Redux: AI-expansion & trends

[tmodelsk]

Topic: the AI-expansion on the stratmap and related trends of the RX-AI, examined and viewed over multiple campaigns. Set on either "standard" or "veteran"-difficulty. No engine restrictions (both VI/v.2.01 and v.1.1-engines relevant). Using RXB1004(e) and beyond...

Context: are there AI-trends and expansion patterns in RX-campaigns? What do these look like? If any? What conclusions can be drawn, using samples from multiple campaigns in such an investigation. This is an open discussion, free for all redux-players and interested to participate by submitting samples/screens, findings/observations, comments or other input on the subject. Please respect the max size for screens as 800x600 pixels, otherwise use spoilers.

------------------------------------------------
ORIGINAL POST....
------------------------------------------------

Hi. As Axalon recently asked for some campaigns progress reports for AI and overall mod analysis, I want to share those results I've got. I've made separate thread for it because main sticky thread IMO is hard to read ( 20 pages). Axalon - feel free to move those my writings or change the thread subject if necessary. I don't know if those reports will give any valuable feedback, I'll just post & write some observations & questions. Of course - SPOILER ALERT.

All those reports are played as Poland. always standard unit size, MTW:VI with 1004e. Poland in MTW:Redux is small weak
nation, placed on north-east map corner, early Poland has not any signifacant impact on overall world situation

First 5 games were on standard difficulty, I abandoned them on some early point for some reasons (I learned something and started new better one :-) )

First was point scenario (not domination), year 809:

What is interesting: Spain is alive :-), they managed to plug Africa-Iberian Pennisula landbridge.
Moore invasion went east, Moors has devastated Saracens and now they'are entering Asia Minor and are beating Bizancjum.
LIthuanians won competition with Russians, France has beaten Germany & company.
This world situation is unique across all my games - unique because of Spain being able to defeat from Moors.

-----------------------------------------------------------------------
Next 4 games are the same conditions (standard diff), except scenario is default : domination.

I loosed somewhere pict & saves, I will post only one, but overall main situation was the same:
MOORS ARE ATTACKING EUROPE, and they progress is realy fast!
845:


This is my current campaign, veteran difficulty, year 746:


On next 2 games on veteran general situation is the same: strong Moors aggression through Spain & the Saracens aggression on Bizancjum.
Axalon, is it intentional ?

Of course some regions situation where different, sometimes Russions won over Lithuanians, sometimes is vice versa, the same with France-German-Lombard-Italians.
But the central Europe winner sooner or later will clash with the Moors hammer.

On this post you share that Redux Spain campaign is hard. Maybe is very or ultra hard, because througt 8 games, only in one Spain manage to survive and mostly Spain is the living dead already about 730 - 750 year, surrounded by Moors & alive only first global conflict arises.

Christian AI factions didn't launch crusade anytime, I thing it's requrements (below) are too high to do it before Moors enter Europe, and Christian AI's don't do it before 800 year, it seems they won't do it ever, because they get crushed by Moors hammer.
----------------------------------------------
Questions:
1. Crusade - is it play a significant role in Redux ?
What are your design goals about crusades?

It's requirements -Order House- are high - Citadel Level Castle (Castle10) - this one costing 12k florines.
In my current game (veteran level) I've lowered requirements to Castle7 - it's 'Castel' as i remember costing 8k florins.
And I enabled crusade for Poland.

2. Difficulty levels in factions descriptions - are they true? Poland - the only one - has hard difficulty, Spain - has standard.

3. Poland has no 'native' good spearman, can't build regular spearmen and higher.
Is it intentional? Anyway - I've found a workaround - Lithuanian Infantry.

4. Scenario modes: Domination vs Points - are AI factions behaves the same?
Does anyone has experience with it?



----------------------------------------------
EDIT: (regarding)
----------------------------------------------

Veteran level
I don't understand the exact reasons, but Veteran level is much more harder that Standard! And it's good.
It's not only harder battles -> wchich of course has impact on strategic map.
It's not only harder gameplay, but somehow ... different.
From my - Poland perspective - AI (HRE, Hungarians) seems to be more aggressive,
they've managed to exterminate me, starting war on 739, and it year 750 Polish King was killed in last great pathetic battle, being overrun 2,5 to 1 because of combined Germany & Hungarians attack.

Ok, regards.

[Axalon]

Hello Tomas, I see you have tried to set up this thread multiple times due to some mishaps beyond your control. Initially I have had some difficulties in grasping what you really was trying to cover here with this thread. However, after some thinking, re-reading all 3 threads, I think I finally get it. Its about RX AI-expansion patterns, trends and tendencies in Redux-campaigns examined/investigated across multiple campaigns and the effort to bring some more clarity to the subject. And, you have basically invited the rest of us to join you in that activity, right here. So... To that end I have now reconstructed this thread to be more clear and explicit on this subject. I also changed the thread-title for the very same reasons. Hopefully I have understood you right - if not - then do whatever necessary changes to further clarify what all this is really about. Until that happens, this thread and its topic is as outlined at the top of your post:1. Any problems send me a PM and we will sort it out. Anyways, below are some general answers to your initial 4 questions...


Question 1. Answer….

Crusades does not play a significant role in Redux and the AI deliberately set to never spawn any as it can not handle them properly. Typically it will do more harm then good to any AI-factions that initiates one. It is for that very reason that I removed the option for crusades for the AI. Had things been different in terms of performance, I would have allowed it. In general I would say that Jihads play a bigger role in Redux – and the AI can usually handle those much better and with a decent enough success rate as well.

Q2. Answer….

The faction challenge-levels in Redux are estimates made by me. I considered the overall location on map, number of bordering foes, threat-intensity, the wealth of the region, troops available to faction, various weaknesses etc. etc. For the most part I think most factions are properly designated. Spain’s status will increase to high as of 1005, the Norse will be lowered to standard as well. I have previously overestimated the actual challenge of the Norse. Thinking that their weaknesses would make a bigger impact on the game, as it turns out it did not, and so I’ll have changed that challenge-assessment because of the fact. As of 1004, the two most difficult factions one can play is Poland and Spain, both providing significantly tougher games then any other factions. The easiest faction around is the Saracens, followed by the Moors and Byzantines.

Q3. Answer....

This is true, they have not – if we totally disregard any use or possibility of Lithuanian Infantry (which would remedy that problem very well). Each faction typically has various weaknesses (and strengths) in their traditional unit-roster. Poland is generally lacking in infantry-options, England and Italy are essentially lacking in cavalry-options, Spain for instance has a rather obvious weakness in ranged units etc. etc. Usually, each faction has a set of troops that makes the factions distinct and different to other factions - creating a special unit-profile that sets it apart from other factions. The only way to bypass this general circumstance is to actively use provincial special units – an example relevant to your context could be Flemish Infantry for instance. However you must first control Flanders (and have the necessary tech) before you can recruit any such units. There are a number of such provincial units scattered all over the map. You must find them first however (or use a GnomeEditor and check the unit-file directly)….

Q4. Answer….

To the best of my knowledge there should not be any actual difference in AI-performance in the two modes. I have never seen or stumbled across anything within the files that suggested otherwise. There don’t seem to be any parameters that separate the two in terms of what the AI should do, it is the same regardless. I think you have your answer right there. That said, I would generally argue that domination is the (somewhat) harder alternative of the two as it expects that you do conquer enough land to win the game in time. Glorious Achievements don’t have that kind of demand/pressure on the player, only that he/she manages to gather more points then the AI – which is typically easier to do then to fight and maintain a massive empire. Generally speaking, of course….

- A

[daigaku]

Hi @Axalon,

whew, - that´ll be a hard one...so, here we go:

the Norse will be lowered to standard as well. I have previously overestimated the actual challenge of the Norse. Thinking that their weaknesses would make a bigger impact on the game, as it turns out it did not, and so I’ll have changed that challenge-assessment because of the fact.

In the "hands" of AI, it´s rather rare the Norse become a really big threat. Having played English and Russians, both an aim for the Norse under certain circumstances (if I´m not careful, they take for example Ireland or Scotland before me, or coming through Prussia into Lith country and EXPLODE), it was rather easy to keep them down. There were, of course, those rare situations when I was occupied in other regions (especially as English) that they took quite a bite of the cake. But this wasn´t a regularly upcoming situation, so my personal opinion is that in the hands of AI the Norse are still quite "hard" to achieve something. On the other side, you´re quite right about this in hands of player. If one knows how to handle their strengthes and weaknesses, they could even be counted "fairly easy". So overall truth lies somewhere in the middle. I hope it´s making sort of sense what I try to say....

England and Italy are essentially lacking in cavalry-options

Can´t talk about Italians, but the English....well, still without setting them to 40men units (after having seen how effective they are I´ll set them back to 20), those ScottishNobles and especially those EnglishKnights are quite a treat, even against those FeudalKnights the French spam out so happily. If it comes to "cavalry melee", I myself do prefer mounted swords to mounted lances - they are, in my experience, more effective in the long run. Will have to play some 1:1 battles to confirm on a not-teched-up base, but with my completely built-up units it works....

Glorious Achievements don’t have that kind of demand/pressure on the player, only that he/she manages to gather more points then the AI – which is typically easier to do then to fight and maintain a massive empire

Sorry, Axalon, can´t follow you here. Keeping the ball low on expansion myself, with GA it´s quite frustrating to see the Byz/HRE/Saracenes/Moors (or, having had that once only, the ARAGONESE), owning a third or half or even more of the map. Playing deliberately with only a few provinces it can become really hard to cope with something like that. And if the AI doesn´t behave stupid, loosing all the won advantages to rebellion (what, pity enough, happens too often), I myself don´t stand a chance to regain control or priority. I had situations when the AI used it´s wealth to really consolidate their provinces to such an extend they became so big this -i´m not sure about exact percentage- 60% penalty struck and they collapsed in internal rebellion. If that wouldn´t have happened, I never would have won the game. btw, is there any chance to take out this for only the AI? Would love to see a game lost to superiour power without sure collaps from a certain point on...

That´s it for today,

I wish you well, greetings daigaku

[Axalon]

...i´m not sure about exact percentage- 60% penalty struck and they collapsed in internal rebellion. If that wouldn´t have happened, I never would have won the game. btw, is there any chance to take out this for only the AI? Would love to see a game lost to superiour power without sure collaps from a certain point on...

Nope, that is hardcoded... Thus we are unfortunately stuck with it, as is...

- A

[Axalon]

Updating this thread to RXB1005....



And...




For additional reference: (in spoiler)

Spoiler Alert, click show to read: 

Info gathered on the old RXB1002, compiled back in 2011...

Any comments or thoughts on this stuff?

- A