(Testing) Leader Personalities Idea

I am very confused.
Okay one, how the hell are those text files that large? Kilobytes? I've written treatises in notepad that didn't get that big for a dozen pages. You have whitespace. I can't even download Personalities.txt, my browser aborts it.

The requested data are SQL files to get your DB settings (which you already have , you used them to mod your game), and a spreadsheet to visualize the changes, which I can manage half of. I have an SQL server on my lappy, so I can make correct tables on my server, run your SQL , and then get text output of the table display. I can process that output and pretty-print it. I 'm not confident in reading the Google Sheets API, so I'll make a data browser and jma can use that to put into Wiley's sheet

You wrote "I tried to use the spreadsheet that is supposed to plug in and make the SQL," and that makes no sense, because you already have your SQL, you attached it some days ago.

---
I worry those policy flavours have constrained AIs to all use the same suboptimal strategy. You have restrained them all to think with minmaxxing, but minmaxxing is not correct always or even generally in this genre. Spaceship needs Rationalism, but all the others just need the resources they need; and the most valuable investment could be different things, depending on any number of other things. If the difference between 'poor' and 'okay' is greater than the difference between 'good' and 'excellent', then curbing your weaknesses is the good choice, which policies can do.

Making literally flavourful choices for policies, as with buildings, is something I could put up with for a while yet.
 
That's why I have the wild card category, so that the AI sees opportunities to use Religion or wealth as an alternative to specializing with Victory policies.

Also, while the AI have mostly taken Might, they have mixed it with Liberty and Tradition with some civs, full Might with others progressing to Piety or Aesthetics.

I have them displayed as text files because as far as I could tell, I can't directly upload them as SQL files, and I just copied and pasted them into a .TXT file. If that isn't working, though, I could try to post them .ZIP files.
 
I really enjoy the idea of the AI being more aggressive but I feel like it doesn't fit perfectly into the game's current mechanics.

For example, In my last Immortal Austria game I was next to Attila and Morocco. Obvious Attila was just being the **** that he usually is and set up all his expansions next to my cities. Later, he started massing a million units and I'm obviously pouring everything into defense to the point that I'm negative happiness and -30 gpt. I barely survive his initial assault and retake my cities. So far so good. I'm behind on science, economy, policy, etc. but I can catch up by grinding Attila's face in the dirt.

Then, the second I take and raze one of the garbage cities he planted on top of my borders to clear some space, Morocco denounces me and starts consoling poor Mr. the Hun for losing his 2 pop 0 lux trash. I have transparent diplomacy and I see (-189) for my moderate warmongering.

This is total . I should have the right to take the cities of my besieger after he goes all-in. I understand this problem was present in the base game, but it's magnified when the AI always expands next to my borders and declares war at the drop of a pin.

Ideally, warmonger penalties for capturing non-capital cities should be drastically reduced based on whether they were planted next to your cities or who declared war on whom. I understand this would be a pain to program and totally outside your purview. As a substitute, would your warmonger penalty values get overwritten by other mods that modify warmonger penalties? Alternatively, can you suggest a way to conditionally reduce warmonger penalties for city captures based on who declared war? The funnest part of this game for me is definitely getting revenge on AIs when they try to **** you, and this warmonger stuff makes the consequences for doing so unrealistically severe.

Otherwise, nice job! I'm curious to see the direction this will take.

*EDIT* I just saw that WHoward has a mod that does just what I want. I'm gonna try it with the new personalities.
 
I really enjoy the idea of the AI being more aggressive but I feel like it doesn't fit perfectly into the game's current mechanics.

For example, In my last Immortal Austria game I was next to Attila and Morocco. Obvious Attila was just being the **** that he usually is and set up all his expansions next to my cities. Later, he started massing a million units and I'm obviously pouring everything into defense to the point that I'm negative happiness and -30 gpt. I barely survive his initial assault and retake my cities. So far so good. I'm behind on science, economy, policy, etc. but I can catch up by grinding Attila's face in the dirt.

Then, the second I take and raze one of the garbage cities he planted on top of my borders to clear some space, Morocco denounces me and starts consoling poor Mr. the Hun for losing his 2 pop 0 lux trash. I have transparent diplomacy and I see (-189) for my moderate warmongering.

This is total . I should have the right to take the cities of my besieger after he goes all-in. I understand this problem was present in the base game, but it's magnified when the AI always expands next to my borders and declares war at the drop of a pin.

Ideally, warmonger penalties for capturing non-capital cities should be drastically reduced based on whether they were planted next to your cities or who declared war on whom. I understand this would be a pain to program and totally outside your purview. As a substitute, would your warmonger penalty values get overwritten by other mods that modify warmonger penalties? Alternatively, can you suggest a way to conditionally reduce warmonger penalties for city captures based on who declared war? The funnest part of this game for me is definitely getting revenge on AIs when they try to **** you, and this warmonger stuff makes the consequences for doing so unrealistically severe.

Otherwise, nice job! I'm curious to see the direction this will take.

*EDIT* I just saw that WHoward has a mod that does just what I want. I'm gonna try it with the new personalities.

Whoward's warmonger-related mod is already incorporated into the CP.

G
 
Did Morocco denouncing you lead to other denouncing or just them?

Also I'm going to post a list of who is who so that people can choose their rivals however they like. I personally have an even number of all 4 so that there is plenty of competition for all 4 victories. A full group of Federalists with the player as a warmonger would be suffocating lol

Nationalist
Assyria
Denmark
The Huns
Japan
Mongolia
Rome
France
Songhai
Spain
Venice
The Zulu

Imperialist
The Aztecs
China
Germany
Greece
Indonesia
Poland
Russia
Sweden

Federalist
Austria
Brazil
Byzantium
Egypt
Ethiopia
The Inca
India
Korea
The Maya
Morocco
Polynesia
Siam

Hegemony
America
Arabia
Babylon
Carthage
The Celts
England
The Iroquois
The Dutch
The Ottomans
Persia
Portugal
The Shoshone

I'm thinking about moving some of them around to fit history better. For example, America is driven quite heavily by technology and space. The Roman Empire wasn't a unitary state, whereas China pretty much always has been.
 
Sorry how do you apply this update? Whatever I did causes the game to go to desktop when you try to start a new civ.

Also is there anyway to increase ai meanness and boldness by 8 across the board? I had another mod that did this which doesn't seem to work with this mod. However it was just an sql mod.

I wanted to try this change because I tried to play a game with this mod and although it was a good mod, all the ai civs were too passive. Maybe bad luck with random personalities. But no personality should be afraid to go to war. This idea seems to differentiate the leaders without making them passive.
 
Sorry how do you apply this update? Whatever I did causes the game to go to desktop when you try to start a new civ.

Also is there anyway to increase ai meanness and boldness by 8 across the board? I had another mod that did this which doesn't seem to work with this mod. However it was just an sql mod.

I wanted to try this change because I tried to play a game with this mod and although it was a good mod, all the ai civs were too passive. Maybe bad luck with random personalities. But no personality should be afraid to go to war. This idea seems to differentiate the leaders without making them passive.

The latest post I made with the three attachments are .txt files that I copy and paste into the appropriate files in the CBP. LeaderPersonalities.SQL in the AI Personalities Folder, LeaderFlavors.SQL in the AI Flavors folder, and Flavors.XML in the Flavors folder of the Policies folder.
 
Okay I goofed pretty hard on the policy flavors, and made some changes. I had Nuke tied to certain policies, when I missed the file that said Nuke was deleted as a policy flavor lol. This was changed to Offense.

The AI will hopefully pick their Ancient policy based on their starting situation and see the Renaissance and Industrial policies as much more valuable. I also tweaked those to prevent any one policy to be out of sync with the ideas described in the earlier post.

  • Tradition - mainly Growth
  • Liberty - Mainly Tile Improvement
  • Might - mainly expansion
  • Piety - exclusively religion (no point taking it if you don't have one)
  • Patronage - almost exclusively Diplomacy
  • Aesthetics - exclusively Archaeology
  • Industry - exclusively Naval (every civ has it in top tier)
  • Imperialism - exclusively Offense
  • Rationalism - exclusively Spaceship

I'll point out that while policies can be exclusive like this, there should be plenty of replay value with this approach because anything can happen to adjust these. Religion alters policy choices because Piety is more valuable if you have a religion, and Industry is being offered as an alternative to a more focused approach by turning your economy up to 11.

Also, while my last test game was fantastic for economic development of the civs starting out, the new patch has hopefully fixed some serious AI problems with fighting wars. I literally waltzed into Berlin without any incident even though they had a navy capable of bombarding me and Musketmen that fled.

New policy flavors attached.

New game is going to be:

Nationalists
Japan (me)
Mongolia
Imperialists
China
Russia
Federalists
India
Korea
Hegemony
Arabia
Persia

Going Asian this time
 

Attachments

Any other news on this? Was some of this work incorporated into the main build?
 
I've been playing with it two games and AI seemed to behave quite well. Anything I should look into?
 
Cool to see that you want to try it out!

I've been pretty busy lately but I'll upload the latest version and an updated CoreDefines file that has the very important zero Flavor Randomization change. I have these in the appropriate file types in the .RAR file.

Instructions for those who don't know where they go
CoreDefines is in the BalanceChanges folder
Policy Flavors goes in BalanceChanges > Policies > Flavors folder
Leader Flavors goes in the AI > Flavors folder
Personalities goes in the AI > Diplomacy folder
 

Attachments

Has anyone been able to beat the AI past emperor difficulty? If so, any tips?

The AI plays much more like human players with the new flavors and personalities, so they can really take advantage of the extra bonuses they get to just slaughter. I feel like I need to expand aggressively since taking cities early on is so hard with the massive military the AI keeps in their pocket, but everyone just DOWs and it's a huge pain. It's also pretty much impossible to win with under 3 cities unless you're Venice or something. Thoughts?

Also, bravo with the changes! Especially with the updates to defensive pact and military logic, the game has gotten so much more fun. I can't wait until the new turn time bug is fixed.

EDIT: Tradition seems really, really strong right now. AIs that don't start Tradition seem to fall off hard.
 
Civs are properly jealous of each other and ambitious. War , and the paranoia of development before it, is a sensible response to anyone who develops too far in isolation. Seeing AIs do in FFAs what Humans do in FFAs - build armies and prolong the Dark Ages if they gotta do it - is brilliant.

I'm jumping to conclusions all over this post. I know that. It's what I do. I like to look ahead, and I want to show you where I see this situation heading.

What's missing are flavours for AIs to try bipartisan economic plans - building caravans instead of libraries, offsetting the attitude of militarization to despise weak civs and respect equally strong ones (modulo loyalty/forgiveness flavour), and overall a system of recognizing the profit of exacting tribute, necessitating grand strategy subroutines in which a civ provides tribute: relationships of submission.

I would further say, though this is speculative now and certainly not the correct move without more data, that the response to then rebalance things like Tradition, if it does dominate, is to slow down science. The omni-war and fear in every direction is a great recreation of how it really works to be in an FFA between players who rationally know that every advantage snowballs. There are just two things to consider: Players who know that a temporary alliance can remove their rivals and reduce risk, and players who know that turtling for a little bit of time will unlock the next tier of science and make you invulnerable. The AIs have not enough knowledge of the former , and the game is ill-positioned in making correct the strategy of the latter.

The problem is caused by the slingshot. You can 'break out' to technological advancement too sleekly with just a little bit of turtling when you have a growth advantage. War needs to be given more time to operate, not buffed per se. And the third alternative to turtling and war - bipartisan development, developing specifically at a suboptimal rate to appease a rival - if buffed, will, I think, answer the imbalance.

The next tier of science needs to be kept at a difficulty to reach just far enough in time so that doubling down on it with Growth+turtling won't get you there all by itself in time before the FFA can pounce on you, but not so far away that no one can ever approach tech. The Community Balance Patch's tech tree, with its extremely flattened topology (extremely flattened), is good at preventing all forms of tech slingshots or breakout strategies. This is the opposite of what made Civ IV fun, but science just works like that in Civ V at the core, so the opposite dynamic very well is the fun one here.
I believe the topology will afford the necessary balance change with a very subtle tweak.


Of course, if we're not doing a Strong A.I. General Intelligence engine here (we're not), then just hardcoding AIs to pick Tradition in CBP could be entirely sensible since they're psychotic antisocial wrecks anyway. :p If the bonuses were additive to difficulty handicaps (I think they are already?) it would ameliorate the Deity challenge situation.
 
I've played two games on emperor with this and the AI did really well.
 
Babylon belongs to Hegemony.
Nebuchadnezzar doesn't focus on Scientific victory, right?
 
Yeah, Hegemony doesn't have Spaceship flavors so they'll avoid that victory. Babylon probably needs to change to one of the others, either Nationalist or Imperialist.
 
Back
Top Bottom