Home Trending Latest threads New posts Developer diaries

A new AI system for AGCEEP

Jump to latest Follow Reply
  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Just for information, traders values in smallCRC AI files:
0 => 20
1 => 20
2 => 50
3 => 100

If we apply the same changes as for smalltrade AI files, smallCRC0 and smallCRC3 should be amended. But I don't know why traders value was 20 for smallCRC0 and different from smalltrade0.
 
YodaMaster said:
Just for information, traders values in smallCRC AI files:
0 => 20
1 => 20
2 => 50
3 => 100

If we apply the same changes as for smalltrade AI files, smallCRC0 and smallCRC3 should be amended. But I don't know why traders value was 20 for smallCRC0 and different from smalltrade0.
Click to expand...

what does CRC mean?

why have we a set of trader file ( which i can understand) and a set of CRC which I do not understand.

EDIT: CRC means small traders that went to counterreform religion, then i agree that they should be the same as the AI trade file in refrence to the traders number

But who gets the CRC files and how?
 
Difference between both sets is the counterreform = yes item. Without this, an AI Catholic country will never become CRC.

IIRC, Fodoron attributed them and this attribution was tweaked for the 1648 setup (just before the Tolerance event) => most of Italian countries (except Naples and Venice) and Archbishoprics in Germany (in order to avoid conversion to Protestantism).
 
YodaMaster said:
Difference between both sets is the counterreform = yes item. Without this, an AI Catholic country will never become CRC.

IIRC, Fodoron attributed them and this attribution was tweaked for the 1648 setup (just before the Tolerance event) => most of Italian countries (except Naples and Venice) and Archbishoprics in Germany (in order to avoid conversion to Protestantism).
Click to expand...

we agree then changes for one set (traders ) should be copied over to CRC set
 
Info below on my findings for the AI file.

the typical AI file (see below) has 2 multiplying sections which I coloured. IMO these sections never work at the same time. The "green" conquer section is active only in times of war in which the "orange " garrison section is "switched off".
Its reverse in times of peace, that is Garrison is on , while conquer is off.


#
# Europa Universalis. French AI
#
# Huguenots AI
#
region = {}
expansion = 0
tradingpost = 50
neighbour = 10
enemies = 0
traders = 10
monopoly = 40
war = 30
ferocity = no
combat = { FRA FUC }
base = 1.0
front = 5.0
conquer = {
enemy = 1.0
supply = 0.2
distance = 0.4
owner = 5.0
notsupply = 2.0
base = 2.0
}
garrison = {
fortress = 1.0
strategic = 3.0
size = 1.0
supply = 2.0
war = 5.0
}

this can be seen in many many wars.

I did experiment and removed the garrison section form the AI file, with very bad effects ie, the AI never moved even in war.
I also kept the garrison section and made all numbers 0, same effect because its a multypling number.

I recently have tested in the garrison section the number 0.1 for all 5 points and it has some very interesting things happen, for one, in peace the state would position its troops in big numbers along the frontier of a neighbor who either has one of its cores or has a bad relationship with it. Its amusing watching it reposition its troops based on changes in its events. This does benefit the AI in that when it goes to war its troops are already combined.

One minor side issue is that it sometimes leaves its troops in provinces which are part of its alliance partner(s)
 
YodaMaster said:
Interesting result. What do you recommend then?
Click to expand...

i am experimenting at the moment with this setting for garrison , for HUN, VEN and ENG HYW files

garrison = {
fortress = 0.1
strategic = 0.1
size = 0.1
supply = 0.1
war = 0.1
}

I still do not understand the Vanilla desc. for garrison especially the war one, I presume it means ..........AI against the human

# Modifiers for garrison plans
garrison =
{
# low values keep troops stationed in low or no fortress provinces, high values keep troops near the largest fortress you have
fortress = (0 - 100.0)
# low values don't acknowledge the strategic value of a province in deciding where to garrison, high values do
strategic = (0 - 100.0)
# low values don't care if a province is big or small, high values do
size = (0 - 100.0)
# low values don't care if a province has a good supply amount, high values do
supply = (0.0 - 100.0)
# low values don't care if a province is occupied by you, high values do
war = (0.0 - 100.0)
}
 
#
# Murad III of the Ottoman Empire AI
#
region = { }
area = { }
expansion = 0
tradingpost = 50
neighbour = 20
enemies = 0
traders = 100
monopoly = 10 #-500
war = 30 #10
ferocity = no
counterreform = no
combat = { KNI BUL SPR QAR KAR KAL HAB HUN SER BOS ALB MAM CRO }
base = 0.1
front = 5.0
conquer = {
enemy = 5.0
supply = 0.1
distance = 6.0
owner = 1.0
notsupply = 2.0
base = 0.3
}
garrison = {
fortress = 5.0
strategic = 12.0
size = 1.0
supply = 9.0
war = 8.0
}


...........................................................


#
# Mehmed III of the Ottoman Empire AI
#
region = { }
area = { }
expansion = 0
tradingpost = 50
neighbour = 20
enemies = 0
traders = 100
monopoly = 20
war = 30 #0
ferocity = no
counterreform = no
combat = { KNI BUL SPR QAR KAR KAL HAB HUN SER BOS ALB MAM }
base = 0.1
front = 5.0
conquer = {
enemy = 5.0
supply = 0.1
distance = 6.0
owner = 1.0
notsupply = 2.0
base = 0.3
}
garrison = {
fortress = 5.0
strategic = 12.0
size = 1.0
supply = 9.0
war = 8.0
}
 
I thought this was proposed by Garbon only if Knights still own Rhodes at this time. Otherwise, Turks don't need it.

As I said in the other thread, if the Turks are not interested in fighting the Knights, there could be a reason not directly tied to combatlist.

But it could be because a too powerful HAB is in the combatlist of the mentioned AI files too.
 
Last edited:
YodaMaster said:
I thought this was proposed by Garbon only if Knights still own Rhodes at this time. Otherwise, Turks don't need it.

As I said in the other thread, if the Turks are not interested in fighting the Knights, there could be a reason not directly tied to combatlist.

But it could be because a too powerful HAB is in the combatlist of the mentioned AI files too.
Click to expand...

i placed these changes before Garbon comments
 
combat list in suleyman 1520 is doubled
 
YodaMaster said:
What do you mean? I don't undestand.
Click to expand...

#
# Suleyman I 'the Magnificent' of the Ottoman Empire AI
#
region = { }
area = { }
expansion = 0
tradingpost = 50
neighbour = 20
enemies = 0
traders = 100
monopoly = 10 #-500
war = 50
ferocity = no
counterreform = no
combat = { BUL ALB HAB HUN VEN CRO LIT GEO TRI ADE OMA KNI AKK MAM QAR ALB BOS RAG SER KAR TRE ATH BYZ CAN DUL TEK SPR TUN XHO POR }
base = 0.1
front = 5.0
conquer = {
enemy = 5.0
supply = 0.1
distance = 6.0
owner = 1.0
notsupply = 2.0
base = 4.0
}
garrison = {
fortress = 5.0
strategic = 12.0
size = 1.0
supply = 9.0
war = 8.0
}
 
i see no reason why we have BUR included in the ENG ai list below.
I assume its in for a deflection of ENG forces from warring DAU/FRA

I tested this below and have found no issue..............i would like others to test.

note , I am experimenting with garrison , and have forgotten the "old" numbers

#
# English HYW AI
#
continent = { }
region = { }
area = { }
expansion = 0
traders = 100
monopoly = 10 #-500
war = 20
ferocity = no
combat = { EIR FRA BUR }
base = 0.1
front = 9.0
conquer = {
enemy = 5.0
supply = 5.2
distance = 18.0
owner = 4.0 9.0
notsupply = 3.0
base = 9.0
}
garrison = {
fortress = 1.0
strategic = 9.0
size = 1.0
supply = 1.0
war = 1.0
}
 
Last edited:
post 1116 edited/amended
 
Still not entirley happy with post 1116, I decided to go back to old, old files and see what "would happen" if part or all of these where used in this beta version.

I decided on Ribbon22 HYW file and found some interesting "theories", firstly the english are more active (fleet) , but the army marches a lot in france and there is a slightly higher number of rebels appear in England.
Below is the file, ( mostly ribbon22 idea with the AGCEEP trading, monopoly system)

#
# English HYW AI
#
continent = { }
region = { }
area = { }
expansion = 0
traders = 100
monopoly = 10 #-500
war = 20
ferocity = no
combat = { FRA }
base = 5.0
front = 1.0
conquer = {
enemy = 0.1
supply = 5.0
distance = 10.0
owner = 1.0
notsupply = 5.0
base = 10.0
}
garrison = {
fortress = 1.0
strategic = 10.0
size = 5.0
supply = 10.0
war = 1.0
}


the base and front are reversed, this leads to a quicker "shipping" of troops to france.:confused:

On testing about 4 games I was pleased , I then amended the garrison/strategic to 20 and ENG became very placid, I then amended the 20 to 5 and it was slightly better than the "20". I returned to the number 10 and again the AI was "active".

hard , hard work these file

Anyway can someone do some test with the file in post 1116 and this file and give me some notes. Its pointless always "leaving my head on the chopping block".
 
You must log in or register to reply here.
Top Bottom
Back