained and equipped than line infantry.

Fair enough.

As you indicated yourself, comparing 50% with walls to 30% without is not fair, as walls also halve the siege efficiency. What you're really getting is 30% vs de facto 100%, at least pre-gunpowder (and even more starkly, at the next cultural level it's nominally 45% vs 50%, while in reality it's 45% vs 100% - so 55% difference rather than 5%!).

That's the case when looking at how long it takes to whittle defenses down - and I'm generally fine with how long it takes to whittle defenses down with a city wall using battering rams or catapults (it can even feel too slow with just rams) - but when it comes to the potential for a quick assault using an important numerical advantage, 50% is most definitely not 100%. The part that specifically bugs me is how 1 or 2 defensive units in garrison can be quickly overwhelmed without having to sacrifice that many attackers.

An idea would be for city walls to give a modifier, on the model of the forts that workers can build. One first strike from walls would really enhance their defensive value. Although at the same time, with how archers are currently balanced, this would probably become obnoxious... So I'm not sure.

Although on the other hand, Civ also doesn't offer way to properly "lay siege" to prevent a city garrison from being reinforced, so that's perhaps an argument against making small garrisons too efficient...

 

[TheBirdMan]

I would also support slightly buffing palisades and walls. Palisades at 25% and being quite cheap don't seem so bad at first glance, but with the basic cultural defense being at 15% and them not adding up, it's often more a meager 10% boost, and making a full unit instead is usually better even if you get the full 25% boost.

I was a little "tired" of having to save "my AIs" from the barbarians, so I made that change with the Palisades and Walls. 20% to the Palisades and 25% more to the Walls.

But-but-but. Then I see that the AIs are actually neglecting to build these defenses. After 174 turns, only 1 out of 21 AIs has made a single palisade at his capital city. The rest of the cities are only defended by 1 or 2 units (one of the AIs had even lost its main city).

Spoiler Lost city :

If it's not possible to get the AI to prioritize the passive defense more then..... well, then it is clear to me, that my idea with increasing the value of the passive defense from palisades or walls not will help much.....unfortunately.

Here at the start I usually help my AIs a bit - I give them back their lost cities - because I'm not interested if they are too weak before the "real" game starts - though it cannot be avoided that some AIs will fall far behind. Usually I check 2-3 times until around turn 300. Around that time the AI should have 4-5 cities - but once the 300 turns are passed, then I have no "pity" for them.


EDIT: So it's more important to get the AI's to build the palisades in the first place.

 

[Walter Hawkwood]

I think since you answered so fast, you might have missed my edit in the previous page. I'm seeing 60 gold upgrade cost for a 50 hammer difference between a 104 hammer unit and a 54 hammer unit. I'd expect a cost of 50. Any idea where the extra 20% might be coming from?

Upgrades always have a base cost, which is currently set to 10 gold. They are always therefore a bit more expensive than the production to gold conversion ratio would indicate.

Although on the other hand, Civ also doesn't offer way to properly "lay siege" to prevent a city garrison from being reinforced, so that's perhaps an argument against making small garrisons too efficient...

Very true; it's quite hard to prevent a city from being reinforced. I don't really see any elegant solutions to that, that's fundamentally how Civ 4 siege combat is built.

 

[[Y]]

Nothing wrong, and as I indicated it was already well possible before the balance changes. But I don't want something to become a no-brainer that people do 100% of the time.

Hmm. If the goal is to prevent early rush of an AI, what if AI's started with a weaker version of the immobile defender barb cities get in world maps? As it is, the AI is its own worst enemy because they fail to use what you give them for defense for that purpose. A unit that can't be used for any other purpose than defending its initial plot may be an appropriate solution.

Even without any additional defensive bonuses, a 5-str unit in a forest defends as 7.5, which is more than a medieval swordsman, and just under a pikeman or man-at-arms. I can see how a 2-move unit like that would be a mighty nuisance to get rid of once it's in your territory (cavalry units are stronger, but they don't get any defensive bonuses ever so dislodging them is generally not a chore).

That's a good point, I hadn't considered that. I think the lack of recon aid is still an issue though. Could Horsemen provide recon aid so that every civ has access to a 5-strength unit providing it?

t's using "Power", which is separate from strength (though it usually roughly corresponds to it), which I didn't change for archers, as their utility in the intended role (city defence) didn't decrease all that much, especially if we go with +100% instead of +75%, so one's total power rating for AI purposes didn't change at all with that revision.

Ahh, interesting. I never knew there was a separate value for that. Since managing power levels is a critical aspect of gameplay, is it possible to expose the power value in the pedia somewhere? I get that in most cases it's effectively the same, but knowing when building a unit actually contributes more or less to power than expected can potentially make a difference.

---

Re: walls -- I think they and palisades are fine as is. Buffing walls will make hilltop cities even more of a pain (not necessary). And palisades, as I see them, is more of a defense for new and vulnerable cities that don't generate culture. They're all serving their purpose well, I think. The only buff I can see making sense is providing a minimum defense (10%?), even after siege units have their turn. No catapult or trebuchet perfectly eliminates every advantage a defender has, after all. Even rubble is a cover of sorts.

 

[Walter Hawkwood]

Oh, I took it as implicit that I hadn't updated since my last post was concerning the same game where I specified that: 5430.

I took a look at the save and even determined the culprit (the settler group in Mongol capital), but unfortunately, I wasn't really able to fix the underlying cause of CTD, as it seems to be something really obscure in settler AI logic (I think AI wanted to settle on the same landmass and got kind of torn up between going to the target tile by boat or by land). If you'd like to continue that save I can easily produce a non-crashing save for you by deleting one of the ships.

 

[AspiringScholar]

Please do. I am excited to try the new changes but wanted to get a bit further in this one first.

 

[TheBirdMan]

And palisades, as I see them, is more of a defense for new and vulnerable cities

And Palisades would probably work as intended as long as you do not play with Barbarians on - maybe with a single exception from time to time. But with Barbarians On..... well those barbarians comes too early - or the palisades comes too late.

I would prefer if the AIs would build Palisades somewhat earlier than they do. But what value in the CIV4BuildingInfos.xml is used? The Modiki writes it's the <iAIWeight>, that set the importance of a building to the AI - but can that be true since this value is 0 (Zero) for all buildings except for GHall (-2) and Harbor (2)?!?!

 

[arizzi]

Both of those are already majorly the case in vanilla and in RI as a consequence: https://www.civfanatics.com/civ4/strategy/game-mechanics/war-weariness-mechanics/ (requests to add things that are already there are my favourite kind of requests - fulfilling them is as quick as writing an answer!)

Interesting so it does work somewhat similarly to how I was suggesting, however there are important nuances.

1. I still think that the civ that declares war should have a modifier giving them more war weariness for everything, even if it's on their territory. Say + 25% to 50% After all, war weariness should also indicate dissatisfaction with your leadership. "The war you declared is going badly and we are losing territory because of you" vs "we got invaded and are defending ourselves/punishing the aggressor"

2. Defending is +2 regardless of whether you win or lose. I think this should be changed to how it works for attacking. Don't really see a reason why it should be a flat rate.

3. WW is gained is based off the culture of the tile rather than borders. That's reasonable. I guess I've been running into an edge case in my games.

That would basically make the losing civ lose more. Not necessarily a bad thing in all cases, but has to be kept in mind - do we really want to make wars more decisive that way?

Ideally this would be in conjunction with aggressors suffering more WW due to changes above. Then aggressor civ would be penalized more.
It would also be beneficial if this prod penalty and WW unhappiness is felt for many turns afterwards, and doesn't go away immediately after peace treaties. I know WW remains and decreases over time according to the formula, however the effects go away until war is re-declared. Losing a portion of your population should have long term implications.


Modeling population loss this isn't really possible with civ 4's existing population mechanic (since we can gain population incredibly quickly with enough food).
Just for fun I was thinking about some more radical solutions. Likely impractical but I will think about these some more.
-Building a military unit could cost a population point.
-Each military unit consumes food


Oh one last thing, I noticed a small inconsistency - maybe slaves should be strength 2 or 3 until warband is unlocked?

Edit: 100% agree war weariness should be shown. Maybe civilopedia could have the formula.

 

[AspiringScholar]

If you play with revolutions enabled, it is shown. The same indication in the military advisor would probably be a good fit for those players who prefer the mechanic off, though.

 

[Walter Hawkwood]

Please do. I am excited to try the new changes but wanted to get a bit further in this one first.

No problem. See attached. I let it autorun for 5 turns without any crashes, and generally it should be fine, as the settler and ship in question live happily ever after.

I would prefer if the AIs would build Palisades somewhat earlier than they do. But what value in the CIV4BuildingInfos.xml is used? The Modiki writes it's the <iAIWeight>, that set the importance of a building to the AI - but can that be true since this value is 0 (Zero) for all buildings except for GHall (-2) and Harbor (2)?!?!

Generally speaking, AI evaluates buildings based on their merit. But if you want to nudge them to like something more or less, that's the exact tag you need to use. You'll have to experiment with the exact numbers to see what's effective, might be up to tens or even hundreds.

Interesting so it does work somewhat similarly to how I was suggesting, however there are important nuances.

1. I still think that the civ that declares war should have a modifier giving them more war weariness for everything, even if it's on their territory. Say + 25% to 50% After all, war weariness should also indicate dissatisfaction with your leadership. "The war you declared is going badly and we are losing territory because of you" vs "we got invaded and are defending ourselves/punishing the aggressor"

2. Defending is +2 regardless of whether you win or lose. I think this should be changed to how it works for attacking. Don't really see a reason why it should be a flat rate.

3. WW is gained is based off the culture of the tile rather than borders. That's reasonable. I guess I've been running into an edge case in my games.

Ideally this would be in conjunction with aggressors suffering more WW due to changes above. Then aggressor civ would be penalized more.
It would also be beneficial if this prod penalty and WW unhappiness is felt for many turns afterwards, and doesn't go away immediately after peace treaties. I know WW remains and decreases over time according to the formula, however the effects go away until war is re-declared. Losing a portion of your population should have long term implications.

I'll consider all those, but I feel that's not going to make it into this upcoming version in any case.

Modeling population loss this isn't really possible with civ 4's existing population mechanic (since we can gain population incredibly quickly with enough food).

I'd say it's actually pretty accurate, historically speaking. Before modern era, populations did recover surprisingly (by our standards) quickly. Even Genghis Khan could only make a dent in real population numbers for a couple of generations, and he tried hard.

Just for fun I was thinking about some more radical solutions. Likely impractical but I will think about these some more.
-Building a military unit could cost a population point.
-Each military unit consumes food

Radical indeed. Also, not very internally consistent - why would a unit of planes consume a population point or eat food, when compared to a unit of infantry? Certainly, modern infantry units are divisions at minimum (so tens of thousands of people) while a unit of planes is hundreds at best.

Oh one last thing, I noticed a small inconsistency - maybe slaves should be strength 2 or 3 until warband is unlocked?

You can't have slavery before Warbands are unlocked. Warbands come with Bronze Working, which is a requirement for Mining, which unlocks Slavery.

If you play with revolutions enabled, it is shown. The same indication in the military advisor would probably be a good fit for those players who prefer the mechanic off, though.

I would even say it is important enough to be shown on the main screen, like the GP tracker, at least while you're at war.

 

[shkitur1]

This mod has the best graphics then any other mod Ihave played. Buildings Improvements and especialy units are amazing. Only 1 Improvement stands out. Fort. For ancient and medieval eras it is fine, but in renessainse industrial and modern eras it looks obsoleted on map. In industrial era more suitable model for fort can be something like star fortress, and in industrial/modern eras it could be bunker or air base. I dont know is it simple to make such changes or it is complicated and needs a lot of time?

 

[PKSRoman]

Just for fun I was thinking about some more radical solutions. Likely impractical but I will think about these some more.
-Building a military unit could cost a population point.
-Each military unit consumes food

Very, very bad idea in my opinion.Because if you are warmonger and build big army - you literally stop your population from growth what in long term will be your downfall (because you usually dont win a war in 20 turns)
Also how to program AI to properly maintain that?
What about "Nemezis" and constant fight each other? esecially when both of you have 20 cities and no, you cannot concquer them without big slowing down your economy and all

Also i find Minor bug - if you and another civ have "Slavery" civic and somehow on teritory of second civ you capture a slave (lets say slave rebellion in his town,its your turn and you attack these slaves for free EXP enslaving them again) - you can preview what another civ build using button "Sacrfice slave" - as you would normally do in your city. Using that - you can just send 7-8 slaves (in ancient era) and use them as free town view of another country (No spy points spending)

 

[AllTheLand]

- The stock exchange tech has global trade has a requirement, but it also requires classical economy which requires global trade already. Glancing at the tech tree in the Renaissance, the tweaks that have been made around tech requirements, tech names and tech effects appear promising. I will have to play more to comment more.

Very true; it's quite hard to prevent a city from being reinforced. I don't really see any elegant solutions to that, that's fundamentally how Civ 4 siege combat is built.

Other games that somewhat manage it have the besieging units on the tile with the siege target, and the garrison being abstracted off the strategic map. I don't see any elegant solution for it either, as it touches aspects of the game that are too fundamental.

The closest I can imagine are some form of zone-of-control that could help limit reinforcements without a full surround, but that would introduce major complications, and the benefit is dubious.

Hmm. If the goal is to prevent early rush of an AI, what if AI's started with a weaker version of the immobile defender barb cities get in world maps? As it is, the AI is its own worst enemy because they fail to use what you give them for defense for that purpose. A unit that can't be used for any other purpose than defending its initial plot may be an appropriate solution.

I had been considering similar ideas.

The only buff I can see making sense is providing a minimum defense (10%?), even after siege units have their turn. No catapult or trebuchet perfectly eliminates every advantage a defender has, after all. Even rubble is a cover of sorts.

I could see that make sense too, although some defenders can get enough first strike bonuses to be oppressive even with defenses at 0.

Support bonuses lifting units that have 0 first strikes to get 2 or 3 first strikes are not so bad, but protective leader + archer on a hill can be nightmarish when it gets aid bonuses from other archers and recon units.

I wasn't really able to fix the underlying cause of CTD, as it seems to be something really obscure in settler AI logic (I think AI wanted to settle on the same landmass and got kind of torn up between going to the target tile by boat or by land).

In which function do you think the issue happened?

I'd say it's actually pretty accurate, historically speaking. Before modern era, populations did recover surprisingly (by our standards) quickly. Even Genghis Khan could only make a dent in real population numbers for a couple of generations, and he tried hard.

I would agree, I think that the disconnect comes from the difference between the timescale of the general civilization progress and the timescale of the in-game wars. A war between two civs that would realistically last a few years can last three hundred years based on the in-game calendar.

You can't have slavery before Warbands are unlocked. Warbands come with Bronze Working, which is a requirement for Mining, which unlocks Slavery.

That's not true anymore since you moved Warbands to Weapon smithing.

 

[AspiringScholar]

So, I decided to shelf the new save and finish it relatively soon with an SVN revert, but to go ahead and experiment with the new early game unit balance now, while conversation about it is still fresh. On the note of that, since English units were just repolished in recent commits, I also thought I'd make a few quick comments along those lines for the same reason:

- The English frigate has white masts.

- The "Royal" Canon for England appears to be a siege piece following a Vallière template, not a field gun. While I understand that the "cannon" unit functionally represents several classes of siege weapons in use during its era, it seems that a field piece is more appropriate to represent with a mobile unit, especially considering that emplaced artillery already separately exists (though admittedly only for defensive purposes in siege warfare, and I don't think we have any distinctive units for it either). I may be ignorant and this is actually emblematic of field pieces, but the unit model resembles a 24 pound Vallière, which would have been notoriously unwieldy and immobile for battlefield purposes. As always, I'd enjoy learning more if I am mistaken about this.

- The English man-at-arms and foot knight look exceptionally cool. The houndskull visor is intimidating and ornamentally "late-medieval," while I am also surprised that I believe the Bardiche is the only other halberd-wielding unit in the whole game.

Yup. In reality, a sword would likely not even be the second weapon of choice in most situations for a knight.

- The English knight, however, does still wield a sword at any rate... Intentional?

- The English ship of the line has an insignia painted onto the sails. Was this still in practice during the reign of George III when the Age of Sail was in its culmination?

- And, finally, a gameplay-contingent question from a stray thought that occurred to me: how about disabling the ability to heal ships in the ocean? After all, it both doesn't really make sense (as mid-sea repairs were always rudimentary and done with stocks and supplies brought aboard already, otherwise fully dependent on access to fresh materials on a coast somewhere) and also too convenient in gameplay, where you can easily sail off into the ocean in a random direction unlikely to be pursued. While my previous request to have early ocean-going sailing ships have a risk of sinking at sea was refused on the basis of how it could disrupt AI pathfinding, this would somewhat make exploration harder, as retreating from pirate-infested coastlines in the New World and then just healing in the ocean for several turns and then returning would no longer be possible.


EDIT: The axeman appears to have been moved to weapon smithing anyway? Surely that is an oversight, given the comment that they were deliberately supposed to stay with bronze working.

 

[Sathar]

It indirectly does, by making them hate you less when it's the case. But if they are already plotting to break free from you, then "thanks for removing direct oversight on us, it'll make it so much easier to organize a rebellion against you!". Since we're talking about ancient eras specifically, I'd say broader autonomy actually made it easier to break away - it's not "giving in to the people's demands" as there is no concept of a nation-state yet, but rather "allowing a local lord freer rein", which, depending on their personal views, may well be a recipe for them breaking away.

Perfectly reasonable.

All doctrines generally focus on one line of units. There are no air-specific doctrines, and I might actually add something for that, but overall they are deliberately limited in scope.

Also reasonable. It's just a little jarring since the splash picture for Mass Production is literally a hanger full of airplanes

 

[pecheneg]

"Civil Service" is udeful due entire game, so why do not make "Aristocracy" similiar, but with diminishing effect in each new era

Probably because there is an example of a combination of feudal liberties with the monarchy, and it is called "Poland". That is, a country that disappeared from the map because that it could not even maintain the technology level of the "powder empire".

building2 (RENAISSANCE): bonus to culture (+1) in renessaince era and bonus to defense (building1 obstolete)

Actually, it should be the other way around. Because
1. Cultural dominance in the 15th - the first half 17th centuries belongs to the least feudal countries. Italy and Holland. The The French "cultural occupation" of Europe takes place exactly when classical feudalism in France has already been successfully suppressed by the Old Regime.
2. As for defense, then.
a) The feudal aristocracy actually has a very vague national identity. As a result, going over to the enemy's side or organizing a feud at the most inopportune moment is a kind of gold standard. b) Feudal castles lose their military importance. At the same time, the construction and maintenance of bastion fortifications with an artillery park is too expensive for the average feudal lord, and at the same time the feudal regime prevents the state from concentrating the necessary resources.

, In combo Autocracy + Plutocracy you have tyrant who depend on mightiest or most wealth -proto "Nobility" and more advanced system is Monarchy with their nobles, private local armies and stuff.
“We do not rule how dare you question the authority of the Ruler? all your sorrows and worries are the fault of the King / Tyrant”.
Oligarchy in a nutshell.The King/Tyrant reigns but does not rule.A group of powerful and influential people rule.Unofficially.

Interesting fantasies. But, actually, according to definitions (from Socrates to modern science), plutocracy is a system where the right to hold public office is limited by a high property qualification. That is, monarchy + plutocracy is, for example, England before the reforms of the 30s of the 19th century, where the number of voters was 2% of the population.

(Oligarchy sounds better for me)

The oligarchy is just the "power of the few." As the history textbook and any news site informs us, THE RULING GROUP CAN BE FORMED NOT ONLY ON THE BASIS OF PROPERTY.

 

[TheBirdMan]

Generally speaking, AI evaluates buildings based on their merit. But if you want to nudge them to like something more or less, that's the exact tag you need to use. You'll have to experiment with the exact numbers to see what's effective, might be up to tens or even hundreds.

Thanks for that info.

I had already changed my settings here to 3, but that might not have worked. Now I changed it to 100 - and that surely works.

Not to the point where the AI is putting everything aside to build palisades, but the palisades have definitely gained attention (I've replayed the turns from 2520BC to 2380BC to see if it showed any difference). So from only 1 civ having 1 city with a palisade and none of the 21 AI's were building on more, it looks quite different now (see screenshots).

Spoiler Palisades in AI-land :

A good start for everyone is, in my opinion, essential for a good game, so I would recommend that in the upcoming version and forward, a value is set to "what is needed", so that the AI prioritizes the early defense with a palisade.


Edit: I see I forgot to undo/reduce the changes I made for <iDefense>. I have this <iDefense>45</iDefense>, the default is <iDefense>25</iDefense> - so I try with a value in the middle <iDefense>35</iDefense>. And no changes for the Walls.

 

[pecheneg]

"Royal" Canon for England appears to be a siege piece following a Vallière template, not a field gun. While I understand that the "cannon" unit functionally represents several classes of siege weapons in use during its era, it seems that a field piece is more appropriate to represent with a mobile unit, especially considering that emplaced artillery already separately exists (though admittedly only for defensive purposes in siege warfare, and I don't think we have any distinctive units for it either). I may be ignorant and this is actually emblematic of field pieces, but the unit model resembles a 24 pound Vallière, which would have been notoriously unwieldy and immobile for battlefield purposes.

In fact, this is a typical "saker." That is, it is a field gun with a caliber of less than 6 pounds (5.25). Just guns "up to Griboval" are, in principle, "abnormally" long-barreled compared to guns familiar from the Napoleonic wars, etc.
For, firstly, the normal technology of drilling the cannon barrel channel (horizontal machine) appeared only in 1734. Before that, the trunks were overwhelmingly cast with a ready-made channel. With the precision of the casting, it was all very sad.
Similarly, everything was very bad with standardization, even in cases where it was possible to carry it out.
As a result, "individualistic" barrels met with "independent" ammunition. The gaps were large, and the efficiency of the gun was low. As a result, the only option to achieve normal ballistic characteristics were large charges of gunpowder (from where the thick breech) and long barrels. At the same time, due to the unpredictability of the relationship between the barrel and ammunition, the range and length of the barrel had to be done with a margin.
Valliere received drilled guns and standardized them, but the design remained almost the same. There is a rational explanation for this (he wanted to preserve the siege capabilities). However, among Valliere's brilliant ideas is the complete eradication of cap loading. Therefore, it is quite difficult to determine where Jean Florent's rationally stupid ideas end and where his irrational stupidity begins.
Then, yes, the Lichtenstein and Griboval systems appeared and the guns acquired their usual appearance. At the same time, the siege artillery retained almost the same design – precisely because it still required large charges and high core velocity / long barrel.

 

[shkitur1]

...

- And, finally, a gameplay-contingent question from a stray thought that occurred to me: how about disabling the ability to heal ships in the ocean? After all, it both doesn't really make sense (as mid-sea repairs were always rudimentary and done with stocks and supplies brought aboard already, otherwise fully dependent on access to fresh materials on a coast somewhere) and also too convenient in gameplay, where you can easily sail off into the ocean in a random direction unlikely to be pursued. While my previous request to have early ocean-going sailing ships have a risk of sinking at sea was refused on the basis of how it could disrupt AI pathfinding, this would somewhat make exploration harder, as retreating from pirate-infested coastlines in the New World and then just healing in the ocean for several turns and then returning would no longer be possible.


EDIT: The axeman appears to have been moved to weapon smithing anyway? Surely that is an oversight, given the comment that they were deliberately supposed to stay with bronze working.

I can add that in my opinion ability to heal on coastal tiles should have only sailing ships. Iron ships can be repaired only in cities (maybe forts also). I can hardly imagine modern battleship taken damages in battle and repearing itself just near coast or in open sea. In that case promotion 'hospital ship' (promotion for additional heal for ships, I dont remember exact name) become unusable.

 

[shkitur1]

Hmm. If the goal is to prevent early rush of an AI, what if AI's started with a weaker version of the immobile defender barb cities get in world maps? As it is, the AI is its own worst enemy because they fail to use what you give them for defense for that purpose. A unit that can't be used for any other purpose than defending its initial plot may be an appropriate solution.

I put one barbarian fort in every AI capital on Huge earth map, because in another case barbarians elimanate half of all civs before reaching ancient era. especially suffer civs living lonely far away other civs: vikings, russian, india, mongols, china, khmer, korea, mali. they die at the start of the game in 100 % cases.
also I give barb forts high lvl but low xp. It prevent forts from getting promotions from battles with barb and other AI.

BeginUnit
UnitType=UNIT_BARBARIAN_FORT, UnitOwner=24, (Joao)
Level=20, Experience=0
FacingDirection=4
UnitAIType=UNITAI_FORTIFICATION
EndUnit