Advanced Civ

Oh, right; no mystery then. 😕 Probably, since keldath did all the work, my memory of the matter has been (especially) transient.

That rather looks like a comparison between Emperor and Immortal. Pretty sure AdvCiv never had a free AI worker at Emperor. That would make a big difference, I expect. Here's a diff between 1.03 and 1.11: compare/1.03...1.10
Bleh, won't link properly to the handicap info XML. A gist, then: gist.githu.com/.../revisions
No major changes to Emperor according to that.
 
Thanks for looking into it. We must've tinkered with it years ago to create a midpoint between emperor and immortal years ago.
 
I think I discovered a bug. When I was the first to research Liberalism, I got the dialog box to select a free tech. But if I am currently researching the tech I select to get for free, Astronomy in this case, then I don't receive the free tech. If I switch my research off of Astronomy before clicking on it in the free tech dialog, then I receive it. If I both switch the tech I'm researching at the start of the turn, and select that tech in the Liberalism dialog, I still don't receive it. Attached a save of the turn right before I discover Liberalism.
 

Attachments

Thanks. It's this Git commit that introduced the issue, specifically the
eNewActiveResearch = NO_TECH;
toward the end. I hadn't been aware that the free tech popup involves the same widget click handler as the jump to Civilopedia and the tech bar on the main screen. I've pushed a bugfix to GitHub. Sounds like an easy issue to run into, but, then, it's been already in the release version (of AdvCiv and of Taurus) for 1.5 years, so I guess I won't hurry another update. And at least it's easy to work around with autosaves if one realizes that the current tech causes the problem.
 
Thanks for the great mod! :trophy2: It improves the AI by a lot and does have some very reasonable balance changes. Clearly a lot of work has been put to this mod and the results are great.

Although I am generally pleased with the mod I find some changes by it to be questionable and think that there are possible improvements, so I post this feedback. I also post how to perform changes (mostly reverts) to the mod which I prefer, so any like-minded individuals can follow suit. I color my changes to the mod in light green color. I played on version "1.10".

Hereditary rule has been changed to give 1 happiness per 2 military units in a city and 25% extra happiness from resources. This makes happiness levels very dependent on resources, which puts player in a very difficult situation if land doesn't offer many happiness resources. For example in one game my empire of 6 cities covering large land area had only one happiness resource (wine), so I was completely gimped happiness-wise. Happiness from buildings is not a very good alternative, because they usually don't give much happiness, often depend on happiness resources themselves and are quite costly. Hereditary rule offers a way of getting happiness regardless of resources, so it can be used in every game as a fallback if needed. It has good synergy with building garrison units as they give both defense and happiness when player uses hereditary rule.
In manual it is written that "The BtS ability is fiddly, encourages players to train lots of Warriors and never upgrade them (or generally to train more units than needed), makes it too easy to grow cities without buildings and is a poor fit flavor-wise (would fit for a military dictatorship". Actually it is usually a good idea to train archers or other good city defenders as they can additionally defend the city. Outdated units can become valuable in future too as they can be upgraded when needed. Growing cities is usually difficult and buildings are often not viable alternative. Generally I can't understand the rationale for nerfing hereditary rule. I haven't seen anyone say, that it is overpowered. People usually use slavery and whip away unhappy population, but I prefer to run hereditary rule and grow my cities big. It is an alternative strategy, which have been nerfed for no apparent objective reasons.
I reverted changes to hereditary rule by changing in "Assets\XML\GameInfo\Civ4CivicInfos.xml" CIVIC_HEREDITARY_RULE iHappyPerMilitary 1 -> 2 and iLuxuryModifier 25 -> 0.

As I described in here unhappiness from emancipation should be halved by changing in "Assets\XML\GameInfo\Civ4CivicInfos.xml" CIVIC_EMANCIPATION iCivicPercentAnger 400 -> 200.

Revolt mechanic is very frustrating, forcing player to keep in conquered cities disproportionate military garrisons compared to the city sizes in order to avoid chances of city revolting. For example in noble difficulty game at 880 AD for cities of recently destroyed enemy:
4 pop city with 9 power (longbowman + archer) and one previous revolt has 4.7% chance of revolt needing +6 power for suppressing.
8 pop city with 12 power(2 longbowmen) has 9.3% chance of revolt needing +31 power (4 macemen!) for suppressing.
4 pop city with 14 power (macemen + longbowman) has 5.3% chance of revolt needing +11 power to suppress.
Next turn 8 pop city revolted having 2 turns of revolt with 29.9% chance to decrease turns, which means on average 6 real turns of revolt.
Even 10% chance of revolt per turn has 47% chance of revolt per 6 turns and 5% chance of revolt per turn has 40% chance of revolt per 10 turns, which means that pretty much total suppressing is necessary for safety. In case of revolt city doesn't produce anything (so can't increase its culture too) and loses culture from all its tiles which can be very dangerous because allows enemy units to use roads on those tiles (so they could easily attack defenders damaged by revolt). I can't be expected to keep 3 macemen in 4 pop city to eliminate chance of revolt potentially disabling entire city and allowing enemy units free roaming.
There is unhappiness penalty for foreign culture in city (for example 4 unhappy faces for that 8 pop city I mentioned earlier) in addition to needing to fight against culture of other AIs in order to keep control of tiles which is enough to motivate player to produce its own culture. There is no need for this extremely punishing and unfun revolt mechanic. Attacker is disadvantaged in Civ IV combat already, needing to have numerical/technological advantage over defender in order to successfully conquer. Having to keep entire armies in conquered cities to suppress revolts is just not feasible. 2 city garrison units from current era should be totally enough for keeping order in city of any size.
I changed in "Assets\XML\GlobalDefines_advc.xml" REVOLTS_IGNORE_CULTURE_RANGE 1 -> 0, which removed revolt chances from cities I mentioned earlier, making gameplay much more pleasant. I suggest making it the default or heavily reducing needed unit power for suppressing revolts. Changing REVOLTS_IGNORE_CULTURE_RANGE has a side-effect though, that decreasing occupation timer becomes too easy, usually being 100% chance for decrease.

Mod makes it so that workers have a 50% chance of being destroyed on capture. I found this to be unnecessary mechanic, because workers are usually needed for transforming captured lands according to the wishes of conqueror and this change only makes that more troublesome. AIs seem to have on average around 1 worker per city, which is usually too little (I prefer 1.5 per city although it is okay to have less late game). I often felt that I captured too few workers from conquered AIs, making building improvements and roads/railroads take too much time. Losing 50% of workers is also bad from the perspective of the conquered civ, because he can't capture them back.
I reverted this change by changing in "Assets\XML\GlobalDefines_advc.xml" BASE_UNIT_CAPTURE_CHANCE 50 -> 100.

I couldn't nuke an enemy city because it had 13% of my culture. I don't understand why should my culture (especially so low) in enemy city affect my ability to nuke it. Especially when I could raze cities regardless of culture.
I reverted this change to BTS behavior by changing in "Assets\XML\GlobalDefines_advc.xml" CITY_NUKE_CULTURE_THRESH 10 -> 101.

When AI is attacked, he should leave minimal garrisons in his cities, collecting rest of units into giant SOD and use it against the attacker (preferably attacking first in order to benefit from collateral damage). Currently they split their units, often even moving some units away from city I am currently attacking, which makes it easy for me to take that city and kill defenders effectively, which leaves even less units for the AI to defend next cities against me. Same happens against next city, etc. Often I am worried when approaching a city with many defenders but to my relief and surprise AI moves many of them away, letting me easily take the city. With each lost city AI becomes weaker by losing commerce, production and units. It is like AI performs divide and conquer tactic against itself, helping me by a lot.
If it is too difficult to make AI use the SOD tactic I described earlier, then at least avoid retreating units from cities, because it is not worth it to lose the cities and land. What is the point of AI losing all his cities except the last one, while retreating to that last city? I could just make peace at that point, leaving AI to huddle in his last city while I gain the benefits from having all his other cities.
In here Napoleon keeps 5 units defending Paris (easily crushed), while there are 9 units safe in Marseilles. Wouldn't it be much more logical for him to defend Paris with all these 14 units?
1736035793025.png


I gave AI ICBMs in worldbuilder but it didn't use them against other AI who had captured many of cities and was about to vassalize him. On couple more attempts I managed to get AIs to use 1-2 ICBMs out of many I gave them, which is weird, as they are clearly not taking full advantage of them.
For example AI used 2 ICBMs out of many against my border city, reducing its garrison to 2 macemen, super medic and worker, but didn't use any more ICBMs or invade with conventional units. After I used ICBMs against all his cities, he responded with one more ICBM against my capital. Only when I once more used ICBMs against all his cities, did he respond with attacking most of my cities with most of his ICBMs (although used 1 ICBM against each city, while should have used 2 for more effect)

AI (Napoleon from Southwest) settled city (Marseilles) behind both my and Suryavarman's (Jayendranagari) borders. He sent settler with archer through my land, even using my roads. Is this normal?
AI settled city behind my borders.jpg


AI often doesn't remove jungle from ivory even after has built camp there, so that tile unnecessarily loses 1 food.

Once AI told me that he has to cancel the deal of one resource vs another, but when I offered the exact same deal again to him, he agreed.

AI offered me 5 gold for dye, 5 gold for incense and 9 gold for both, but when I made deal for one of them, the AI didn't want the other anymore, saying that he would have nothing to gain. It is weird, that I can sell both of them together, but can't sell them both individually.

AdvCiv doesn't calculate commerce bonus of financial trait during golden age correctly. In manual it is said, that financial trait adds 1 commerce to tiles with a natural yield (terrain, feature, river and hill) of at least 2 commerce or a total yield of at least 3 commerce.
With financial trait during golden age:
1. Riverside grassland/plains tile without improvement has 2 commerce (1 from river and 1 from golden age), with cottage has 3 commerce (+1 from cottage, but should have +1 from financial bonus too because total commerce is 3), with hamlet has 5 commerce (+1 from growing to hamlet and +1 from financial bonus)
2. Non-riverside grassland/plains with cottage has 2 commerce (1 from cottage and 1 from golden age), with hamlet has 3 commerce (+1 from growing to hamlet, but should have +1 from financial bonus too because total commerce is 3), with village has 5 commerce (+1 from growing to village and +1 from financial bonus)
Apparently AdvCiv doesn't include golden age commerce bonus into total yield before deciding whether to add financial bonus. This contradicts statement from manual "or a total yield of at least 3 commerce" and unmodded game (in it any tile with at least 1 commerce gets +1 from golden age and then +1 from financial trait too (because total yield is 2)).
This is a problem for my work-in-progress mod for automatic population assigning because it doesn't know whether the tile with 3+ commerce has financial bonus already applied or not during golden age, which affects calculations for potential future extra value of cottage with financial trait. Currently it uses logic of "If cottaged tile currently gives less than 3 commerce and player has financial trait, then add 1 to potential commerce in anticipation of financial bonus", which doesn't work properly during golden age due to aforementioned bug.

Sometimes on Pangaea I start in a position where I am immediately boxed in by nearest AI (his first city covers entire land bridge) with quite small peninsula for myself. This is very unfair, immediately leaving me in a disadvantegous position. I think that on Pangaea and Continents it should be guaranteed that no player/AI starts boxed in. Land bridge should be wide enough to let me quickly place a city there to guarantee access to the rest of the map. If it can't be guaranteed during map generation, then try to detect this situation and regenerate map in that case.
In this game Washington immediately boxes me in:
1736034509501.png


Sometimes on Pangaea one civ is placed on another landmass, much smaller than the main landmass, which severely handicaps him and gives civs on main landmass one less civ to interact with, reducing variety and fun in the game. Primary reason for playing on Pangaea is to play with all civs on the same landmass, giving everyone possibility to expand and interact with each other. Such isolated starts work against the main characteristics of Pangaea maps and so should not be generated.
In this game on Pangaea Alexander was spawned on his own landmass, where he built 5 cities and never interacted with anyone, until I won domination victory in AD 1640:
1736034865601.png


There appears to be too few grassland tiles and food resources on maps, which makes finding good cities (Heroic Epic, etc) too hard. Especially all starting sites should be very good in order to guarantee fair gameplay, but currently they are sometimes very bad. I saw in unmodded game starting site with 17 surplus food thanks to 3 seafood, etc while in AdvCiv games I often see starting sites which don't have enough food to work all their tiles, like this:
Boudica only plains start.jpg

I reverted some changes to resource placement by changing in "Assets\XML\GlobalDefines_advc.xml":
1. SUBLINEAR_BONUS_QUANTITIES 1 -> 0. Not sure whether had any effect, but should result in somewhat more resources on maps by using unmodded resource placement.
2. NORMALIZE_STARTPLOTS_AGGRESSIVELY 0 -> 1. Great effect! Now basically all starting sites are good like in unmodded game, usually having 8+ food surplus and often being great Heroic Epic cities or great person farms (depending on whether are inland or coastal).


Currently when building an improvement on a tile where there is forest and when building that improvement causes forest to be chopped down, then forest will be removed (and hammers added to the city) at the end of building an improvement. I think that it is better to get hammers as quickly as possible so in such cases I always first queue chopping of forest and then building an improvement, which gives city hammers sooner. I would like that this would be done automatically, so when I would order building of an improvement on a forested tile, then chopping of forest would be done first.

It is weird that I have to temporarily decrease science slider in order to get positive gold surplus for offering gold per turn to AI in negotiations, even though I have large amount of gold and can afterwards immediately increase science slider back to previous level.

Why does one religion almost always die out in a city when another one spreads there? Currently all my cities have either 0 or 1 religions. In real life and in unmodded game there are often multiple religions in the same city.

In random event of fire destroying the forge in a city I picked the option of paying gold in order to preserve the forge. Forge was preserved, but game displayed incorrect message of "A fire has destroyed the forge in the ...".
 
Do victory conditions affect how the AI behaves? I began to notice that the AI often does not finish off an opponent who has one city left, and simply makes peace with him, does not even take him into vassal dependence. In general, it seemed to me that the AI has become much more peaceful
 
Do victory conditions affect how the AI behaves?
Yes, the AI counts its progress toward each of the victory conditions in "stages" from 0 to 4, and those stages affect its behavior in a bunch of ways. If a victory condition is disabled, then the progress toward that victory is stuck at stage 0.
I began to notice that the AI often does not finish off an opponent who has one city left, and simply makes peace with him, does not even take him into vassal dependence.
That's probably not a new phenomenon. I think, in the war utility calculation, a single (perhaps small) city is sometimes not considered worth continuing a war for. This may often be good thinking, especially when that city has a large garrison. Moving through enemy territory and bombarding defenses causes supply costs, perhaps busies units that could otherwise suppress revolts. Enemy units may still pose some threat to workers. And completely defeating a civ doesn't generally provide any special reward (as it doesn't remove all its culture in the mod). Also, despite my efforts to let the AI handle multiple war plans, having just one war at a time tends to work best for the AI. All that being said, the AI might also sometimes behave in this way when the final city would be somewhat valuable and quick and easy to conquer. And these apparent acts of mercy seem to be a little puzzling – it's been reported a number of times –, which isn't good. I'd need a concrete situation though, preferably with a savegame, to do something about it.
In general, it seemed to me that the AI has become much more peaceful
Well ... noted. Lanstro was saying that too, but now it seems to be mostly the result of his difficulty settings. Let's see if greater peacefulness gets corroborated. I don't think I've made AI changes in that vein, but I also haven't tested v1.11 much, so maybe something's not working right.
 
@Yorok: Thanks for your feedback. I guess it's no surprise that I don't intend to just undo some major balance changes. I reckon that AdvCiv is more hostile toward some established patterns of play than one might expect or welcome from a mod focused on AI changes. I'm glad that you've found ways to adjust things to your liking. Perhaps, on that note, to get this out of the way:
Changing REVOLTS_IGNORE_CULTURE_RANGE has a side-effect though, that decreasing occupation timer becomes too easy, usually being 100% chance for decrease.
The occupation countdown has been tied to revolt chance. So I think one really needs to revert those changes as well, going back to longer countdowns that don't depend on the size of the occupying force. That should be possible through OCCUPATION_COUNTDOWN_EXPONENT and OCCUPATION_TURNS_POPULATION_PERCENT. I'll add a note about that to GlobalDefines_advc.xml.

Hereditary Rule: I'm not happy with the revised HR effect either. The main reason I felt that a change was necessary (which the manual also states) is that the BtS effect gives the AI civs too much happiness, which causes an oversupply of luxury resources offered for trade for a long stretch of the game. So the change to HR is a consequence of letting the AI offer and acquire resources more rationally.
People usually use slavery and whip away unhappy population, but I prefer to run hereditary rule and grow my cities big.
Perhaps a good time to point out that the abundance of food resources on the map – that you propose to restore – contributes to the pressure to either sacrifice population, or to grow cities large already by the Medieval era, or to keep churning out settlers and workers.

Revolt chance (in non-border cities): It's supposed to slow the pace of military expansion, or really to discourage an alteration between long (boring) phases of peace and long (boring) phases of war in favor of smaller, perhaps opportunistic campaigns. Having recently destroyed an enemy, as you write, would seem like a reasonable time to take a break. Although I do think that largely ignoring revolts for a while is possible too. Without a nearby foreign city, revolts can't cause a flip, they only delay the city's economical integration and exploitation.

As for your specific numbers, I don't know how focused the conquered civ had been on culture or if city religions or even sacrificed population play a role. Would really need a savegame. (I don't think the changes you've made will break save-compatibility.) Though I'm not sure if there's a point in getting into details when you find the whole idea largely uncalled for or counterproductive. It's of course also been discussed before, the most thoroughly perhaps around this post. (This isn't meant to prevent more discussions; just saying they may not be so interesting to me.)

AI city evacuation: In case that this makes no sense to you at all, I'll refer you to change ID 139 in the manual for a general explanation. I've mostly found this AI change to work well and to the advantage of the AI. I can't tell from the screenshot whether the AI stands a chance against your stack one way or another. Would need a savegame to make adjustments. That would also be helpful for the matter of ICBMs.

AI (Napoleon from Southwest) settled city (Marseilles) behind both my and Suryavarman's (Jayendranagari) borders. He sent settler with archer through my land, even using my roads. Is this normal?
Yes, K-Mod changed that. Not sure how deliberate this was; the K-Mod changelog doesn't mention it and the Git commit message only speaks of "pathing efficiency improvements." Also not sure how deliberate the original behavior was. The name of the pathfinder flag "MOVE_SAFE_TERRITORY" sounds like safety may have been the main concern. It gets used for most worker moves too. Although moving civilian units through foreign territory is generally not risky. I've seen posts speculating that Firaxis didn't want players to be on guard against sneaky AI settlers. Imo players who cut off AI access to unclaimed land tend to be quite aware of that fact, and I don't think it's much of a burden to watch out for settlers on the move for a while. Conversely, the AI just politely ignoring good land feels rather toothless to me. Yes, the player can just cancel Open Borders, but that comes at a diplomatic and economic cost. So I like the K-Mod change. One needs to be aware of course. But, then, I don't think it's all that widely known that the BtS AI will assuredly not traverse foreign borders to settle.
Once AI told me that he has to cancel the deal of one resource vs another, but when I offered the exact same deal again to him, he agreed.
This used to happen more frequently. It can, in theory, always be a result of the cancellation decision happening on the AI turn and the renegotiation on the human turn; i.e. something could've happened during the intervening AI turns. So long as it's rare, I'm not worried. Though I would want to take a look if you have a savegame. It would have to be from one turn before though, i.e. (just) before the AI decides to cancel the deal. This may not be workable when using slightly different versions of the mod because the AI decision to cancel is randomized and even small, unrelated differences in the mod's rules could interfere with that roll.
AI offered me 5 gold for dye, 5 gold for incense and 9 gold for both, but when I made deal for one of them, the AI didn't want the other anymore, saying that he would have nothing to gain. It is weird, that I can sell both of them together, but can't sell them both individually.
The trade evaluation works best when considering one resource at a time. The AI will accept two luxury or two food resources at once (no more if I remember my own code correctly), but it's a bit of wild guess as to whether that much extra happiness/ health will be useful. I suppose, in your example, the AI thought it might need 2 luxuries but thought better of it after receiving the first.
It is weird that I have to temporarily decrease science slider in order to get positive gold surplus for offering gold per turn to AI in negotiations, even though I have large amount of gold and can afterwards immediately increase science slider back to previous level.
A similar behavior (that I've introduced) is ... I'll just quote the loading screen hint that I've added about this: When asked to make a trade offer for a resource, the AI will try to offer gold per turn if your income is negative, and a resource otherwise. Adjust your gold slider in order to get the offer you want.
Maybe keeping that slightly clunky restriction in place makes it a little more likely that players will also figure out that the income determines the AI proposals. And I'm not seeing that restriction implemented in CvPlayer::canTradeItem in the DLL (where a corresponding restriction on lump sums of gold is enforced). Might actually be imposed by the EXE and difficult to work around.
Why does one religion almost always die out in a city when another one spreads there? [...]
I think how it should work is a 100% chance of outright success (adding the Missionary's religion and removing none) minus 10 percentage points per religion already present. In the case of a failure, one religion is chosen (partially at random) for removal. That could be the religion of the Missionary – in which case the Missionary fails without spreading or removing anything – or one of the existing religions. In that last case, the Missionary succeeds at spreading its own religion after all - and removes one other religion. So it should not almost always remove a religion. It's a K-Mod mechanism.
Mod makes it so that workers have a 50% chance of being destroyed on capture. I found this to be unnecessary mechanic, because workers are usually needed for transforming captured lands according to the wishes of conqueror and this change only makes that more troublesome.
I find that there isn't usually much to be gained by replacing lots of improvements in captured lands. Not having the workers for that should make it an easier decision not to bother. Which isn't to say that the reduced capture chance was a necessary change.

Sounds like the mod is not working as intended in these two cases. I'll look into it:
AI often doesn't remove jungle from ivory even after has built camp there, so that tile unnecessarily loses 1 food.
AdvCiv doesn't calculate commerce bonus of financial trait during golden age correctly.

In random event of fire destroying the forge in a city I picked the option of paying gold in order to preserve the forge. Forge was preserved, but game displayed incorrect message of "A fire has destroyed the forge in the ...".
I guess I haven't paid enough attention to events (usually I disable them anyway) to decide if that's unusual. From a quick glance at the code, it looks to me like one more way in which events were carelessly implemented – displaying a text that doesn't take into account how the event was handled. I'd only take a closer look if I could do so via a savegame. (Which, again, due to the randomness involved may not be practical.) Minor issues with events aren't generally worth fixing to me.

Sometimes on Pangaea I start in a position where I am immediately boxed in by nearest AI (his first city covers entire land bridge) with quite small peninsula for myself. This is very unfair, immediately leaving me in a disadvantegous position. [...] In this game Washington immediately boxes me in: [screenshot]
In this case, the starting position algorithm should be aware that you'll have little room to expand – given that America could settle a part of the peninsula too. The more difficult case (that I won't be doing anything about) is when the second AI city could seal a bottleneck. It might be that your (shallow-)water access to the main continent and to the numerous islands in the west is being overestimated by the algorithm. But, generally, AdvCiv maps are somewhat crowded, and a fair starting position can't be guaranteed. I wouldn't say that this one is very unfair. Apart from the possibility (opportunity, even, given the short distance between the capitals) of an early war, settling toward the western coast and then on the islands should be quite feasible. Perhaps also some snowy seafood sites above the northern edge of the screenshot? Well ... :undecide:

Maybe the Solid Shoreline option of Pangaea will make such starts less likely. Peninsulas still seem to occur, but a shorter coastline, maybe no islands (not sure if those are affected by this setting), thus less seafood – should make it clearer to the mod that a small peninsula is not an acceptable starting site. You could also try if the original algorithm (much simpler, focused on air distances) suits your priorities better by setting ENABLE_STARTING_POSITION_ITERATION to 0 in GlobalDefines_advc.xml.
Sometimes on Pangaea one civ is placed on another landmass, much smaller than the main landmass, [...]
I had not been aware that Normal-size Pangaea can generate such large minor continents, and even separated by Ocean. Starting that way as a human player would seem like a bad joke after opting for Pangaea. Hopefully, that's exceedingly rare. (The mod's starting position algorithm has a bias against starting human players in isolation.) Personally, even though I respect Pangaea not being a script for players who crave variety, I think 1 out 8 starting on another continent occasionally is fine. The creator of the script (Sirian) clearly didn't feel that way. It explicitly rules out starts on minor continents. AdvCiv ignores that, primarily because it's awkward (implementation-wise) to recognize such special restrictions imposed by individual map scripts. Which means that, by disabling the AdvCiv starting position algorithm, you should get the behavior that you seek.
[...] I always first queue chopping of forest and then building an improvement, which gives city hammers sooner. I would like that this would be done automatically [...]
Somehow this subject seems familiar. I think there's a bit of an open issue (and someone may have pointed this out in this very thread) with invested worker turns not adding up when combining chopping and improving. E.g. when telling a worker to build a Mine on a Forest Hill and canceling that mission right away (the worker moves already having been spent), the one invested turn won't count toward just chopping the Forest. And iirc fixing that seemed too complicated.

So this also means that, if a chop mission is stacked automatically, then the player can't really cancel that mission to get the original behavior back. One turn will already have been wasted. So perhaps we're really talking about taking away the choice to delay the chop at the level of the game rules. Cleaner then to change the effect of building an improvement so that the terrain feature gets removed already midway? Sounds like a headache to implement, actually, and won't interact well with the pre-chop BUG option/ shortcut. But the approach with two missions will allow the chop to be canceled. And, if that's allowed, then it's unsatisfactory because it wastes a worker turn. The stacked mission could just be a BUG option, but then that option will have to be toggled on and off, or some clumsy/ obscure key press will have to enable the original behavior.

To be clear, I don't think it's all that rare that delaying the chop is more efficient. The typical situation that comes to mind is a city with no decent tiles in its inner ring working on a Library to spread its borders. And maybe just one Forest in the inner ring, on a Hill, and that being the best production tile. A single chop won't come close to finishing the Library. Then it's best to build a Mine on the Forest Hill and to keep the Forest alive until the Mine is finished. The delayed chop does seem almost universally pointless for Jungle, however.
Well, I don't think any such changes will deliver a result that is nice enough to justify the effort.
 
Hi folks,
Having read all of the above,
Just wanted to comment that with manipulating the advciv devel file and another one i think,plus changing leader values, got me to play vs very aggressive ai.
That much that in the latest version of my mod i decided to tune it down.
So, the ai can be changes to be less peaceful as said.
 
My 2c on some of Yorok's feedback:
  • making it much harder to pacify conquered towns is a good feature of this mod. Warfare is too easy otherwise
  • for similar reasons, worker capturing shouldn't be as rewarded as it is in vanilla
  • AI city evacuation works pretty well in my experience. Makes a stronger stack to fight later instead of losing chunks of their army in hopeless fights
  • I did see a failure to chop an ivory jungle once or twice too
  • i don't have any major issues with the map generation - I understand Yorok would like lower variance in general, my personal preference is the higher variance that's currently implemented; given I play it as a PvE game I prefer variety to fairness
 
I tested it, set up 29 AI, fast speed, ancient era, donut map (huge), only victory by conquest, and looked at how many opponents would remain after every 100 turns with the game.AIPlay command.

After 100 turns 29 AI
After 200 turns 28 AI
After 300 turns 24 AI
After 400 (you can easily find civilizations with one city) turns 19 AI
After 500 turns 19 AI
After 600 turns 16 AI
After 700 turns 16 AI
After 800 turns 11 AI
After 900 turns 7 AI
After 955 turns 5 AI (Victory under AI control)

How normal is this?
I also attach a link with intermediate saves https://drive.google.com/file/d/1njBQJ1m02pEoKs2w0i00iP-xy-4QfCHZ/view?usp=sharing
 
I guess it's no surprise that I don't intend to just undo some major balance changes.
Indeed, because clearly you thought them through. But people have different preferences and it doesn't hurt that people voice their opinions :). Maybe could some alterations be done if some mechanics are too extreme according to majority? Discussions can highlight possible improvements. It is very good, that you have explained your changes, rationales for them and ways to alter them so thoroughly so that people can modify the mod according to their preferences.
I have to note though, that generally I have been pleased with base BTS, mostly playing this mod because of AI and BUG mod. So any major overhauls of game mechanics are likely to alter things in way that I am not fond of.

The occupation countdown has been tied to revolt chance. So I think one really needs to revert those changes as well, going back to longer countdowns that don't depend on the size of the occupying force. That should be possible through OCCUPATION_COUNTDOWN_EXPONENT and OCCUPATION_TURNS_POPULATION_PERCENT. I'll add a note about that to GlobalDefines_advc.xml.
I think that having countdown depend on size of occupying force is reasonable. Having cities endlessly remain in occupation is not fun. 3-4 turns is reasonable if I have 2-3 decent units in a modest city.

I'm not happy with the revised HR effect either. The main reason I felt that a change was necessary (which the manual also states) is that the BtS effect gives the AI civs too much happiness, which causes an oversupply of luxury resources offered for trade for a long stretch of the game. So the change to HR is a consequence of letting the AI offer and acquire resources more rationally.
Why exactly is oversupply of luxury resources a problem? Usually I am eager to import happiness resources as long as price is reasonable. There should always be civs interested in luxury resources as long as some of them don't have such resources. Price will drop of course. Generally I have been struggling to import resources from AIs, partly because they are too unreliable and can easily cancel trade after 10 turns have passed, potentially forcing my cities into unhealth/unhappiness. This means that I can't depend on imported resources and have heavy motivation to ensure that I have the resources myself ... along with HR :)
HR is incredibly useful also because it allows to shift happiness around. For example if spy forces my city to temporary unhappiness, then I will immediately move units from cities with spare happiness to that city to counter the unhappiness. I can manage happiness levels across my empire very precisely with units. So I am used to running HR till end of the game.

I don't think that this AI luxury resource issue is enough to justify such hardcore nerf to HR, which makes player very-very dependent on happiness resources. I was pretty much forced to change HR mechanics back to BTS ones in that game with only wine resource in order to grow populations of my cities.

Perhaps a good time to point out that the abundance of food resources on the map – that you propose to restore – contributes to the pressure to either sacrifice population, or to grow cities large already by the Medieval era, or to keep churning out settlers and workers.
Without food resources it is very difficult to grow cities. Non-resource farm has very weak yield (1 food), so I only build them when city can't reach 3-4 food surplus otherwise. Food resources make it possible to get much more hammers/commerce out of the land. Especially difficult is to get any hammers as hammer tiles are mostly food negative in the beginning of the game.

Revolt chance (in non-border cities): It's supposed to slow the pace of military expansion, or really to discourage an alteration between long (boring) phases of peace and long (boring) phases of war in favor of smaller, perhaps opportunistic campaigns. Having recently destroyed an enemy, as you write, would seem like a reasonable time to take a break. Although I do think that largely ignoring revolts for a while is possible too. Without a nearby foreign city, revolts can't cause a flip, they only delay the city's economical integration and exploitation.

As for your specific numbers, I don't know how focused the conquered civ had been on culture or if city religions or even sacrificed population play a role. Would really need a savegame. (I don't think the changes you've made will break save-compatibility.) Though I'm not sure if there's a point in getting into details when you find the whole idea largely uncalled for or counterproductive. It's of course also been discussed before, the most thoroughly perhaps around this post. (This isn't meant to prevent more discussions; just saying they may not be so interesting to me.)
To be honest I don't find parking my units in cities for ages until I manage to gather enough culture interesting. My preferred way of playing is to build sufficiently powerful army to completely conquer weakest neighbor while preserving enough of my army to keep steamrolling over rest of the civs from weakest to strongest in perpetual war until end of the game. I don't want to lose steam by having to stop conquest while my units get outdated and enemy builds defensive units.

Currently culture revolt chance seems artificial and disproportionately strong. I pretty much had to keep 3/4 as many units as city has population to quench revolt chance. This is clearly too many from both gameplay and realism sense and makes it unfeasible to keep perpetual war campaign going. Revolting doesn't just delay growth of city. It also allows enemies access to roads in area of city, which can be very dangerous. I looked at how my conquest was stopped by revolt mechanic and thought, that this ain't fun or fair and decided to get rid of it.

War campaign should be stopped by smart AIs managing to defend themselves, not artificial revolt mechanic making me sit my entire army in captured cities for ages. Perhaps making needed power for quenching revolt chance 1/3 of current requirement would be better balanced, so 1 unit for 3 population, 2 for 6 and 4 for 12?

In "Yorok AD-0760" savegame Lakamha revolted next turn which it definitely shouldn't, as it had archer and longbow guarding 4 population.

AI city evacuation: In case that this makes no sense to you at all, I'll refer you to change ID 139 in the manual for a general explanation. I've mostly found this AI change to work well and to the advantage of the AI. I can't tell from the screenshot whether the AI stands a chance against your stack one way or another.
Alright, then it even makes sense. AI had no chance, as I had 42 units going for Paris. Still as a player it was relieving to take Paris that easily. Losing cities with a lot of hammers/commerce is devastating

Generally I would like to see AI building proper SODs (50% siege - currently they build too few siege units) and use them in defensive wars against attackers by attacking first (thanks to roads in own culture) - first with siege, then with rest of units. Usually AI does random attacks with scattered siege units, which is a complete waste because they don't attack immediately afterwards with non-siege units, so I can heal damage off.

The trade evaluation works best when considering one resource at a time. The AI will accept two luxury or two food resources at once (no more if I remember my own code correctly), but it's a bit of wild guess as to whether that much extra happiness/ health will be useful. I suppose, in your example, the AI thought it might need 2 luxuries but thought better of it after receiving the first.
It seems inconsistent, like the AI has randomness in its decisions, not logical calculations. Surely if he thinks that 2 resources are good for him, then it should not change after reciving one of them.

A similar behavior (that I've introduced) is ... I'll just quote the loading screen hint that I've added about this: When asked to make a trade offer for a resource, the AI will try to offer gold per turn if your income is negative, and a resource otherwise. Adjust your gold slider in order to get the offer you want.
Maybe keeping that slightly clunky restriction in place makes it a little more likely that players will also figure out that the income determines the AI proposals. And I'm not seeing that restriction implemented in CvPlayer::canTradeItem in the DLL (where a corresponding restriction on lump sums of gold is enforced). Might actually be imposed by the EXE and difficult to work around.
It is clunky to have to exit trade screen, temporarily alter research rate, enter trade screen, make trade, exit trade screen and alter research rate back to previous level. Also I sometimes forget that I have to do this. And I can't do this at all when it happens during AI turn?
Do you mean by "income determines the AI proposals", that if player doesn't have enough income, then AI doesn't offer some deals, it would otherwise offer?

I think how it should work is a 100% chance of outright success (adding the Missionary's religion and removing none) minus 10 percentage points per religion already present. In the case of a failure, one religion is chosen (partially at random) for removal. That could be the religion of the Missionary – in which case the Missionary fails without spreading or removing anything – or one of the existing religions. In that last case, the Missionary succeeds at spreading its own religion after all - and removes one other religion. So it should not almost always remove a religion. It's a K-Mod mechanism.
Hmm, usually I don't see any missionary. I just see a notification, that one religion has spread to the city while other one has disappeared from it. It seems weird. I would like to have more religions in cities for more happiness from Free Religion, so this behavior irritates me :lol:

I find that there isn't usually much to be gained by replacing lots of improvements in captured lands. Not having the workers for that should make it an easier decision not to bother. Which isn't to say that the reduced capture chance was a necessary change.
I have a lot of work for workers in captured lands, because I usually convert captured cities to workshop-powered production cities. Often I find unimproved tiles. AIs build too many non-resource farms, which need to be replaced with more useful improvements. Building railroads need many worker turns. If I would not have a use for worker, then I would just delete him. I keep at least 1 worker per city, so I have a possibility to quickly change something around the city or rebuild destroyed improvement.

Maybe the Solid Shoreline option of Pangaea will make such starts less likely. Peninsulas still seem to occur, but a shorter coastline, maybe no islands (not sure if those are affected by this setting), thus less seafood – should make it clearer to the mod that a small peninsula is not an acceptable starting site. You could also try if the original algorithm (much simpler, focused on air distances) suits your priorities better by setting ENABLE_STARTING_POSITION_ITERATION to 0 in GlobalDefines_advc.xml.
Hmm, Solid Shoreline indeed gets rid of trap-peninsulas, but as a side-effect makes the map kinda boring :lol: Is there indeed considerably less seafood next to main landmass than with other shoreline types, or am I imagining things? I think that amount of seafood next to main landmass should not depend on islands. I am unsure what ENABLE_STARTING_POSITION_ITERATION exactly does and should I really disable it.

Personally, even though I respect Pangaea not being a script for players who crave variety, I think 1 out 8 starting on another continent occasionally is fine. The creator of the script (Sirian) clearly didn't feel that way. It explicitly rules out starts on minor continents. AdvCiv ignores that, primarily because it's awkward (implementation-wise) to recognize such special restrictions imposed by individual map scripts. Which means that, by disabling the AdvCiv starting position algorithm, you should get the behavior that you seek.
Civ, which starts isolated, is crippled, which is especially bad if it is human and still bad if is AI (less variety in game). In my example Alexander was basically out of the game the entire time. I could disable ENABLE_STARTING_POSITION_ITERATION, but what are side effects?
I will play Continents or Fractal if I want variety. There would be unbalanced too if 1 civ would start isolated while all the rest are on the same landmass.

Why in "Yorok AD-1500" savegame has Peter not sent more of his units (there are plenty of units all around his empire, especially in Verlamion) to his SOD (2 tiles SW of Bibracte)? It ended with disaster like can be seen in "Yorok AD-1530 - Railroad" savegame. Sending more units would surely have let him capture Tolosa and maybe Bibracte.

making it much harder to pacify conquered towns is a good feature of this mod. Warfare is too easy otherwise
Difficulty of warfare should come from smart opponents, not annoying mechanics. What if FPS games would work like this - after shooting 5 enemies you have to do 30 push-ups, 60 squats and 10 pull-ups before getting your rifle back.

for similar reasons, worker capturing shouldn't be as rewarded as it is in vanilla
I find it reasonable that workers come with the land, so I don't have to train new workers at late stadiums of the game (stalls growth of cities).

AI city evacuation works pretty well in my experience. Makes a stronger stack to fight later instead of losing chunks of their army in hopeless fights
Would you behave the same way when AI attacks you? Let him capture most of your cities while gathering your army for the last stand in the last city ... only to discover that AI is not interested in a fight anymore and instead garrisons in cities captured from you.

i don't have any major issues with the map generation - I understand Yorok would like lower variance in general, my personal preference is the higher variance that's currently implemented; given I play it as a PvE game I prefer variety to fairness
I play Pangaea in order to have interactions with many AIs and not have to hassle with moving/fighting over water. Starting isolated on an island or being limited in interactions to only one civ, which has boxed me into a pensinsula from the start does not fit into proper Pangaea gameplay. I would play Fractal if I would want more variety.
 

Attachments

Hello! I have just finished playing my first game with this mod on, and I got absolutely destroyed in the midgame, which is a good thing of course. Love the changes to the AI and mostly like the gameplay/balance changes. However, I've encountered several issues that I would like to clear out:
1) By far the biggest issue I've encountered is when I set workers to "Build Trade Network" automation they for some reason start ruining my improvements on silver, spices and other resources and instead build forts on them. I think they're doing that because those resources are outside my workable tiles, but I still want to have access to them. This is pretty inconvenient in the mid to late game because I'm forced to build roads manually. Is this a known issue, or could I be missing something? I would just like them to build roads and leave my improvement choices alone. Edit: Damn, I'm an idiot. I didn't know that forts still give you access to the resource. Then this change is a good one, of course.
2) Another thing I've noticed is that the map size is smaller in the mod compared to vanilla. I generated a few Large maps in vanilla and the mod and the difference is very noticeable. I skimmed though the manual and found some references to map changes, so that seems intentional. This is not a big deal since the Huge map is more that enough for me, but still I wonder what rationale for that change is? With the default number of Civs it felt like I was elbow-to-elbow with my rivals.
3) Update: found a bug. In Civilopedia leader descriptions are absent, and the message about the function "getCiv" not being defined is displayed.
 
Last edited:
Hello!
Accept my praise for this mod!
One question: is there a possibility to turn off an espionage like in vanilla version? I hate spies! In vanilla i can turn it off before i start the game. Then espionage becomes a culture. But here i can't.
Maybe i can change it in another way?
Or it is necessary to make changes to the mod itself?
If so, can i ask the creator (or anyone who can do it) to add such an option?
Or maybe someone can tell me how to do it myself?
I'd be grateful.
 
Last edited:
I just want to point out that Mercantilism might not be working as intended. There are no trade routes between master and vassal. The description in the manual is:

Only with the master. (And the master can have
trade routes with its vassals – no change.)
Help text says that trade routes with all foreign
civs are blocked. I don't think the exception for
vassal/ master (presumably introduced by
Warlords) is mentioned anywhere.


So it seems it is supposed to be possible to have trade routes between masters and vassals, but in-game this never happens.


Yes, indeed it does actually work. Turns out the cities were reprioritising the trade routes in a weird way that made them lose the foreign route, but when conditions are right, they will still have trade routes to vassal civs.
 
Last edited:
@Rawwwrr: Stepping through the AI code for Conquest victory, I did find a bug that usually prevented the higher stages (3 and 4) of the Conquest victory strategy from being adopted. This has been in the code since AdvCiv 0.97b. However, I don't think this bug has affected the first 400-500 turns of your test much, if at all. I do think it's normal that no AI will win Conquest within the normal time frame of the game on such a large map and on Quick speed – and also normal that Conquest being the only victory doesn't cause the AI to work directly toward the elimination of rivals. Space being disabled likely means a good deal more military production in the late Modern era. And the personality-based weights for victory strategies get redistributed when victory conditions are disabled (in AdvCiv; not in BBAI, which introduced those weights). However, those weights tend to affect mostly the early stages of the victory strategies. For Conquest stage 3, in particular, a civ needs to be either a lot more powerful militarily than the average remaining rival or half of the civs must have already been defeated (by anyone) or vassalized by the civ considering the Conquest strategy. Personality doesn't relax these conditions. Maybe they should in fact be relaxed somewhat when the personality weight is very high.

But, fundamentally, when a victory is such a monumental effort as Conquest on a Huge map, the AI imo does well to mostly ignore that goal and wage war based on economical considerations – until, however late in the game, victory will be a matter of just a few more wars (or a lot of easy ones). I've also run a couple of all-AI games with your settings myself - but with only 18 civs on a Large map (Donut has a lot of land). With the bug fixed, these still ran into the 22nd century. I don't think this is a meaningful measures of the overall aggressiveness of the AI. And I did not see a striking number of leftover civs with just one or two cities. I think there are usually a couple of those around, also in your first few savegames (edit: the long distances on such a large map might make the AI more inclined to end a war before taking every last city; Quick speed also could contribute). If they stick around for a long time, then this could interfere with the half-of-the-rivals-gone condition for Conquest; on the other hand, they'll make the alternative condition based on military power easier to fulfill. Normally, Domination will be the much more relevant victory strategy anyway.

Edit: I'd recommend AI Auto Play (Ctrl+Shift+X) instead of game.aiplay. The latter kills the human civ, if I'm not mistaken.
 
Last edited:
In Civilopedia leader descriptions are absent, and the message about the function "getCiv" not being defined is displayed.
Ugh, this is an easy bug to run into. Perhaps another release will have to come sooner than I'd hoped. Thanks though.
Another thing I've noticed is that the map size is smaller in the mod compared to vanilla. [...] With the default number of Civs it felt like I was elbow-to-elbow with my rivals.
Given the limited tactical abilities of the AI, expanding peacefully - even into mediocre land - is normally more efficient than waging war. So, until space gets scarce, the AI tends to lack good reasons for starting a war. I find it jarring when, on the default settings, there is no warfare until, say, late antiquity – or almost none. Especially with a mod whose biggest feature is a rewrite and expansion of AI decision making about war and peace. On the other hand, I don't want the starting positions to be so dense that human players will often feel compelled to start a war of their own before the Medieval era. Given the variability of maps, I don't think these goals are reliably reconcilable, but the default player counts and map dimensions are my best bet. I first increased the player counts to come close, the map dimensions are fine-tuning. Well, for some map scripts, the size change is more pronounced. I've tried to make the same player counts work for all of them. This only goes for the rather small selection of scripts that AdvCiv contains customized copies of. I haven't e.g. bothered to make Donut smaller. (Maybe I should. I've mainly avoided the more obscure scripts because I don't want them to show up at the top of the menu on the Custom Game screen, – which is implemented out of reach in the EXE. But I've learned that the PrivateMaps folder will get listed before PublicMaps.)

@arcvoodal: I've loaded an old savegame where I had one vassal, converted to Mercantilism, ended the turn to make sure that trade routes get updated, and I got a total of 12 commerce from my vassal according to the Info tab of the Foreign Advisor and was also able to see those trade routes at a few of my cities, including the capital. Seems to work fine based on this quick investigation. I suppose an Open Borders agreement is necessary too. (Not necessary for entering the vassal's borders.)
 
@Nocovic: Thanks. The No Espionage option can be restored at the end of Assets\XML\GameInfo\Civ4GameOptionInfos.xml. bVisible needs to be set to 1. I hope it still works correctly, I haven't been testing it. My aim has been to make espionage (even) easier to (largely) ignore than in BtS, e.g. by moving Great Spy points from the Great Wall to the unpopular Versailles and allowing the espionage slider to be hidden (BUG menu). The game option has the problem of unbalancing Culture victory, and it's weird that e.g. Jails produce culture. Well, pick your poison. 🙂 I rather keep getting asked about the absence of the option than why I haven't improved it.
 
Given the limited tactical abilities of the AI, expanding peacefully - even into mediocre land - is normally more efficient than waging war. So, until space gets scarce, the AI tends to lack good reasons for starting a war. I find it jarring when, on the default settings, there is no warfare until, say, late antiquity – or almost none. Especially with a mod whose biggest feature is a rewrite and expansion of AI decision making about war and peace. On the other hand, I don't want the starting positions to be so dense that human players will often feel compelled to start a war of their own before the Medieval era. Given the variability of maps, I don't think these goals are reliably reconcilable, but the default player counts and map dimensions are my best bet. I first increased the player counts to come close, the map dimensions are fine-tuning. Well, for some map scripts, the size change is more pronounced. I've tried to make the same player counts work for all of them. This only goes for the rather small selection of scripts that AdvCiv contains customized copies of. I haven't e.g. bothered to make Donut smaller. (Maybe I should. I've mainly avoided the more obscure scripts because I don't want them to show up at the top of the menu on the Custom Game screen, – which is implemented out of reach in the EXE. But I've learned that the PrivateMaps folder will get listed before PublicMaps.)
This is a compelling explanation, however, in practice on huge maps with 12 AIs (your default is 16 AIs) I found the war/peace balance quite satisfying. In Ancient there are no wars because the barbarian invasions are too rampant (this is also another thing I've noticed, due to your changes to fog-busting there are so many barbarians on maps with such relatively sparse starts that it feels like you're waging a full-scale war against them). In Antiquity there were still a lot of potential spots for new cities, but only a few really great ones and the AI was rushing them and stealing them from me, so I had a choice of swallowing my pride and settling on subpar locations instead, or starting a war against the AI to reclaim those good city sites.
So, all in all, I found the huge fractal map with 12 AIs the most satisfying way to play this game. With the default number of civs I felt forced to wage war against the AI, there was no chance to win the game otherwise. I still find it a little bit weird that you felt forced to decrease the maps themselves instead of just tweaking the number of civs and their starting positions, thus decreasing the possible maximum size of the map.
 
It's 10080 tiles vs. 10240 for Huge Fractal maps. (Fractal being one of the few scripts that doesn't override the grid dimensions in Civ4WorldInfos.xml.) It's a smaller tweak than adding a 17th player. Besides, AdvCiv also changes the grid dimensions to make the map taller and less wide (ca. 7:5 aspect ratio instead of 8:5). This is to make climate zones feel more authentic and relevant on smaller (than Huge) maps. I don't find it strange to also adjust the area a little. The mod also uses a (very) slightly lower sea level at the Medium setting and the resource and terrain distribution differs a bit from BtS. 16 players is also close to the maximal player count. There ought to be some leeway for players to go above the default density and for Low sea level. In that regard, the limit of 18 civs really is too tight. I would've switched to 31 (i.e. 32 with the Barbarians) by now if that didn't break savegame compatibility.
In Ancient there are no wars because the barbarian invasions are too rampant [...].
None in the Classical era either? The Ancient era usually ends by 1000 BC (around turn 80 on Normal speed) for the AI. Warfare quite that early is a tall order.

I'm attaching a screenshot from turn 100 of an all-AI game I've just run on a Huge Fractal map on Monarch difficulty with default settings otherwise. The game's first war has just been launched in the SW by Hammurabi against "my" civ, the French, immediately razing one city. It looks like a bunch of civs will get no more than 5-6 cities by peaceful means, but that, if we imagine a human player in one of those tight spots, should be enough to delay a break-out war until Medieval times (if one so chooses), for longer if the player can snatch up a city here or there as the AI neighbors war against each other – or against the player. Some also have an option to settle quite a few cities in rather marginal-looking terrain. Exploring the oceans isn't rewarding on this specific map, but that can be an option too. Positions with fairly little room and few interesting options, e.g. the Dutch on the NE continent, arguably occur too frequently, but I don't see much room for giving civs more space. The first war – among 16 civs – just before turn 100 is already on the late side of things for me. Especially for civs without much land to themselves, AI-initiated wars present opportunities. If Gilgamesh attacks Spain or the Inca, things get interesting for the Dutch.

If 12 civs on Huge meet your needs, that's fine of course. Just wondering if it's also down to other settings or unspoken assumptions. If the Barbarians are really too much of a good thing at 12 players, then I guess the only remedy will be changing the difficulty-based settings in Civ4HandicapInfos.xml.
 

Attachments

  • Civ4ScreenShot0000.JPG
    Civ4ScreenShot0000.JPG
    286.3 KB · Views: 17
Back
Top Bottom