How does the AI pick army composition?

[Chris1959]

Just wondering how the AI forces "pick" themselves, as sometimes stacks seem quite well balanced othertimes an odd mix of elite or dross troops.
Is it me or despite being renowned for their cavalry Gallic armies in 1.1 seem to field very little, in my current campaign at 170BC and I don't think I've seen any Gallic cavalry apart from FM bodyguards. The exception being the start up rebel forces.
I try to stick to a sort of "historical" force as recommended by some in the forums.

[fenix3279]

A historical Gallic army would look something like 15-20 units consisting of one of every type of available unit to better represent the different tribes. So if the AI seems to be fielding messy armies with random compositions then they're on the right track.

[Ezephkiel]

Im knee deep in baktrians at the moment and they're fielding quite good troops. Mainly a core of medium phalangites and a few elites with eastern crap to bulk up their formidable armies. Also they're building cataphracts, by god they're annoying.

[teh1337tim]

say that to me, allied with makedonia...who crushed greece (theyre bosphorion kindgom now)
and epiros (out of greece to pergamon, atm trying to take pergamon back after a pyrich victory against 3 armies)

they would have a greek army of like 5-8 hoplites, 2 toxaitas a couple skirmisher/calvry and a phaangita deutoria or something
but while at demetria and pella they got elite stacks just sitting the the city
in demetria alone, silver chevronned agrispidaes (8 units),2/3 vet hydapsists/peltsti makedonia(2units)
a couple pezhetairoi and a general with 6 and 8 stars with high management etc...
while there faction leader is all maxed out with 10 stars and a crazy army full of silver/gold units...
while theyre armies are fought to standstill in getai and pergamon... theyre just leaving elite stacks in their main cities doing nothing... haha no wonder theyre right under me seleukid and carthage
(effen carthage conqured all of africa and started into ptolemaio...now i conquered 1 city and a fken whole stack of sacred bands and elites are besieging me)

ya i think theyre pretty accurate... considering most of there field armies are pretty balanced...
anyone think that when those pesky 3 units stacks form into something (takes like 6 years for ai to do that lol) it actaully is pretty good?

now what do i do with that carthage beseiging me with 7 full elite stacks and 5 balnced ones...

[fenix3279]

now what do i do with that carthage beseiging me with 7 full elite stacks and 5 balnced ones...
Uh... hire alot of Gaesatae?

[satalexton]

believe it or not, bataroas are very good at gutting a phalanx up. Don't place all your units in a line, space them all out in small groups.. then promptly advance and flank each section of their line one by one. U can gut a phalanx army of equal numbers easily, taking less than 8% casaulties.

[konny]

Just wondering how the AI forces "pick" themselves, as sometimes stacks seem quite well balanced othertimes an odd mix of elite or dross troops.

Difficult to say.

The settings in descr_strat (Caesar, Mao, Genhis etc) do play a role. Stats, size and price seem to be important too (but to a lesser extend). "Class" as defined by EDU is also important, usually "light infantry" is prefered over "spearmen" or "missile". Then we have the element of random, that is "what's available" in the specific town the army was raised; but the AI is moving around stacks and merging armies, even though it often fails to do so and leaves behind "an ocean" of mini-armies.

After all, the army comopsition is an element of several decisions and so looks very much like random.

[mcantu]

i've also read that the AI will recruit based on what it is being attacked by (ie. if you are fielding cavalry, it will recruit more spearmen).

[konny]

I don't think so. In this case the armies would differ depending on who is the human player, what is - to my obeservations - not true. AS, for example, has always about the same army composition, regardless if I play the Ptolemaioi, Pontos or Armenia. When Rorarii were classified as "light infantry" the AI spammed stacks of them, since they are classified as "spearmen" they do disappear form AI SPQR armies as soon as the AI can field something better in the respective provinces.

In general there is a tendancy of what could be called "top or flop": either loads of elite units or masses of cheap peasants (sometimes mixed in a stack), omitting the "professional" class. For example when the AI is able to field Pantodapoi Phalangitai, Klerouchoi Phalangitai and Argyraspides from a single settlement it will most likely recruite either Pantodapoi Phalangitai or Argyraspides (or even both) but not the Klerouchoi, who would be most likely the backbone of a human raised army from the same settlement. When it recruites Klerouchoi Phalangitai it usualy isn't able to get Argyraspides and so goes for the best there is.

[mcantu]

I don't think so. In this case the armies would differ depending on who is the human player, what is - to my obeservations - not true. AS, for example, has always about the same army composition, regardless if I play the Ptolemaioi, Pontos or Armenia. When Rorarii were classified as "light infantry" the AI spammed stacks of them, since they are classified as "spearmen" they do disappear form AI SPQR armies as soon as the AI can field something better in the respective provinces.

In general there is a tendancy of what could be called "top or flop": either loads of elite units or masses of cheap peasants (sometimes mixed in a stack), omitting the "professional" class. For example when the AI is able to field Pantodapoi Phalangitai, Klerouchoi Phalangitai and Argyraspides from a single settlement it will most likely recruite either Pantodapoi Phalangitai or Argyraspides (or even both) but not the Klerouchoi, who would be most likely the backbone of a human raised army from the same settlement. When it recruites Klerouchoi Phalangitai it usualy isn't able to get Argyraspides and so goes for the best there is.

This is where I read it in the Scriptorium...


Originally Posted by therother
These control a set of AI production personalities, which contribute a bias towards building and training (but not retraining or repairing). This bias is fairly small compared to game-generated factors such as "the enemy is attacking me with lots of cavalry, build me some spearmen". Explaining the weighting system which drives the production AI in full is beyond the scope of this document as it would take several days to write.

So in short, the building construction personalities are these: (ranked highest to lowest - therother)

balanced - biases towards growth, taxable income, trade level bonuses (roads), walls and xp bonus buildings

religious - biases towards growth, loyalty, taxable income, farming, walls and law

trader - biases towards growth, trade level, trade base, weapon upgrades, games, races and xp bonus buildings

comfort - biases towards growth, farming, games, races, xp bonus and happiness

bureaucrat - biases towards taxable income, growth, pop health, trade, walls, improved bodyguards and law

craftsman - biases towards walls, races, taxable income, weapon upgrades, xp bonuses, mines, health and growth

sailor - biases towards sea trade, taxable income, walls, growth, trade

fortified - biases towards walls, taxable income, growth, loyalty, defenses, bodyguards and law

These biases are towards building properties, rather than buildings themselves. The game does not know what a "Blacksmith" is, for example, it only knows that it is a building which provides a weapon upgrade, and hence a Craftsman AI would be more likely to build it than another AI personality type.

These are then combined with a troop production personality, as follows:

smith - exactly level

mao - biased towards mass troops, light infantry

genghis - biased towards missile cavalry and light cavalry

stalin - biased towards heavy infantry, mass troops and artillery

napoleon - biased towards a mix of light and heavy infantry, light cavalry

henry - biased towards heavy and light cavalry, missile infantry

Caesar - biased towards heavy infantry, light cavalry, siege artillery

The same system as for the buildings applies. Troop category and class are combined at the time the unit database is loaded to give a unit production type, and the likelihood of the AI choosing to produce a given unit type which can be produced is then modified by the unit type weighting. There is also a random element in the choosing of which building or troop type to produce next, so the effect of the bias is a statistical thing. Another factor that is applied over the top which may obscure the bias is a tendency towards producing troop mixtures (according to what is already in the garrison) and a weighting according to unit strength.

The two sets of types can be freely combined; for example, although Fortified Caesar does not appear in the list of options currently used by the vanilla RTW game, it is a valid combination.

[Chris1959]

Konny, thanks for the answer I suppose in a way the AI (or rather the developer's programme) is on a hiding to nothing. Like you, and using your advice, I want to play "historical" and so face balance opponents, other players want Blitz and world conquest building super-elite armies that will crush anything, the best stats being the only issue.
I suppose a great feature, impossible I know, would be to choose at the start of a campaign what style you wished to play and the AI would re-act accordingly!!!

[konny]

This is where I read it in the Scriptorium...
This bias is fairly small compared to game-generated factors such as "the enemy is attacking me with lots of cavalry, build me some spearmen".

I realy cannot confirm this. To my obersvation the armies of certain AI factions are always the same.

Let's take the Aedui and Averni for example: absolut the same line-up, save for some nearly identical elites. The early armies are most of the times composed of spearmen (Lugoae), the reaction of both sides should be to field more swordsmen, what should result in the recruiting of more cavalry. So, in the consequence the two Gaul factions should field tons of cavalry and close to no spearmen, also as a reaction of the Romans and their full stacks of swordsmen. But the exact opposite is true: cavalry is very rare in their AI armies and the spear the most prominent weapon.