Buffed AI for BNW

[glider1]

Problem is your game was spoiled because you were able to grow population like that so fast. On some maps it is not possible to do, and then the tech funnel works fine as is. Another problem is that it depends on AI population. If they have a smallish population, no matter how wide or strong the funnel is, it won't pull them in enough.

I don't exactly understand how the diplo modifier works, like when it kicks in for what type of civ. I'll have to look through. I just played a test game where I grew tall quick enough that Venice should have complained, but they didn't.

 

[glider1]

Slowly ratcheting up the difficulty. This version could well keep the AI's in the funnel because of the boosts to the passive buffs. Really got my hopes up for this version!

V18 #9:

• AIs resent both tall and wide reckless expansion
• AIs more wary about human reckless expansion either tall or wide
• AIs within tech funnel will accelerate back in tech quicker and de-accelerate past (cubed pull back instead of squared)
• AI declare war probability on player no longer flat 120% but increases with difficulty
• AI-AI research agreement bonus buff back in
• AI offensive attack bonus slightly increased (to help the AI clinch the deal when capturing cities)
• AI passive per era bonuses increased

EDIT: So this version will make things tougher for players with terrestrial neighbours, and makes things a bit tougher for players on their own islands (but not as much, which gives the possibility of playing less war dominated games).

 

[glider1]

Ok, test successful, after just 14 turns, for the first time in the history of Civ5, an AI resents my tall expansion:

Spoiler :

I out grew Germany 4 pop verses 3 pop because of the citrus. So 4 / 3 = 1.33 which is greater than the threshold for emperor. Bismarck's army is bigger than mine which means that the negative diplomacy hit is not inhibited.

EDIT: the 3 in the above calculation is the average of all other civs total population except for target civ. It is the same calculation as wide substituting number of cities for population.

 

[joncnunn]

Slowly ratcheting up the difficulty. This version could well keep the AI's in the funnel because of the boosts to the passive buffs. Really got my hopes up for this version!

Downloaded, with me recently hearing on the main discussion forum complaints about if you fully exploit the Irq UU & UA their food growth being too low and also the last time I've played a full game with them was back in Vanilla, I'll play them with Beta 9.
Their forest bias could be a sharp contrast with India' grassland bias. (Or alternatively it could just show once again how much food cargo ships are overpowered.)

I'm thinking I'll be a bit slow on purpose sharing embassies as well. AI can't complain about my bigger city if they can't see it, right?

 

[glider1]

Good try, but I think lack of embassies won't make a difference! That sure will be interesting to try a low pop growth civ.

The calculation I used for tall is the same style as the wide code that already exists. So I think all the AI needs is to have met you. That gives it access to the number of cities you have (for wide) or your total population (for tall). I think it has been like that all along in Civ5. Actually, yes because it is no different for us humans. As soon as we meet a civ, we know their city count and total population via the score breakdown in the mouse-over.

EDIT: So the best way not to get the penalty, is to not meet the civ at all (which makes terrestrial neighbour maps harder than island maps, but at least a player can choose to take a bit of refuge in a more peaceful game by choosing an island map).

 

[joncnunn]

Good try, but I think lack of embassies won't make a difference! That sure will be interesting to try a low pop growth civ.

Yeah, I saw that when on turn 85 Rome met me via finding my western city on my landmass and while they were showing Friendly, they also showed "They believe we are populating recklessly".

Would you believe that at look at the demographic screen showed them in last place in food production?

It does appear though that they need to see your city to complain about any wonders you built because exchanging embassies caused them to also get World Wonder envy. They also got land envy the next turn, but that would have happened soon anyway since they are my closest neighbor.

The Irq starting bias actually placed me "near" a forest instead of in the middle of one. Growth in my capital would actually be pretty good with 2 cows.

The large island script also likes to place capitals in coves, my own capital & Rome are both in coves.

 

[glider1]

Interesting on the world wonder envy. I haven't had that one. How many wonders have you been able to build on immortal to cause that?

The AI's growth rate is the concern worth keeping an eye on. So far in my own game it seems good. What would have caused Rome to have the lowest population? Maybe keep an eye out on the possibility that on island maps, the AI is not building fishing boats enough, something like that?

EDIT: If you got the reckless expansion hit, that means they have a bigger army than you and therefore are not repressing their opinions. Question is why would Rome have a small pop and big army on an island map with no neighbours?

 

[joncnunn]

Interesting on the world wonder envy. I haven't had that one. How many wonders have you been able to build on immortal to cause that?

The AI's growth rate is the concern worth keeping an eye on. So far in my own game it seems good. What would have caused Rome to have the lowest population? Maybe keep an eye out on the possibility that on island maps, the AI is not building fishing boats enough, something like that?

EDIT: If you got the reckless expansion hit, that means they have a bigger army than you and therefore are not repressing their opinions. Question is why would Rome have a small pop and big army on an island map with no neighbours?

I had actually only built 1 World Wonder. (And still only have 1 world wonder). My guess is that Rome had been trying to build it. Wonder envy has never been them comparing about number of wonders you've built, but instead specific ones. It does stack though.

Rome is no longer in last place for food, that "honor" now goes to the Incas.

Immortal as Irq Beta version 9 with my National Wonder changes.

Turn 0: No forest in sight even with Irq forest bias. However, with 4 Copper, 1 Marble, 1 Stone, 2 Cows, and on a river hill tile on a mountain, it's a good starting location.

Turn 4: The first ruin I pop is a cultural ruin.

Turn 5: I open Tradition

Turn 9: Mt Kalish discovered, 4 NW left.
Barb camp is near an unmet city state. The barb camp is too close to comfort to my own city.

Turn 10: Jerusalem greets me so I don't have to sneak past the camp. +8 faith

Turn 11: The promised forest from my starting bias finally appears.

Turn 13: Add legalism.

Turn 18: I found Mother Earth, Jerseulum is going to take in so many useful desert tiles that I'll get more faith out of Mother Earth than Desert Faith, and 7 sources beats both 1 NW and 2 Queries.

Turn 23: Monarchy added

Turn 26: Barb camp formed west of my capital.

Turn 31: Between having 2 nearby barb camps on opposite sides and a very high city site to found (Natural Wonder with built in happiness bonus), I'm building the normal library.

Turn 41: Landed Elite added

Turn 44: Unknown civ adopts Land of the Open Sky

Turn 52: Western barb camp destroyed. Killing the naval units it spawned will take a few more turns.

Turn 53: Eastern barb camp destroyed for Jerusalem. Killing the land units it spawned will take a few more turns.

Turn 54: I kill one of those barb units, becoming allies for a few turns (and friends for several more)

Turn 56: 2 unmet civs enters Classical era
Osininka founded to the E adjacent to natural wonder. I cash rush a granary so it will work it immediately.

Turn 58: Pop a ruin near there. Darn map ruin

Turn 60: Unmet civ enters classical era.

Turn 63: Temple of Artemis built far away
I found Grand River to the west.

Turn 64: Unmet civ wastes pantheon on Goddess of Protection

Turn 66: Stonehenge built in faraway land.

Turn 68: Great Prophet born. I'm first to a religion and found Protestantism.
Tithe as founder belief, and with faith natural wonder + Earth Mother I also choose Pagoda.

Turn 69: Unmet civ enters Classical era.
Artistocracy added.

Turn 70: Great Library built far away.
Unknown civ adopts Oral Tradition
Unmet civ enters Classical Era

Turn 71: I enter Classical Era via optics. I've not yet met anybody.

Turn 78: Unknown civ enters Classical Era.

Turn 81: Pyramids built far away

Turn 83: Unknown civ enters Classical Era

Turn 84: I complete Mas of Hal - two tiles with direct benefit (+ normal spawning bonus). A Golden Age was going on.

Turn 85: Rome meets me. They are showing "Friendly" but also have "They believe we are populating the region recklessly". Of course they would with the worst crop yields and 4 subpolicies into Liberty. They haven't built any wonders nor do they have a pantheon.

Turn 86: Rome is now Guarded with me and in addition they now have land envy & wonder envy.

Turn 88: Ter Army built in a far away land.

Turn 89: Great Lighthouse built it a far away land.

Turn 90: Unmet civ founds Catholicism

Turn 96: I add oligarchy, completing Tradition.
GP born. I enhance Protestantism adding Religious Texts and Feed the World

Turn 100: Parthenon built in a far away land.
I meet Carthage. Their homeland is far to the NW even though I saw their ship NE of mine.
They built the Great Lighthouse. No pantheon

Turn 102: I found Akweasane to take in the northern part of my landmass.

Turn 104: Hanging Gardens built in a far away land.
Carthage completes Oracle
Greece meets me. No wonders, no pantheons.

Turn 105: Incas meet me. They've built Stonehenge and founded Catholicism.

Turn 107: Carthage enters Midevil era

Turn 108: Spain meets me. They have Parthenon and a pantheon.

Turn 111: Greece enters midevil era.

Turn 115: Spain completes Chicken Izta.

Turn 116: Unmete civ founds Islam

Turn 119: Incas built Great Wall

Turn 120: I meet Ottomans. They have Great Library & Terra Army. They founded Islam.

Turn 123: I open Aesthetics as I'm planning a cultural victory

Turn 124: I discover Civil Service, entering Midevil era.

Turn 124 status:

Me: 20 techs. #1 pop, crops, man goods, land, and science

Rome: 19 techs. Worst approval. Guarded

Ottomans: 19 techs. Worst population & troops

Incas: 19 techs. Worst crops, man goods, GNP, land, and science.

Carthage: 20 techs. #1 approval

Greece: 20 techs. #1 troops

Spain: 20 techs.

1 unmet civ: #1 GNP

6 city states & 4 natural wonders left to discover.

Screenshots of beginning of Classical & Midevil Eras.

 

[glider1]

Thanks Joncnunn the AI's in your game look to be locked into the tech funnel because their passive buffs are sufficient.

I have autoplayed through the first 50 turns of a large islands map just to see if I could replicate the sluggishness in the AI you are noticing. First thing that struck me is that the AI doesn't know how to play that map. It seems to assume that it is on a standard map like continents. It doesn't adjust. Therefore it is slow to develop sea resources (except for some civs that do), and it has extra difficulty deciding on settlement location and so takes longer to settle (sometimes 10-20 turns longer than on a continents map). It also seems not to clear out the barb camps as well either.

I doubt that it is a good idea for increasing the intercity spacing like you do for that type of a map with so little land area. I think it might hurt wide playing AIs that need to plonk a city down without the additional restriction. They may even have more confusion about where to settle because of the extra spacing requirement (not sure).

I'm only speculating based on my observations of watching the AI play on that map. It is not because of a bug in this mod. Workers are building improvements quickly, and the AI city focus is growth oriented. It is seems to be a map that is beyond the limit of the AI's programmed ability. It should probably not be picking liberty it seems to me!

EDIT: There are even AI's that go full honor with no rivals to conquer being by themselves on a little island. By the time they should be enjoying the benefits of honor, they have developed too slowly. The adaptive buff algorithms in this mod then cannot help them, because their population size relative to you is too small. So things fall back onto the passive buffs to keep them hanging in there, and at least I have increased those substantially for this beta.

 

[joncnunn]

Thanks Joncnunn the AI's in your game look to be locked into the tech funnel because their passive buffs are sufficient.

I have autoplayed through the first 50 turns of a large islands map just to see if I could replicate the sluggishness in the AI you are noticing. First thing that struck me is that the AI doesn't know how to play that map. It seems to assume that it is on a standard map like continents. It doesn't adjust. Therefore it is slow to develop sea resources (except for some civs that do), and it has extra difficulty deciding on settlement location and so takes longer to settle (sometimes 10-20 turns longer than on a continents map). It also seems not to clear out the barb camps as well either.

I doubt that it is a good idea for increasing the intercity spacing like you do for that type of a map with so little land area. I think it might hurt wide playing AIs that need to plonk a city down without the additional restriction. They may even have more confusion about where to settle because of the extra spacing requirement (not sure).

I'm only speculating based on my observations of watching the AI play on that map. It is not because of a bug in this mod. Workers are building improvements quickly, and the AI city focus is growth oriented. It is seems to be a map that is beyond the limit of the AI's programmed ability. It should probably not be picking liberty it seems to me!

EDIT: There are even AI's that go full honor with no rivals to conquer being by themselves on a little island. By the time they should be enjoying the benefits of honor, they have developed too slowly. The adaptive buff algorithms in this mod then cannot help them, because their population size relative to you is too small. So things fall back onto the passive buffs to keep them hanging in there, and at least I have increased those substantially for this beta.

It's indeed the case that the AI treats every map the same, from Tiny Islands to Pangena and from Ice Age to Oasis.
Fish actually should be slow to develop, there's really no point in a work boat for fish untli after the light house is been built. (+2 f & +1h for every resource tile from a lighthouse vs only +1f from workboat.) In fact I personally prioritize a library over the work boats going to fish. (By contrast lighthouse is right after the granary) It's spare luxuries they are slow to connect (and perhaps even slow to construct the lighthouse, I'm not sure.)

Whenever the primary luxury nearby is a sea resource, it's going to be a slow start for all map types, including if you started on the coast of pangena.

I also play with 8 city states instead of 16, that restores some of those land tiles (by not having as many city states in the way)

I find that on standard spacing that the AI founds cities too close together, founding a city whose only key tiles it will take in are those that its existing cities will take in and work, and so if I when I went on a conquest spree I'd have to raze about half of them to the ground. Under this enforced setting, I only find myself razing 20% of the time.

Now if it's possible to have city state's exclusion zone being smaller than major players, that would be preferred.

On policies, it's never considered the map settings when picking which trees to open, so that's a big area for possible improvement. In fact, I found out early in Vanilla that if you turn all city states off, some of the AIs will still go full patronage.

Large Islands does appear custom made for a Tradition player, especially with the legendary advanced setting I use.

But Tiny Islands is actually a map type popular among Liberty players.
While both are water maps, the typical placement patterns of the islands on "Tiny Islands" tend to support self founding a larger number of cities than Large Islands.

Edit: Rome just went from Guarded to Friendly with all negative modifiers except for the new one missing. Did you remember to have it be hidden when the AI is "pretending to be friends"?

 

[joncnunn]

Midevil Era of above game.

Turn 125: Ottomans, Irq, and unknown Civ all advanced to Midevil era.
Greece declares war on Ottomans

Turn 126: Carthage completes Hagia Sophia
Carthage founds found Messenger of the Gods & Buddhism
This means Spain is now locked into getting the 5th & final religion and so as a result, my religion is going to passively take over Rome.

Turn 128: Petra built far away.
Spain completes Manchu

Turn 130: Inca enhances Catholicism
I discover Old Faithful. 3 natural wonders left

Turn 131: I discover a city state, 5 left

Turn 132: Incas want to be friends. Why not?

Turn 133: Rome enters midevil era
Rome has happily adopted my religion, but is still guarded

Turn 138: Spain enters Rean era.

Turn 139: Spain just went from Guarded to Friendly, with all negative modifiers hidden except for the new populating the region too aggressively. I think they are actually pretending to be friends.

Turn 140: I meet America. They've built 4 world wonders.

Turn 142: Spain founds Confucianism

Turn 144: I add cultural center

Turn 145: Rome completes Colossus

Turn 146: Spain completes Alham

Turn 147: Ottomans enhance Islam.
Carthage enhances Buddhism

Turn 149: Carthage enters Rean era (first)

Turn 150: Greece enters Rean era

Turn 151: I discover a city state. 4 left.

Turn 152: Ottomans want to be friends. Why not?

Turn 153: I discover a city state. 3 left.

Turn 154: America completes A. Wat
America enters Rean era.

Turn 156: I discover a city state. 2 left.
Great Prophet born, I build a holy shrine with him.

Turn 157: GS born, I build an academy.

Turn 158: Carthage discovers war on Ottomans

Turn 159: Greece & Ottomans sign peace treaty.

Turn 160: I discover a city state. 1 left.

Turn 161: Ottomans denounce Greece

Turn 162: America wants to be friends. OK
Spain completes Great Mosque
I discover Solomon's Mines. 2 left

Turn 163: Rome enters Rean era.

Turn 164: Incas enter Rean era.

Turn 165: Spain enhances Confucianism

Turn 166: I add Fine Arts

Turn 167: America completes Notre Dame

Turn 169: RA completes, completing the rest of Printing Press and advancing me to the Rean era.
World Congress also founded, I'm the host.
I propose Arts Funding.

Status, everybody is in Rean era.

Me: 30 techs. #1 population & land

Rome: 32 techs. They are back to Guarded. They no longer think I've populated the region too aggressively, but still have wonder envy over the 1 built, land envy, and now city state envy over my one ally.

Ottomans: 30 techs. Worst in everything. At war with Carthage. Friends with me & America

Incas: 32 techs. Friends with me.

America: 33 techs. #1 crops & man goods. Friends with me & Ottomans

Carthage: 34 techs. #1 troops, approval, and science. At war with Ottomans.

Greece: 33 techs. #1 GNP.

I have 1 city state & 2 natural wonders left to discover.

Weird for me not to be #1 in techs at the beginning of an era.

How I've used the UA: You may notice that there's no road on the forest, but both the SE city and the NE city are still connected. (The NE one also via city state). Jerusalem had asked for a road to my capital, and this was enough to grant it. (It's fulfilled when you place a road from the city state to any of your cities that have a city connection)

Use of my UB: The SE city has the most potential for this (if I can get it to grow enough, as you can see Jerusalem grabbed the really fertile tiles.) My capital, western city, and northern city are all limited to one hex each. (The western most forest tile is off the map, the northern city use will be after culture expands onto it)

Use of my UU: You can't really tell the quantity from the screen shot, but that was only a 2 Iron. I did however sell it since my UU doesn't need it and so it won't be until Longswords that I need it back. Their forest & jungle promotions though are unlikely to come into play with Rome having been to my north instead to to my south or east.

Screenshot attached

 

[glider1]

Wow, big difference in AI tech rate, good to see. I'm getting similar performance on emperor.

The negative diplo I made is actually the same diplo for reckless expansion, just code added to detect reckless tall as well as reckless wide and reworded to cover both cases. If you expand both tall and wide the AI treats it as one diplomatic negative. Separating tall and wide diplo modifiers is a lot more work and I'm also uncertain what effects it would have on diplomacy so prefer taking the conservative approach for now. So the way that the modifier works in terms of being hidden to the player is unchanged from the original function.

There are some ways that the code can be sensitive to map type because of the function GetAIMapHint() which gives the AI some ability to distinguish map types. At the moment the AI will adjust by map type for:

• formation size
• quick colonizing
• calculating best settler plot for colonization
• when expansion is enough
• tactical naval operation
• helping to choose a great person

EDIT:
I guess it could be extended. At least some map knowledge does exist, but yeah, it basically plays the same way for most maps.

 

[joosegoose25]

Hey! I like what this mod is trying to accomplish but have a few questions about some aspects.

With regards to the AI science buffs - if I understand this correctly the AI tech rate gets boosted if they fall a couple techs behind? Does this mean it's almost better to take the foot off the gas pedal a bit when it comes to your science? Or should you still be pursuing science as hard as ever?

I see the AI's combat is supposed to be boosted in more ways than tactical decisions. Is there a flat buff to AI strength that's applied directly to combat strength? Will the player's troops always be 20% weaker or so than the AI's simply because they are controlled by a human?

Lastly, regarding the DLL - what particular aspects of the mod is the DLL responsible for (other than ninakoru's contribution of course)? If one were to play without it, would certain aspects not work, or would the entire mod be null and void?

Thanks in advance! Like what you're doing here.

 

[glider1]

Hey Joosegoose, no probs, happy to answer. The tech rate is not adaptively boosted, but rather the science yield. The AI still chooses whatever tech it wants I don't interfere.

This mod is basically set and forget. Pick your normal difficulty level and go. After a while, you don't even realise that there is a difference between this mod and the standard game. Try different things out. You don't always have to play the science race game.

Brave New World was actually designed to be like that, but the AI is too weak, and so the whole wide ranging concept of the game is narrowed not because of the mechanics, but because of the handicapping system for the AI.

So onto the details:
If the AI is ahead of the player on its own terms, the adaptive buffs in this mod turn off. If the AI would otherwise start to fall behind the player, the adaptive buffs in this mod turn on. So in the latter case, the more you put your foot on the pedal, the more core yield leaks to the AI. The more techs you are ahead of the AI, the more science yield leaks to the AI yet again. So yes if the AI needs help, if you take your foot off the pedal, the AI will slowdown too! It depends on how much the AI is being propped up by the adaptive code, and that depends on difficulty. A weak player playing on deity, will mean that the adaptive buff to the AI will almost be totally off. A strong player playing on warlord, means that the adaptive buff will almost always be on. If the AI is unhappy, the adaptive buffs progressively start turning off as well. Rationalism also leaks to the AI, to prevent humans always beelining it. Not only that, but AI-AI research agreements end with extra science yield to the AIs if they need it relative to the player.

There is basically not much you can do to stop the AI getting buffed, unless it is ahead of you, or it has a smaller population than you. The smaller its population relative to you, the less buff gets fed to it. So the AI's to watch out for in this mod are the under performers with a big population that are able to stay happy. You can delay your quantity of techs researched, so that it appears to the code that you are teching slowly. That would be one trick you could use, but that would only partially work as well. Another trick would be to save up your great scientists and don't use them at all, expending them off for a scientific revolution in one turn, right when you need it. Problem is that as time goes by, the AIs will catch up in science and pass the player quicker as well.

So the difficulty level is trying to find a balance between getting an AI ahead of the player enough to make the endgame interesting. This requires it to pull its weight to some extent, otherwise it will never pull enough of a gap out to win. That is where the sensitivity of difficulty level matters.

The combat buffs are static buffs applied once at the start and do not change.

Nerfs to humans:
All human units and cities are weakened in strength, and human ranged attack is weakened as well. Range-3 human attack is weakened on top of the nerf, but retained against cities. So for example, artillery and battleships will be randomly weak attacking units at range-3, but fine against cities and anything at closer ranges except for the nerfs.

Buffs to AI attack:
The AI gets a buff when it attacks in whatever way, units or cities. So this buff affects human-AI play as well as AI-AI play. It helps offensively attacking AI's to do better. The AI is also more likely to declare war on humans too for three reasons:

1. the AI human declare war prob is increased
2. AIs resent rapid tall or wide expansion based on difficulty level
3. AIs know about the human nerfs to combat and adjust their calculations accordingly.

Also, AIs are not as easy to bribe into third party wars.

The combat buffs and nerfs are designed to make unit losses in battle between humans and AI roughly equal at emperor level. Emperor level is where it is balanced. Really good combat players will loose less units, weaker player will loose more. But at low levels, the AI doesn't produce as many units and slower, at higher levels the AI builds more units and faster. The system is also designed so that humans bring in a balanced composition of army (no longer just a wall of ranged units and one melee, or a wall of artillery at Dynamite tech).

The DLL runs the show totally. Without it, this mod is nothing much.

 

[joosegoose25]

First of all thanks for the in depth reply, and the swiftness of it!

The tech rate is not adaptively boosted, but rather the science yield.

So, if the AI is receiving an adaptive bonus from being behind (I don't know how big it is so let's just say 10%), its science yield would go from say 100 beakers to 110 per turn? Although I hate the artificial feel of it (if you're right with or ahead of the AI any extra science you generate is somewhat pointless) boosting the raw yield is probably as good a way as any to compensate for the AI. I certainly can't come up with a better way outside of a near impossible complete overhaul of the AI.

Is the adaptive rate a flat percent based solely on difficulty level that is received as long as the AI is within a certain range behind the player (i.e. 10% or so like I said that is received no matter how far behind they are), does it vary depending on how far behind (i.e. 10 techs behind gets a bigger boost than 2 behind), or is it calculated from the player's science yield? I'm under the impression that it varies depending on exactly how far behind the player a given AI is plus the AI's population relative to the player.

On that note, let's say it's late in the game and because the mod is doing an awesome job, I'm only 1-2 techs ahead of all AIs. I receive 2 free social policies, and I decide to use them on the Rationalism opener and Secularism, which boosts my raw science by say, 20%. Obviously some of this is leaked to the AI, but how much? Are the policies completely pointless or do I still gain an advantage, albeit a much smaller one?

Regarding the bonus to AI offensive combat, is it a combat modifier similar to a unit with Drill that is attacking into rough terrain? Is it present every time an AI is the initiator of a battle? For example, if I am the aggressor invading an AI's capital and they attack my crossbowman with a musketman, do they get an offensive boost?

Also, if I edit the Handicap.xml does that change the combat nerfs/buffs, or is there more to edit? I'm personally terrible with combat so I'll probably give a slight nerf to human ranged but keep everything else intact. On that note, how would you feel about giving melee units (at least AI, potentially all units) a slight boost vs cities? I've seen it before in combat mods and it certainly helps the AI with their "throw melee units at the city and see what sticks" strategy.

Sorry for the exorbitant number of questions, but this does more than change AI tactics and tendencies than the typical AI solutions so I just wanted to be sure of the gameplay changes From what I can gather, a lot of the information seems to be rather spread around this thread so I wasn't 100% sure what all was in the current build.

I really like giving free Cover to siege units and the edit to make AI's more aggressive and less willing to be bribed. I'm a little disappointed that everything is dependent on the dll, but then again I'd be shocked if you were able to do all this without it. I currently use whoward's dll merged with the new quiet diplomacy mod (my favorite new UI mod), but these changes sound so good I may have to forget about those for a while

Edit: I see in the zip file you uploaded there are 2 main folders. Do we need to do anything with the CvGameCoreDLL_Expansion2 folder?

 

[glider1]

Rationalism could still be useful as part of your strategy. It still can be good, just no where near as good as the base game.

I'll try and do a merge with Whoward's DLL once he releases the code for V62. The answer to most of your questions are basically yes. You can make adjustments to combat in the handicap.xml go for it.

With this mod, suggest not to think about specific yields that flow to the AI, how much and when. Just play the game the way you want to play it and how it was intended to be played. There is no real difference in anything, other than it being more punishing at times.

You don't have to push the science throttle, but in many respects you have to push the population throttle instead...
(Some could say that I have turned the game from a science race, into a population race....)

Only science, production and gold is leaked to the AI nothing else. Everything revolves around the AI's population and happiness. If an AI is weak in population growth or has suffered major population loss from war or is desperately unhappy, they are not going to get as much buff. The thing is, that some AI's are going to do well, even without a big population, just because they have managed to optimise everything by accident!

I'd say that is the simplest approach. Enjoy it would be the other important point! Be wary about trading luxuries to civs that potentially have a big population. If they are happy, they could potentially be buffed more. It depends on how well you are doing. Make sense right? The game was supposed to have negatives to selling the AI luxuries!

EDIT:
I enjoy being beaten by the AI and am glad when it happens. Suggest to do that as well. Find a difficulty level that repeatedly produces good game for you. That is all there is to it

 

[joosegoose25]

With this mod, suggest not to think about specific yields that flow to the AI, how much and when. Just play the game the way you want to play it and how it was intended to be played. There is no real difference in anything, other than it being more punishing at times.

I think this is all I needed to hear The only reason I was so specific in my questions was to make sure the game should still be played normally. If it was pointless to prioritize science techs/policies like usual I was going to to focus more on the lower half of the tree for a change. I'll definitely be more flexible than usual but it's good to know human science isn't largely irrelevant.

I enjoy being beaten by the AI and am glad when it happens. Suggest to do that as well. Find a difficulty level that repeatedly produces good game for you. That is all there is to it

Same here, it's the reason I exclusively play Deity or modded Deity equivalents. Only problem is it always pidgeonholes you into a certain strategy which gets boring after a while. I'm looking forward to hopefully playing Emperor or so and being challenged

You don't have to push the science throttle, but in many respects you have to push the population throttle instead...
(Some could say that I have turned the game from a science race, into a population race....)

Looks like I'll be rushing for Temple of Artemis for once, and for the first time since maybe my first game ever I'll be pushing for Physics/Notre Dame. Woohoo!

I'm assuming there's nothing that needs to be done with the CvGameCoreDLL_Expansion2 folder?
Also happy to hear you plan to merge this with whoward's dll, though I still wouldn't be able to use Quiet Diplomacy unless there'd be a way to merge that, too. Looks like I have some learning to do

Time to jump in, I'll be sure to report if I have any interesting observations!

 

[glider1]

Thanks Joosegoose. I would be keen to hear your feedback good or bad. If you are a deity player, I doubt emperor will challenge you sufficiently, but it would depend on the map settings. Yes do concentrate on Notre Dame if you feel like it, clinching it is a huge prize. Concentrating on the bottom half of the tech tree is actually a very good idea, because it means you can grow population up safely.

I've had a bit of a think about how to simply state what this mod does.

Yes I have basically turned it into a population race instead of a science race (simply by how I have buffed the AI). So if you want to know how well you are doing, take a look at your rivals city population numbers, that will tell you a lot right there. I don't boost population in anyway. The AI's population is real. So too is its happiness...

In my defense, that I have just turned the game from one type of race to another, I would argue that a population race is *much* better than a science race because it is multifaceted where as a science race is one dimensional. Population can be gained either peacefully or through war. Optimising population carries inherent risk as well. It can all fall apart very quickly. I think it is what Firaxis actually intended civ 5 to be (not a science race).

So here is a simple rule of thumb then. A Green start means you can play tall. A brown start means you should consider playing wide.

EDIT:
On the subject of combat, the simplest way to think of this mod is just to focus on the units and their promotions, and underplay the importance of ranged, I have devalued it. The AI gets no free promotions. What you see is what you get. Also, protect the ranged you do have. They are not as powerful, but if they last the entire game, they will get a bucket load of promotions.

 

[joosegoose25]

Good explanation, I totally get what your goal is here and I like the direction.

One last question (well I'm sorta re-asking a question) what exactly does the offensive bonus for the AI mean? Is it how I described, where they receive it whenever they initiate an attack (regardless of melee/ranged attack and even if I DoW and we're fighting in their territory)? Just curious so I know if I want to play around with the value for it or not.

 

[glider1]

Good question Joosegoose, I had to look back into the code to remind myself. I boost the values these functions output for the AI:

GetMaxRangedCombatStrength
GetMaxAttackStrength

I also thought that I boosted this function:
CvCity::rangeCombatDamage

so that the AIs city ranged attack was stronger, but I appear not to have done that. Possibly because generally I don't really like the whole concept of cities firing bullets on their own, so just left it at base game level. So the "offensive attack" which is actually just "combat attack strength" is boosted, and applies only to AI units.

EDIT:
It always applies and is added on top of what the AI already has through traits, promotions whatever. It includes the barbarians too so factor that in. In this mod they are a real menace. At times I have had to allocate units just for the sake of acting as outposts to prevent barbarian camps from sprouting, and that can be a drain on the economy (but also adds more depth to what we have to factor in).