Win Conditions

[Redmeth]

I found this interesting and decided to post it here.
I inserted your win conditions in my ongoing Sauromatae campaign and some years later I see Carhage doing something I have not seen them do before crossing the Gibraltar straits to retake Iberia (which they lost 30 years ago or so in the game)

And they're also aggressively taking on the Romans in Sicily.

I'm playing with BI and I know naval invasions are more common but I've never seen the Gibraltar stuff before. I guess it could be a coincidence but this started recently after introducing the code.
Also if in my campaign I hold all the settlements needed for the win conditions in descr_win_conditions and choose to keep playing could I still get the more complex victory conditions in EB?

[Foot]

I found this interesting and decided to post it here.
I inserted your win conditions in my ongoing Sauromatae campaign and some years later I see Carhage doing something I have not seen them do before crossing the Gibraltar straits to retake Iberia (which they lost 30 years ago or so in the game)

And they're also aggressively taking on the Romans in Sicily.

I'm playing with BI and I know naval invasions are more common but I've never seen the Gibraltar stuff before. I guess it could be a coincidence but this started recently after introducing the code.

It is a coincidence:

No, sorry. You have to start a new game for the changes in DWC to be loaded into the new savegame.

It is not savegame compatible, so the actions of carthage are not being influenced by the DWC.

Also if in my campaign I hold all the settlements needed for the win conditions in descr_win_conditions and choose to keep playing could I still get the more complex victory conditions in EB?

As I said before:

In BI this file was used for the new type of victory conditions, tailored for each faction, where a certain number of provinces AND one or two specific provinces were required for victory. This such makes sure that the AI would go for the specific provinces and thus be able to win. Basically its the modifiable version of the code that made all the factions in RTW go for the Rome province.

Foot

[Spoofa]

Ive heard that when Carthage expands all the way to the strait of Gibraltar then they would shift their focus to Iberia.

[JMRC]

I found this interesting and decided to post it here.
I inserted your win conditions in my ongoing Sauromatae campaign and some years later I see Carhage doing something I have not seen them do before crossing the Gibraltar straits to retake Iberia (which they lost 30 years ago or so in the game)
(...)
I guess it could be a coincidence but this started recently after introducing the code.
Also if in my campaign I hold all the settlements needed for the win conditions in descr_win_conditions and choose to keep playing could I still get the more complex victory conditions in EB?

This must be a coincidence, because DWC is a campaign file, so it is loaded when the campaign starts.

As for your 2nd question, when you have all the required number of settlements (which must include the specific settlements) declared in DWC, the Victory screen allows you to keep playing. This means that the conditions in the EBBS file may still be met (because they are scripted) and thus you will receive a second victory screen.

As an academic thought, I think you may get unlimited victory screens if you accomplish the objectives and keep playing, then lose one of the objectives and get it again afterwards, re-accomplishing the objectives, and thus getting another VS. I don't think the AI will be smart enough to understand that the conditions have already been met before. NOTE: I haven't tested this, so I can't say it for sure.

[Redmeth]

To stop that you could follow MAA's suggestion from the previous page:

If you set that to 199, though, you wouldn't have to worry about setting off the victory conditions and wouldn't have to cast off the complex EB victory conditions.

About setting the specific settlements to take by name and then using take_regions 199 to make sure you don't trigger them abruptly.

[JMRC]

Ive heard that when Carthage expands all the way to the strait of Gibraltar then they would shift their focus to Iberia.

This is highly unlikely, unless it's hardcoded in the game (which I also don't believe). Unless you're using DWC to make the AI go for certain objectives, the AI will move on a "turn by turn" evaluation of targets using heuristics (which can go a few turns into the future, though).
This means that if they go as far as the strait of Gibraltar, their next target will be going through the strait or moving south to the desert. The heuristics (if correctly implemented) should point to Iberia because it surely is a richer land than the desert, so the AI moves there. But if for some reason, the heuristics point to the desert, then that's where the AI will go.

[JMRC]

To stop that you could follow MAA's suggestion from the previous page:

About setting the specific settlements to take by name and then using take_regions 199 to make sure you don't trigger them abruptly.

I just made some tests and noticed that the victory conditions set in EBBS only apply to the faction controlled by the human player due to the condition:

 and CharacterIsLocal

I don't know why this was made, but if you remove this line, the conditions also apply to the AI controlled factions.

As for the DWC, it doesn't trigger victory screens for the AI controlled factions. The VS only appear to the human player.

When I removed the line above from the EBBS, the VS appeared for the AI controlled faction, but it still gave me the option to keep playing.

So, I don't think you need to put the

take_regions 199

since even if the VS appears, the player always has the chance to keep playing.

[JMRC]

Hi. This is an update to my Test Run, taken in 242BC (120 turns into the game).

In this latest count, from 69 regions conquered, 58 were objective regions.

Note: starting regions are in Italic; Conquered objective regions are in Bold; Lost starting regions are Underlined/Italic; Conquered non-objective regions are in Bold/Italic. The factions that didn't have changes since 262BC are marked in red, otherwise are marked in blue. I'm playing with Casse, so it doesn't count.

Romani
hold_regions: Latium2 Etruria Umbria Apulia Campania Aemilia Venetia Korsim Elimya Liguria Sardin Greseoallra Insubramrog Lacetania Baleares Pannonia_Illyrica Kalabria Brettia Sicilia Trinakrie Syrthim Byzacena Zeugitana Edetania Celtiberia Carpetania Bastetania Turdetania Dalmatia Illyria_Hellenike Epeiros Makedonia Attike Delta_Neilou Ioudaia Syria
non-target regions: Volcallra Aquitae Pannonia

Saka
hold_regions: Sai_Yavuga Kangha Saka_Yabgu Xiyu Dayuan Dahyu_Haomavarga Sogdiane Margiana Dahyu_Alanna Aria Baktria Paropamisadai Drangiane Arachosia Gandhara Sind Sattagydia

Arverni
hold_regions: Arvernotorg Sequallra Lugonesis Batromorgan Mrogaedu Lemorisae Armoriae Mrogaule Aquitae Volcallra Greseoallra Bellovacaea_Belgae Nervaea_Belgae Vindelicos Helvetis Rhaetia Insubramrog Venetia Latium2

Saba
hold_regions: Saba Qataban Main Tadmor Zufar Hadramaut Maketa Arabia Gerrhaia_Arabia Sinai Nabataia Syria Syria_Koile Phoenicia Ioudaia Assyrie Babylonia Charax_Spasinou

Makedon
hold_regions: Makedonia Euboia Lesbos Thessalia Peloponnesos Dardanoia Aitolia Odrysai Attike Lakonike Epeiros Chersonesos_Thraikia Illyria_Hellenike Mikra_Skythia Mysia Lydia Karia Bithynia Thraikia_Hypertera

Ptolemaioi
hold_regions: Erythraia Triakontaschoinos Thebais Heptanomis Delta_Neilou Oasis_Megale Marmarike Ioudaia Phoenicia Kypros Kilikia Pamphylia Libye Sinai Nabataia Kush Syria Kyrenaia Diamat Syria_Koile Tadmor Assyrie Babylonia
non-target regions: Arabia Augila

Arche Seleukeia
hold_regions: Syria_Koile Assyrie Babylonia Lydia Elymais Media Kappadokia Persis Gabiene Karmania Drangiane Aria Margiana Charax_Spasinou Parthava Khoarene Phrygia Sophene Adiabene Hyrkania Syria Astauene Karia Pamphylia Kilikia Kypros Phoenicia Ioudaia Tadmor Nabataia Sinai Delta_Neilou Heptanomis Thebais Marmarike
non-target regions: Gerrhaia_Arabia

Carthage
hold_regions: Zeugitana Mashiliem Byzacena Syrthim Turdetania
Bastetania Baleares Edetania Korsim Sardin Elimya Carpetania Turdulia Lacetania Volcallra Greseoallra Liguria Etruria Latium2 Sicilia Trinakrie Mauretania_Massaesili Mauretania_Tingitana
non-target regions: Numidia Phasania

Parthia
hold_regions: Khwarazm Hyrkania Astauene Dahyu_Daha Aghvan Parthava Gabiene Karmania Gedrosia Adurbadagan Media Elymais Khoarene Babylonia Adiabene Sophene Assyrie
non-target regions: Dahyu_Mazsakata Dahyu_Alanna Saka_Yagbu

Pontus
hold_regions: Kappadokia_Pontika Galatia Paphlagonia Bithynia Pontos_Paralios Pokr_Hayk Kappadokia
non-target regions: Chersonesos_Thraikia

Aedui
hold_regions: Mrogaedu Mrogaule Lugonesis [U]Insubramrog8/U] Sequallra Arvernotorg Batromorgan Lemorisae Armoriae Aquitae Volcallra Greseoallra Bellovacaea_Belgae Nervaea_Belgae Vindelicos Helvetis Rhaetia Venetia Latium2
non-target regions: Pannonia_Illyrica

Sweboz
hold_regions: Swebolandam Kimbrolandam Hattolandam Heruskolandam Habukolandam Mrogbonna Skandza Rugolandam Silengolandam Vindelicos Bastarnolandam Sequallra Helvetis Bellovacaea_Belgae Nervaea_Belgae Rhaetia Noricae
non-target regions: Eravacouw

Casse
hold_regions: Cassemorg Corieltauvae Cornovae Cambriae Brigantiae Caledryn Cruddain Erain

Hayasdan
hold_regions: Hayasdan Adurbadagan Kartli Pokr_Hayk Egrisi Aghvan Adiabene Sophene Pontos_Paralios

Getai
hold_regions: Getia Thraikia_Hypertera Scorcouw Getia_Koile Mikra_Skythia Bastarnolandam Coutinoe Eravacouw Dardanoia Odrysai Chersonesos_Thraikia

Koinon Hellenon
hold_regions: Lakonike Rhodos Attike Peloponnesos Thessalia Euboia Krete Aitolia Makedonia Epeiros Illyria_Hellenike Dalmatia Odrysai Chersonesos_Thraikia Mikra_Skythia Bosporion_Tyrranesis Taurike_Chersonesos Bithynia Mysia Lesbos Lydia Karia Kypros Kalabria Brettia Sicilia Trinakrie Greseoallra

Baktria
hold_regions: Baktria Dayuan Sogdiane Dayhu_Daha Margiana Dahyu_Haomavarga Xiyu Sai_Yavuga Saka_Yabgu Kangha Dahyu_Alanna Aria Paropamisadai Drangiane Arachosia Gandhara Sind Sattagydia

Sauromatae
hold_regions: Dahyu_Aursa Dahyu_Roxsalanna Dahyu_Yazyga Dahyu_Siraca Maeotis Dahyu_Yugra Budinije Neurije Dahyu_Thissakata Dahyu_Alanna Dahyu_Mazsakata Bosporion_Tyrranesis Taurike_Chersonesos Skythia

Lusotannan
hold_regions: Lusitania Turdulia Celtiberia Galaecia Asturia Cantabria Carpetania Turdetania Bastetania Edetania Lacetania Baleares

Epeiros
hold_regions Epeiros Illyria_Hellenike Kalabria Dalmatia Brettia Sicilia Trinakrie Dardanoia Makedonia Odrysai Mikra_Skythia Chersonesos_Thraikia Thessalia Aitolia Euboia Attike Peloponnesos Lakonike Scorcouw Pannonia Pannonia_Illyrica Elimya Thraikia_Hypertera


Check the Test Run pic here.

[Imperator]

Wow- great find! This could mean a whole WORLD of improved campaign map AI behavior! Well done!

[Ludens]

Has there been any progress on this? It certainly sounds like a potential remedy for unrealistic A.I. expansion.

[Imperator]

I think a moderator should probably move this to the EB forums, where it belongs so it could get the attention it deserves. If it works the way it seems it does, then the EB team should definitely look into it.

[MarcusAureliusAntoninus]

Does it work in v1.5?

[Anarzius]

...

[Puupertti Ruma]

I'm going to try if this indeed works with version 1.5 later today. First I will do some test runs playing Casse and doing nothing but pressing end turn and taking screenshots. After couple of those, I will insert the DMC exactly as made by JMRC and see if the AI expansion differs.

I'dd do it right now, but alas I have some other things to do so this has to wait for the evening.

[Puupertti Ruma]

Ok, it took a while but after one game of only pressing end turn and with slightly modded JMRC's DWC it seems that the DMC doesn't do anything. I pressed the end turn until about 212 BC IIRC and during this Makedonia first seemed to win the Greek wars but then it took of and went to the steppes, leaving both Epeiros and Koinon alive. So in other words, Makedonia did not try to conquer it's victory conditions before going to some weird places.

Fortunately this isn't conclusive yet! I mentioned I slightly modded JMRC's DWC. The only thing I modded was that every faction would have to own 234 provinces. That might put the AI off of the region owning requirements as it also has to conquer huge amount of provinces, and of course it could nullify the whole DWC file as the region count is more than the hard coded limit (which is 199 right?).

To remedy those things I started another campaign earlier today with quite extraordinary DWC. The point is to make the victory conditions so weird that if the AI actually follows them this should show quite clearly on the map. Here is the DWC so others can see it, correct it and perhaps even use it to figure out if the DWC 1.5.

;Romani
seleucid
hold_regions Latium2 Etruria Umbria Apulia Campania Lakonike Attike Aitolia Peloponnesos
take_regions 10

;Saka
pontus
hold_regions Sind Sattagydia Gandhara
take_regions 10

;Arverni
scythia
hold_regions Turdetania Carpetania Turdulia
take_regions 10

;Saba
saba
hold_regions Karmania Gablene Karmania
take_regions 10

;Makedon
macedon
hold_regions Neurije Budinije Skythia
take_regions 10

;Ptolemaioi
numidia
hold_regions Kyrenaia Syrthim Byzacena Zeugitania
take_regions 10

;Arche Seleukeia
romans_julii
hold_regions Phrygia Lydia Karia Pamphylia Kilikia Kypros Sophene Syria Syria_Koile Assyrie
take_regions 199

;Carthage
egypt
hold_regions Liguria Etruria Latium2 Sicilia Helvetis
take_regions 10

;Parthia
parthia
hold_regions Saba Hadramaut Arabia Tadmor Babylonia Elymais Gablene
take_regions 10

;Pontus
carthage
hold_regions Kappadokia_Pontika Paphlagonia Pontos_Paralios Pokr_Hayk Kappadokia Galatia Bithynia
take_regions 10
;Aedui
gauls
hold_regions Rhaetia Noricae Pannonia Scorcouw Dardanoia Odrysai Bithynia Galatia
take_regions 10

;Sweboz
germans
hold_regions Liguria Aemilia Etruria Umbria Latium Apulia Venetia Campania
take_regions 10

;Casse
britons
hold_regions Cassemorg Corieltauvae Cornovae Cambriae Brigantiae Caledryn Cruddain Erain
take_regions 10

;Hayasdan
romans_scipii
hold_regions Dahyu_Aursa Dahyu_Roxsalanna Dahyu_Yazyga Dahyu_Siraca Dahyu_Yugra Dahyu_Thissakata Dahyu_Alanna Dahyu_Mazsakata Maeotis Bosporion_Tyrranesis Taurike_Chersonesos Skythia Budinije Neurije
take_regions 10

;Getai
dacia
hold_regions Krete Rhodos Karia Pamphylia Kilikia Galatia
take_regions 10

;Koinon Hellenon
greek_cities
hold_regions Egrisi Kartli Aghvan Hayasdan Maeotis Kypros Syria Phoenicia
take_regions 10

;Baktria
romans_brutii
hold_regions Sogdiane Dayuan Kangha Saka_Yabgu Sai_Yavuga Dahya_Alanna Dahya_Aursa Dahya_Yugra
take_regions 10

;Sauromatae
armenia
hold_regions Neurije Lugouw Venedu_Tauta Mrogbonna Vindelicos Coutinoe Eravacouw Pannonia Scorcouw
take_regions 10

;Lusotannan
spain
hold_regions Lusitania Galaecia Asturia Cantabria Turdulia Carpetania Edetania Lacetania Baleares Erain Cruddain
take_regions 10

;Epeiros
thrace
hold_regions Epeiros Makedonia Aitolia Euboia Attike Peloponnesos Lakonike Krete Rhodos Marmarike Heptanomis
take_regions 10

Notice how my modded DWC is full of errors? If you do, please point them out so we can count out the possibility of DWC not working because of syntax errors.

Also I'd like to point out some things about JMRC's research in to the DWC. First and foremost, his test games weren't in my opinion long enough. Those posted here lasted to 252 BC at their longest. In my first campaign everything looked fine until little after 240 BC the Makedonia made it's plunge for the steppes. Second problem is the DWC itself. It doesn't differ enough from the normal unmodified AI expansion patterns. Epeiros kicked out of Italy by Rome? In my first unmodded reference game, Rome did it by the end of 270 BC! Third problem is the lack of repeats. Two games don't tell enough about the norm, to see the norm there must be more repeats.

All this in mind, I'd like to have some help with trying this thing out, not only with 1.5, also with BI. I my self don't have BI, so if you could test my DWC with it we could see if the DWC actually has any influence on anything, even in BI.

PS. My second test game is now in 237 and it seems my edited DWC has no effect at all. I will also post raw research data of my first game to the AI expansion thread when I have edited them to publishable form. This might take a day or week, depending on RL stuff.

:edit: Some wordings and some such:

[Puupertti Ruma]

Hmm.... I just realised I hadn't deleted the map.rwm. I guess that should nullify my test result pretty much completely. Nice. About 6 hours of just pressing end turn down the drain.

[madmatg]

im really interested in this hope it works!

[JMRC]

Hi!

Nice to know that someone went back to look at this thread. I always used BI in my tests, so DWC might not work with 1.5 (which is what you are confirming).

However, I would like to add a few issues, that I found since the time that I made my last post here:

- It seems that the I slightest syntax error in the DWC file turns it into unreadable, which means it won't be taken into account. So, please, start with few (and correctly introduced syntax and regions) before adding the rest.

- It also seems that, despite using BI, the DWC file which is read is the one found in the

imperial_campaign.txt

file. I haven't confirmed this one.

- The way the AI has to achieve its objectives might not be direct. It probably has to do with enemy strength on surrounding settlements, etc. Which means that they don't rush into the objectives, and if the objective settlement is very far, the AI might get "lost" in pathfinding and never go for that objective.

- The simplest way to test this is to give objectives which are 2 regions deep away from the "border" and without massive defenses. The AI should not get lost in pathfinding and might consider it a good target. After that, you can start making more difficult situations.

[Puupertti Ruma]

I've edited my DWC to have a good syntax but due to RL stuff haven't had a chance to play EB in almost a month and haven't had the time to test it. I've constructed it in a way that quite a lot of factions have a "trail" of victory conditions. That means there is one near or even next to their starting territories and from that there is a trail to follow to some very odd place where the AI faction should never wander without guidance. For example, Aedui have a trail from Helvetia to Galatia, with a neighboring victory condition provinces next to each other.

Also, what do you mean by

- It also seems that, despite using BI, the DWC file which is read is the one found in the

imperial_campaign.txt

file. I haven't confirmed this one.

I couldn't find file called "imperial_campaign.txt" from my EB folder, or even RTW folder. Did you mean EB\Data\world\maps\campaign\imperial_campaign folder maybe, or is that a BI only thing?

And last, my latest DWC.txt for interested people to test, mock or enjoy:

Spoiler Alert, click show to read: 

;Romani, goes for greece
seleucid
hold_regions Aemilia Venetia Pannonia_Illyrica Dalmatia Illyria_Hellenike Epeiros Makedonia Thessalia Aitolia Attike Peloponnesos Lakonike
take_regions 8

;Saka, goes for India
pontus
hold_regions Sind Sattagydia Gandhara Gedrosia
take_regions 4

;Arverni, goes for Iberia
scythia
hold_regions Turdetania Carpetania Edetania Lacetania Volcallra
take_regions 6

;Saba, goes for Persia
saba
hold_regions Arabia Charax_Spasinou Elymais Gabiene Persis
take_regions 4

;Makedon, goes for the amber route
macedon
hold_regions Chersonesos_Thraikia Odrysai Mikra_Skythia Getia Thraikia_Hypertera Scorcouw Eravacouw Mrogbonna Silengolandam Rugolandam Venedu_Tauta
take_regions 7

;Ptolemaioi, goes for Carthage
numidia
hold_regions Kyrenaia Phasania Syrthim Byzacena Zeugitana Mashiliem
take_regions 10

;Arche Seleukeia, goes for Steppes
romans_julii
hold_regions Adiabene Hayasdan Egrisi Dahyu_Yugra Dahyu_Mazsakata Dahyu_Aursa Dahyu_Daha Dahyu_Yazyga Dahyu_Roxsalanna Dahyu_Siraca Dahyu_Thissakata Dahyu_Alanna Dahyu_Haomavarga
take_regions 18

;Carthage, goes for Italy
egypt
hold_regions Sicilia Elimya Trinakrie Brettia Kalabria Campania Apulia Latium2 Umbria Etruria
take_regions 10

;Parthia, goes for Persia and Babylonia
parthia
hold_regions Hyrkania Khoarene Charax_Spasinou Elymais Gabiene Persis Babylonia Assyrie
take_regions 7

;Pontus, goes for Syria
carthage
hold_regions Sinai Ioudaia Phoenicia Syria Syria_Koile Kilikia Kappadokia
take_regions 5

;Aedui, goes for Galatia
gauls
hold_regions Sequallra Helvetis Rhaetia Noricae Pannonia Scorcouw Thraikia_Hypertera Odrysai Chersonesos_Thraikia Bithynia Galatia
take_regions 6

;Sweboz, goes for Italy
germans
hold_regions Kalabria Campania Apulia Latium2 Umbria Etruria Aemilia Venetia Noricae Mrogbonna Hattolandam
take_regions 6

;Casse, goes for Iberia
britons
hold_regions Carpetania
take_regions 6

;Hayasdan, goes for persia
romans_scipii
hold_regions Charax_Spasinou Elymais Gabiene Persis Babylonia Assyrie Sophene
take_regions 6

;Getai, goes for modern Turkey
dacia
hold_regions Karia Pamphylia Kilikia Lydia Phrygia Mysia Bithynia Chersonesos_Thraikia
take_regions 4

;Koinon Hellenon, goes for Kaukasus
greek_cities
hold_regions Karia Pamphylia Kilikia Lydia Phrygia Mysia Bithynia Chersonesos_Thraikia Kappadokia Pokr_Hayk Hayasdan
take_regions 8

;Baktria, goes for the steppes
romans_brutii
hold_regions Dahyu_Yugra Dahyu_Mazsakata Dahyu_Aursa Dahyu_Daha Dahyu_Yazyga Dahyu_Roxsalanna Dahyu_Siraca Dahyu_Thissakata Dahyu_Alanna Dahyu_Haomavarga
take_regions 8

;Sauromatae, goes for Pusta
armenia
hold_regions Skythia Bastarnolandam Eravacouw Scorcouw
take_regions 5

;Lusotannan, goes for Ireland and Normandy
spain
hold_regions Armoriae Mrogaule Cruddain Erain
take_regions 6

;Epeiros, goes for modern Turkey
thrace
hold_regions Karia Pamphylia Kilikia Lydia Phrygia Mysia Bithynia Chersonesos_Thraikia
take_regions 7

:edit: Made the DWC a "spoiler".

[Tarkus]

BUMP...this is a really interesting and very promising thread that deserves some more attention...JMRC / Puupertti, anything new?!

[C.LVCIANVS]

JMRC this is woonderful!!!!
The order of listed provinces to be conquered could influence AI expansion? For ex: if I put korsim before massilia in the list the AI will try to take korsim first then moves to massilia?
Everything is possible. I'm using imagination!

EDIT: And what about the "outlive_faction" command, used for vanilla RTW imperial_campaign?

[Anarzius]

...

[JMRC]

Hi! Sorry, I didn't see that there was still some action in this thread.

First, you're right, I meant imperial_campaign directory.

Second, many people say that DWC doesn't work, but I'm still not convinced of that, because I think I saw a pattern of movement during my testing. I do recognize that sometimes the AI didn't went straight to the targets assigned for it, but I reckon it might be related to easier targets around, which draw the AI.

I'll try to make a few more tests.

[Anarzius]

...

[Callicles]

To really test to see if it works like we hope, you should give factions strange win conditions. For example, see if you can get the Seleukids to focus on conquering the Pahlava, Baktrians, and Saka. Meanwhile have the Ptolemaioi focus on Carthage and have Pontos and Armenia focus on conquering Syria and Babylon.

If the Win Conditions file works like we hope it does, then after 20 turns you should begin to a big difference as the Koele-Syria region is vacated by the traditional Ptolemaioi v. Seleukid war and replaced with Pontos v. Hayasdan.

Edit: I see I missed the post where this was already suggested. My apologies.

[konny]

It seems to work with RTW as well: I had seen on the modding forum that at least one mod uses the BI win conditions with RTW. I had made a simple test myself: using the code for BI in an RTW.exe run game. I had changed the conditions for >>Selucids<< (the EB Romans) to conquer Kalabria (Taras) and have 6 towns. The other VCs remained unchanged.

- In 6 attempts the Romans had send their army north to Segesta once

- In the other 5 tests they had massed their forces in the south without doing anything north of the Rubicon.

- Of these 5 they had attacked Taras at first in two tests

- Attacked Rhegion first in two tests

- Attacked Rhegion and Taras in the same turn in one test.

I had run this last test for a longer time and in this one they succeded in Rhegion in the second attempt but not in Taras. After that they attacked Taras every turn without doing anything else.


This is not a prove that it 100% works, because for that they should have attacked Taras first in 6 out of 6 times. On the other hand the AI might also take other things into account than just its VCs. But experinced EB might find it notable that the AI Romans did not atttack Bononia in 6 out of 6 tests and that they went south in 5 out of 6 times.


I had tested other things too: I had given Epeiros Makedonia as a target, Makedonia Attika and KH Krete; always with the respective "outlive" command. All three factions have an army next to the respective town from the start but in no occasions ever the AI uses this army to siege the towns (Pella, Athen and Kydonia). What should I say, they didn't do so this time too:

- The KH army didn't do anything at all
- Pyrrhos marched to the mountains and became pinned there
- Antigonos followed him, defeated his army and became pinned on the same position (pathfinding problems in the hills west of Pella?).

Later the Maks trained a new army in Chalkis and attacked Athens (but what else should an army from Chalkis do?), while the Epeirote sent another army across the mountains that did not dare to attack Pella because there was a (second) Makedonian stack guarding the town.

But that must not mean anything: It might be logical for a human player to do use these armies to attack the nearby towns when beeing ordered so; but the AI works different and might as well march a stack from Attike to defeat an enemy army in Makedonia and move in the same time the army from Pella to attack Athens. It might be better for the AI to place all units in towns at the beginning of the game that it can compose its own stacks and asign them targets according the hardcoded preferences?



I definitly think that further tests with the win conditions on RTW will be usefull and, after the results of the Roman tests, I am convinced that the BI codes can be used for RTW as well and that they might change the AI behaviour up to about 80%.

[larsbecks]

Is there a command to not take a region or avoid taking it?

[konny]

Is there a command to not take a region or avoid taking it?

No, there isn't.


Further tests:

I had made the three Northen Italian towns targets for the Aedui. I had made one test and it didn't work. On the other hand, I was sure that it will not work in advance because in all my EB campaigns the Aedui only attacked Bononia once and the other two never, even though they are rebell. May be, before the next tests I will reduce the garrison there, or there is another reason why the AI does not recognize them as legal targets for the Aedui.


descr_strat

Not the same but related: I have noted that for all factions, save for Armenia(!?!), the slave faction has the highest priority. Can it be that this is the reason for "the hardcoded preference of the AI towards the Eleutheroi"? I have made one test with slightly altered preferences: Epeiros and KH got Makedonia as highest priority and Makedonia got KH as highest and Epeiros as second highest. The test was run without using victory conditions. After 5 years neither the Makedonians did attack Serdike nor Epeiros Dalmatia. Both factions had large forces in the mountains between Makedonia and Ill. Helenike, what were fighting each other.

Other changes that seem to be important for the AI:
- The Epeirote army at Pella should start in or around Epidamnos because the AI is always trying to move it there, but usually the stack becomes pinned after the first turn.

- Antigonos army must be moved to either Demetrias or Pella, while the forces next to Korinth must become part of the garrison. In every test Antigonos merges that army with his own force and moves the entire stack north, leaving Greece open for KH.

- The Greek army from Krete must be moved to the mainland because the
AI has absolutly no idea what to do with it (and the player will most likely move it back in the first turn anyway).

[larsbecks]

I've noticed the Greek Army will take Krete but only after a very long time (in one of my games, 35 years). Or it will try and fail and just sit there. I've modded my game so that the Maks will always have a good chance against the greeks. I made the unit spawning for human only players to be permanent for human and AI and I changed their metality to "bureucrat". Also, once they beat the greeks and expanded all the way to Byzantion they took Epidamnos by sea attack, except once they landed they stood around for a couple of years standing between two Epeiros armies before anything happened.
And the lack of Hayasdan expansion seems to be a combination of the 300 rating for aggression towards elutheroi and their "fortified" AI.

[Callicles]

As always, Konny's testing is superb. I've been away from my EB computer for a while, but I'll report back on my own testing tomorrow.

I plan to make Carthage have first priority against Rome and second against Iberia. Likewise, Rome will have first priority against Rome and second against Iberia. If Konny's testing rings true, this should lead to a Punic war in Sicily and after some time, a proxy war in Spain.

I'll give Epeiros a priority against the city of Kyrene to see if they make the attempt to cross the sea.

I'll give Pontos a priority towards Trapezeus and then Cheronesos to, again, see if they cross the sea.

Finally, I'll give Parthia, Baktria, and Armenia, priority against the Seleukids, and give Parthians a target priority of Babylon to see if they cut through the middle of Persia.

Report to follow tomorrow night with screen shots.