Battle AI too easy

[yonyu]

Hello there ,i already started atopic similar to this about a year ago.First of all i like to thanx the EBII team for afabelous mod. I started an haytasan campaing wich the script there with the selekuian empire for example is really cool.

But my complaint here is that even if i give the AI fanctions around me time to develop and attacked only rebels for about 120 turns,and attack only forces that invaded my land, because the battles are too easy i can take the selecuid and ptolemaic empire without too many problems. The AI in battles just eventually rout quite fast.

Maybe the solution will be to slow the routing process? stronger the morale of all units wich will give advantage to AI?
B.T.W I am playing on VH/VH.

[myarta]

Usually people say up the difficulty, but I see you're already at VH. In that case, the next option would be to check your army composition to see if your stacks are considered overpowered and ahistorical. An army with fewer "elite" or "professional" units in it will be more likely to have routers of its own and help even out the playing field. See this thread for the official guide for a number of factions:

http://www.twcenter.net/forums/showt...itions-for-2-2

[yonyu]

I dont think the army composure is the problem most of my inf is Hai Nizamartik medium spear cost effective, ihave skirmishers cappadocian spears no heavy cav only arch and skirmishers cav max two generals on the battlefield maybe one mercenery hoplite.
I played Epiros with the M patch ithink and they were quite challenging but when you defeat koinon and macedonia all the world is open to you . Maybe also trying to weakan alot big empires by loyalty defections maybe? Something similar to BI RTW Western roman empire rebels fanction.

[Jurand of Cracow]

... The AI in battles just eventually rout quite fast.
...Maybe the solution will be to slow the routing process? stronger the morale of all units wich will give advantage to AI?
B.T.W I am playing on VH/VH.

I wonder if it wouldn't be desirable to give a +morale advantage only for the AI for the higher difficulties, if this is a problem.
Concerning the army composition I always think that it's the designers task to pick such values of recruitment refill-rates, upkeep prices, availabilty (ie buildings and their levels) etc. so that the player produces at the end of the day a right army composition. If he doesn't have access to elites units, if he doesn't have enough money, if here're no recruits - he'd recruit worse troops. If the upkeep is too high, he'd disband.
Actually there're many tools in the M2TW engine to make it happen.

[QuintusSertorius]

I dont think the army composure is the problem most of my inf is Hai Nizamartik medium spear cost effective, ihave skirmishers cappadocian spears no heavy cav only arch and skirmishers cav max two generals on the battlefield maybe one mercenery hoplite.
I played Epiros with the M patch ithink and they were quite challenging but when you defeat koinon and macedonia all the world is open to you . Maybe also trying to weakan alot big empires by loyalty defections maybe? Something similar to BI RTW Western roman empire rebels fanction.

How strong are your garrisons? How many armies do you have? Are they full stacks? How much cavalry is in that army?

Note if you rush the AI, no matter how long you give them to build up first, they will not provide as effective a resistance as if you go slower. Meaning take longer between conquests. Taking 2-3 provinces in as many turns, with an army for each is something the AI cannot respond to.

I wonder if it wouldn't be desirable to give a +morale advantage only for the AI for the higher difficulties, if this is a problem.
Concerning the army composition I always think that it's the designers task to pick such values of recruitment refill-rates, upkeep prices, availabilty (ie buildings and their levels) etc. so that the player produces at the end of the day a right army composition. If he doesn't have access to elites units, if he doesn't have enough money, if here're no recruits - he'd recruit worse troops. If the upkeep is too high, he'd disband.
Actually there're many tools in the M2TW engine to make it happen.

There's only one EDU, so unless we have a different one for each difficulty setting (no thanks!), there's no way to vary the morale of the AI. And higher difficulty settings already give the AI a bonus to morale and stamina.

Recruitment is already balance to provide lots of levies and fewer professionals and elites. I invariably find the people complaining about how easy their economy is, are the same ones who exploit public order in their core regions and leave skeleton garrisons behind. No settlement should have less than four units in its garrison (not including the governor), bigger places should have more. It isn't unreasonable to have a half-stack parked in your capital, especially if it's a city or larger.

[Jurand of Cracow]

And higher difficulty settings already give the AI a bonus to morale and stamina.

I admit having no knowledge of possibility to modify the difficulty levels so I assume that's hardcoded and not possible to be moded.

There's only one EDU, so unless we have a different one for each difficulty setting (no thanks!), there's no way to vary the morale of the AI.

If most of the AI armies are lead by a general (I don't know if it's the case in the EBII, I haven't played so much to tell) then there's one way. You may give the AI general a boost by adding "TroopMorale" benefits as a hidden trait. In the SSHIP we use some AI boosts on different issues. For instance, we know that the AI is not able to educate his generals (it doesn't know it exists, does it) so at the Coming of Age there're initial additional points for education. You can do a similar thing with TroopMorale.

Recruitment is already balance to provide lots of levies and fewer professionals and elites.

This should result in the balanced armies, so there's no need for the calls to the players to artificially restrain themselves to create balanced-composed armies.

I invariably find the people complaining about how easy their economy is, are the same ones who exploit public order in their core regions and leave skeleton garrisons behind. No settlement should have less than four units in its garrison (not including the governor), bigger places should have more. It isn't unreasonable to have a half-stack parked in your capital, especially if it's a city or larger.

I agree on the result, but I believe it should not be the task of the player to show such a behaviour by his nature, but it's the task if the game (mod) to enforce such behaviour on the player. He should simply be better off leaving those units behind, not skeleton garrisons - for many reasons, be it defence from the AI incursions, potential rebell attacks, or keeping the order in the settlements. For instance, the high unrest levels may do the trick - I'm forced to have many troops in the settlements in the SSHIP game (you may see it in this table: number of units and troops, and also the garrison costs). The free upkeep is also beneficial providing the player with positive incentive to keep the units at home.

[QuintusSertorius]

I admit having no knowledge of possibility to modify the difficulty levels so I assume that's hardcoded and not possible to be moded.

Hardcoded and not even confirmed because CA have never said what the effects of battle difficulty are in M2TW.

If most of the AI armies are lead by a general (I don't know if it's the case in the EBII, I haven't played so much to tell) then there's one way. You may give the AI general a boost by adding "TroopMorale" benefits as a hidden trait. In the SSHIP we use some AI boosts on different issues. For instance, we know that the AI is not able to educate his generals (it doesn't know it exists, does it) so at the Coming of Age there're initial additional points for education. You can do a similar thing with TroopMorale.

If you give the AI time to build up its forces, it will be led by an FM. Who get bonuses to morale (and other things) through the AIGeneral trait. If you rush the AI, chances are you'll face fragmented, Captain-led stacks which are easy to defeat.

This should result in the balanced armies, so there's no need for the calls to the players to artificially restrain themselves to create balanced-composed armies.

It's not enough by itself. If you have several settlements and the patience to build up your armies slowly, you can avoid those limitations. Not to mention that recruitment is done on a per-settlement basis, if you have five settlements with the same recruitment (rare in EBII but possible), there's nothing to stop you only recruiting all the heavy cavalry from those five settlements in the same turn.

I agree on the result, but I believe it should not be the task of the player to show such a behaviour by his nature, but it's the task if the game (mod) to enforce such behaviour on the player. He should simply be better off leaving those units behind, not skeleton garrisons - for many reasons, be it defence from the AI incursions, potential rebell attacks, or keeping the order in the settlements. For instance, the high unrest levels may do the trick - I'm forced to have many troops in the settlements in the SSHIP game (you may see it in this table: number of units and troops, and also the garrison costs). The free upkeep is also beneficial providing the player with positive incentive to keep the units at home.

That's an impossible task. High levels of unrest have unintended consequences and forcing the use of larger garrisons will be crippling to small factions early on. Having too high an effect of garrisons negates the balance of unrest from other sources, and CA didn't give us a proper set of controls in the descr_settlement_mechanics over garrisons.

[Jurand of Cracow]

If you give the AI time to build up its forces, it will be led by an FM. Who get bonuses to morale (and other things) through the AIGeneral trait. If you rush the AI, chances are you'll face fragmented, Captain-led stacks which are easy to defeat.

This is why there're initial rebel stacks - they buy time for the AI to buid-up it's forces (but also making it harder also for the faction lead by the AI, right?).
But yes, indeed this is the question: how to ensure the AI is lead by the generals. Do you support the AI somehow by the script as far as the number of the generals is concerned? (I've got this idea for the SSHIP at one point, but wasn't much welcome. Then a fellow moder made it in his submod, but in a way I don't like - making it available also for the player (I'd restrict it only to the AI)

It's not enough by itself. If you have several settlements and the patience to build up your armies slowly, you can avoid those limitations. Not to mention that recruitment is done on a per-settlement basis, if you have five settlements with the same recruitment (rare in EBII but possible), there's nothing to stop you only recruiting all the heavy cavalry from those five settlements in the same turn.

Yeah, but this is still a limit. Besides, you may remember our discussion on the price/upkeep ratio. In the EBII there's a strict 3/16 approach (and you've explained it to me why, thanks again for that!). In the SSHIP there's a different approach: the elites have such a high upkeep prices (plus 1/1 ration) that you disband them after a campaign. When a need comes, you can recruit them but still your're limited by the re-fill rates. The outcome is that the armies are more-or-less balanced.
But even without this variation, you may come up with other solutions. Byg in his BGR had the Professiona Training Staff system which enabled recruitment of the elite units only in 3 places at one time. Coupled with high prices for the elite units it indeed ensured the right army composition (besided limiting the snow-ball effect).

That's an impossible task. High levels of unrest have unintended consequences and forcing the use of larger garrisons will be crippling to small factions early on. Having too high an effect of garrisons negates the balance of unrest from other sources, and CA didn't give us a proper set of controls in the descr_settlement_mechanics over garrisons.

Ok, if you say so. My initial thought would be that we could come up with something for the player playing smaller faction (like a few good governors ensuring order without many units present and living at least for the first 20 years). However, there might be unintended consequences.

[Jimalim]

Just some feedback from a long time player:

Although as QS says there has never been an official note on the difference between battle difficulties, I can personally attest that morale/stamina is higher on higher difficulties. Generally speaking it seems to be that there is more morale and stamina on higher difficulties. Player research estimates that:

Hard: + 4 morale, +4 stamina
VH: +7 morale, +7 stamina.

If VH battle difficulty is too easy I recommend limiting your strategic advantage of the digital format. Play with a restricted camera, for example. Force yourself to assist allies even when it isn't necessarily beneficial.

I always say build your army the way you want too, even at the expense of history, if that is more fun for you. But certainly if challenge is what you are after, make the attempt to build historical armies.

P.S. A horse archer faction like Hayastan is always going to be capable of easily painting the map. Horse Archers are overpowered in all total war mods on these older games. Play using less of these and less foot archers and you will have a more challenging campaign.

[Jurand of Cracow]

Hard: + 4 morale, +4 stamina VH: +7 morale, +7 stamina.

Yep, I've seen it before. However, I haven't read much about stamina. How the points translate into it? What does it mean to lose whole stamina (when a unit is "exahausted")? In Attila you actually see it (iirc: max 35% of strenght)

If you play Medieval 2 and its mods long enough, you will get to a point when you won't lose battles anymore. It's the case for me. Doesn't mean it's not interesting anymore. In fact, the less time and effort I spend commanding, the more time I have to enjoy the show.

Depends on the number of troops you get to a battle. The player should take as little as possible: "just to make it" - and lose from time to time. There should be incentives made by the mod designers for such a behaviour - eg. much higher probabilities for generals to get good traits if they win having less / inferior troops. (I didn't like Byg's solution from the BGR: if you fought with stronger forces, you'd get nasty traits).

A horse archer faction like Hayastan is always going to be capable of easily painting the map. Horse Archers are overpowered in all total war mods on these older games. Play using less of these and less foot archers and you will have a more challenging campaign.

I don't think there's such a rule. It depends on the balance found in the game by the designers (during balancing).

[Rad]

Depends on the number of troops you get to a battle. The player should take as little as possible: "just to make it" - and lose from time to time. There should be incentives made by the mod designers for such a behaviour - eg. much higher probabilities for generals to get good traits if they win having less / inferior troops. (I didn't like Byg's solution from the BGR: if you fought with stronger forces, you'd get nasty traits).

Nah, that would be... wrong. The exact opposite of cheating, but somehow it would feel just as bad. I fight battles to win them, and I will not willingly get drawn into one I cannot win. I usually have few field armies, but they're made to win - within reason, ofc. I don't spam elites.

Sometimes, the enemy will come out of nowhere, or suddenly declare war on me and siege a lightly guarded city or something. Then I'll have a tough fight on me. I'm fine with that.

If I want a challenge, I declare war on a faction that has at least a dozen field armies. For example, the Grey Death or KH if it remains in isolation for too long.

[Rad]

If you play Medieval 2 and its mods long enough, you will get to a point when you won't lose battles anymore. It's the case for me. Doesn't mean it's not interesting anymore. In fact, the less time and effort I spend commanding, the more time I have to enjoy the show.

[QuintusSertorius]

Hayastan aren't a horse archer faction, they simply have a native horse archer unit and lots of cavalry available to them. Much of it, being eastern, of good quality.

[Jimalim]

By arcade style i mean spam HA, not literally choosing arcade no morale battles.

[Jimalim]

I just meant that any faction that has early and easy access to HA can easily wreck the AI, IF u arcade style play. Im not trying to classify anything. Should've been more clear.

[MagusCaligula]

Well the way to make the campaign more challenging is play it as a Paradox game and just auto resolve everything And even more challenging remove Mercenaries

[Hellenikon]

Exhaustion reduces a unit's defense/attack/morale, so the more tired the more useless the unit is.

[Pooploop]

Before anything else, let me say that you EB2 devs have created a really awesome mod that is not only beautiful but also historically accurate. However, there are some CA hard-coded limits and hindrances imposed by the relative "oldness" of the game that makes improving the mod harder to do; stuff like Campaign and Battle AI (which still is unsatisfying to a good portion of Total War players even in CA's more recent titles). Knowing this of AI development, that even the company that makes the original game finds it tedious (either it's really beyond their capabilities or they just find it hard to focus on with investors breathing down their backs), I'm still going to suggest (fully aware that you guys do this in your free time, and I'm truly grateful) to you EB2 devs, to "please make the CAI and BAI more intelligent". The fact that I can't come up with specifics (e.g. like code them like this to improve aggressiveness, or add this line to randomize their time of attacks) shows my baleful ignorance of the technical details (which are of the highest importance) but what I do know is that the mindset that players need, NEED, to LIMIT their play-style to help enemy AI (enemy!!) to fully enjoy the game is wrong, wrong, wrong. The fact that you need to wait 120 turns being peaceful is like fighting a chess match against a computer, but conceding 10 turns to the AI, or giving up your rooks and bishops, just to have a "challenging" match. When one side needs to put on kiddy gloves to let the other side have a chance, it's disparaging. The goal should be (goal, not current state of the game because, as I stress, I know that actually getting there is the hard part) that in some future EB2 release, there will be a need for the player to use all dirty player tricks to help on Very Hard Campaign, and that winning one or surviving for a set amount of turns is grueling but ultimately fulfilling.

To give an example of a MTW mod that gives this experience, I'd like to point out Rise of the Three Kingdoms. There is no point in the game (until the very latter end when you've conquered most of China) in which it's easy. The enemy AI from whatever faction around you is out to take a piece of your land, and fields full stacks and attacks multiple points along the entire front. This fielding and aggressive employment of troops, or close to it, is continuous and doesn't stop whether the AI faction is on it's last breath or is a rival superpower to yours. This means that when the player is attacking a weak nation, he must commit large armies under competent generals to prevent them from being wiped out in ambush, or to make a dent in attacking garrisoned cities, or simply to maintain a presence in every inch of newly conquered ground (as if he turns his back and leaves a token garrison, the AI immediately snatches the settlement back with another attack). When attacking a powerful nation, the player must be creative, and deliberately have chokepoints to defend, and a support and reinforcement system of stacks. The former is needed to stop the powerful nation's Xerxes-esque attack (there was a game, playing as Cao Cao, when finally, FINALLY, having been able to secure the Central Plains and deliberating whether to attack the South-East under the Sun or the North under Yuan, that I was attacked from the West by the Ma of Xiliang. After defending the passes to LuoYang multiple times from a never ending stream of enemies, I kinda used view all cheat to ascertain how many troops they were fielding against me, as this hopeless feeling of futile victory had never plagued me before in playing games like Stainless Steel, or EB2 - early releases. Haha, lo and behold, the whole NorthWest of China under one banner and fielding nigh 15+ close-to-full stacks all streaming towards the poor battered capital or LOOPING around the LouYang Corridor to attack the main chunk of the Central Plains, with myself only having 5 stacks, spread out, and some useless as they were recuperating in my most eastern provinces. Needless to say, I quit after that, and laughed, as that was the most epic MTW2 game I've ever played.) and the latter is needed to have even a bit of a hold in their empires.

Now I'm not saying that EB2 must be exactly like ROTK. But, there are some pointers to be taken from the game that can be applied to make EB2 harder, without the need to give enemy AI 120 turns of peace. In this, EB2 has an edge in it's new and improved BAI, as ROTK only has Germanicus's AI. One, complete limitation of elite units. In ROTK, elite tier units can only be recruited once. Maybe units like hetairoi or Parthian cataphracts can be hard-limited as well by the number of unique structures present in the whole faction? Like one "King's Palace" means fielding one elite unit. Build one more in a different settlement, you can support two. This completely eliminates spamming, or temptations to spam, and forces players to cope with crappier units that rout as well. Two, more AI troops, and continuous fielding of troops, of crappy tiers, all around. I don't really buy the argument that ye olde ancient nations could only field one to two armies and after defeating them, they'll fall right over. Can one really apply that to situations like the Rome that Hannibal had to face, the Jews of their great revolts, the Afghans against Alexander, the Gauls against Ceasar, the various Indian Kingdoms against each other, the nomad tribes against everybody. You study their victories and defeats, and the number of troops they could field and their overall performance and note that they were NOT like Alexander's Persia that capitulated only after three major battles (and a vast number of sieges and skirmishes). It took tens of battles over a long period of time and then some, because as long as their spirit was unbroken and their racial hatred against their conquerors remained (everybody was racist back then to a greater or lesser degree) then expect stacks being thrown at you. How many troops did Rome require to keep watch over Britain and over the Rhine frontier? Nearly 50,000+ and more on EACH front? Against foes that were partially subjugated (on and off), disunited, and of vastly inferior logistics and urbanization? Yet Rome REQUIRED that many just to keep the peace and prevent or DEAL with invasion or rebellion? Then you tell me that in EB2, the AI, to be historically accurate, can only field one stack and after that it must be easy for the player to conquer their cities? The resistance the Persians put up against Alexander was not the norm.

Look at the Greeks against the Persian invasions. Their coalition didn't include Crete, Thessaly, Thebes, Argos, Arcadia, Rhodes, the cities of the Aegean, the northern Greek tribes and settlements, and Magna Graecia, but they could summon up to 30,000+ hoplites in battle, and their light troops numbered at least two-to-one to the hoplites, around 60,000 (at least 4 helots per Spartiate, and hoplites of other polises had 2 servants/squires/light troops, so my numbers are conservative). You're not even counting the other rowers of the Athenian fleet (those who didn't rotate into the land army), which had 140+ triremes and 150+ men per ship. Yet, the original Greek Cities were considered inferior in population to their Ionian brethren, and those Asia Minor Greeks were vastly inferior in numbers to the original and ancient Asiatic and Near East nations. Yet, in EB2, an average faction is only supposed to be able to summon one to two stacks and then keel over? Come on! I'm not saying the troops and stacks the AI is supposed to make are filled with medium tier troops. Even stacks 80%+ composed of peasant spearmen and slingers are better as long as they EXIST and can garrison cities and support the main line troops. I want Arche Seleukia to be able to field 20+ stacks even if 3/4th's of them are filled with crap troops. At least it''ll finally be a challenge to the player, and the player can't do stupid things like garrison a city with one unit and expect he'll get away with that.

The bottom line is, is that, whatever system EB2 devs will implement, whatever techniques, or scripts, or bonuses, or limitations, or whatever, it has to result in a game which makes the player play "conservatively", because it is the smart thing to do and NOT because he has to coddle the AI.

P.S. Going back on the unit limitation thing, one can conceive of a system in which it applies to ALL units except those of peasant-tier. Every grade, from elite to medium, will be limited by the number of "buildings" in total present in the entire faction. You can, for example, only train three units of levy phalangites (not the pro's) for every polis (example) and can only train another three more from the same settlement only if those original three were all slain. This means the player (and the ai) will deal with armies in which even barely competent troops are a valuable resource and the threat of rout is ever present as at least close to half of the troops they use (in proportion, across all stacks) are of levy spearmen or peasant slingers or javelineers. Mercenaries will now be also used extensively as the idea of ready-to-hire decently armored and armed troops are a wonderful commodity, but one can put a general minus to all mercenary morale to represent lack of drive in them as a profession and a higher priority to save their own skins.

P.P.S. There was a suggestion long ago of lessening all cavalry morale to promote the actions of routing and reforming continuously. I have no idea of how to apply it to the rest of the types of troops in EB2, but can the "reforming" part be increased? This will prolong battles and lessen the complaint of winning "easily" as there is an increased chance the enemy troops will reform and attack flanks or now exposed backs, thus preventing players from over-committing themselves in pursuit, or under-committing themselves in any engagement. Basically battles will be more complex and fun.

[myarta]

The goal should be (goal, not current state of the game because, as I stress, I know that actually getting there is the hard part) that in some future EB2 release, there will be a need for the player to use all dirty player tricks to help on Very Hard Campaign, and that winning one or surviving for a set amount of turns is grueling but ultimately fulfilling.

I'll start similarly and say that first, I found the feedback interesting and thoughtful. But I would say that if by dirty player tricks you mean things like blitzing with multiple armies to besiege half of a smaller faction's cities at once or having an all-elites army or using all 6 family members for infinitely free-replenishing cavalry as long as they didn't die, that I don't think such things would be a goal of the mod, in that having to resort to them takes us out of historicity, even if it's optional in being only at the highest setting.

I love the work that Gigantus and his team did for ROTK and it's a beautiful piece, but for me I find that I don't tend to enjoy playing it as much for exactly the same reasons that you like it: it feels like I have to fight multiple battles every turn and I've got a dozen generals with their own small detachments to manage, with fronts in all directions, and you build recruitment buildings fast and just churn out regular or heavy infantry all the time to meet the constant threats, it feels like. For me, it's more stressful and tedious than challenging, but I wish I would enjoy it more since I'd love to explore their traits system and Liu Bei missions more.

I like the slower pace here, though I do agree that the challenge of that is you reach a certain plateau of comfort, where your borders are reasonably secure, your army is in good shape, and your income is buoyant, and it can feel like with no pressure or real goals besides a reform (which not all factions can work towards), it is a slog to get to 500 while taking modest nibbles to avoid becoming too strong. So I suppose there are tradeoffs, and if people do want some sort of "survival" mode, I think it's best implemented by finding the "nigh-impossible" or "very challenging" factions and trying them on VH/VH. Some factions just objectively have an easier position, and that's reflected in the difficulty estimate in the campaign start screen.

[QuintusSertorius]

Pooploop - I'm guessing you've never played EB1? Because what you're describing is an attempt to recreate the way that plays. Which I'm afraid, is boring. Fighting the same 2-3 full stack battles, turn after turn, forever (because factions never accept peace, nor does the military AI ever stop building for attacks once its decided you need to go) is not "epic". It's a chore. It's exactly why I never play games of longer than about 100 turns in EB1, because the whole thing is simply repetitive and un-fun. There's no sense of achievement from winning any battle, because you'll have to do it all again.

There is already a limit on most elites - they come from the top-tier government, and for many factions you can only have one or a small number of those. You're missing the point, though, you don't need elites to make the game easy. Just build full stacks comprised of nothing but medium cavalry and medium infantry. Sure every battle will be the same boring slog, where your tactic is simply to close so your too-numerous cavalry can charge to the rear of everything your melee infantry engage, but it's effective. Especially when the AI tends to recruit variety (and you've no idea how much trouble that's taken to achieve, to stop them either recruiting nothing but crap, or only their highest tier units).

You can't limit recruitment of units on a "per building" basis, that's not possible without a huge and onerous script that would impair performance for only a small gain.

Most nations could only field one "royal army", then if they were defeated that was them done for a few years. They most certainly couldn't maintain the sort of continuous drain on manpower (Rome aside) that is implied by the way EB1 allowed the Seleukids to raise 2-3 full stacks a turn, every turn, forever. The AI is slow to respond, if you hit them in multiple places with multiple stacks, they cannot deal with that. There's nothing we can do to make the AI "cleverer" in that regard, it isn't a human opponent.

Full stacks of levy dross are make-work, they're not hard for the player to beat which turns them into nothing more than a speedbump or a tedious grind. EBII used to have a special "AI bonus" recruitment section, whereby Hellenistic AI factions got extra recruitment that the player did not; it was one of the first things I removed when I took over recruitment. The AI gets a raft of financial support, and the player has plenty of recruitment already available, so I see no reason for the AI to get a host of additional units when it just leads to spam.

P.S. Going back on the unit limitation thing, one can conceive of a system in which it applies to ALL units except those of peasant-tier. Every grade, from elite to medium, will be limited by the number of "buildings" in total present in the entire faction. You can, for example, only train three units of levy phalangites (not the pro's) for every polis (example) and can only train another three more from the same settlement only if those original three were all slain. This means the player (and the ai) will deal with armies in which even barely competent troops are a valuable resource and the threat of rout is ever present as at least close to half of the troops they use (in proportion, across all stacks) are of levy spearmen or peasant slingers or javelineers. Mercenaries will now be also used extensively as the idea of ready-to-hire decently armored and armed troops are a wonderful commodity, but one can put a general minus to all mercenary morale to represent lack of drive in them as a profession and a higher priority to save their own skins.

You might conceive of it, but I'm afraid it's impossible in the system CA has given us. There is no coded flag which notes a specific unit no longer exists. You can track recruitment and disbanding. Anything that happens between those two points (ie units being killed/routed) is not recorded and thus can't be tested.

Again, armies of levies are pointless, they're not a challenge, just battles for the sake of having battles. Which you'll win easily with a moderately balanced army.

P.P.S. There was a suggestion long ago of lessening all cavalry morale to promote the actions of routing and reforming continuously. I have no idea of how to apply it to the rest of the types of troops in EB2, but can the "reforming" part be increased? This will prolong battles and lessen the complaint of winning "easily" as there is an increased chance the enemy troops will reform and attack flanks or now exposed backs, thus preventing players from over-committing themselves in pursuit, or under-committing themselves in any engagement. Basically battles will be more complex and fun.

We have no control over any part of this, except the value we choose for the morale. The point with lowering morale is that if units flee earlier, they are bigger and less tired, which impacts the hardcoded decision over whether to rally or run for good.

Again, most of the time when we probe what army compositions people who complain it's "too easy" are using, they not even remotely balanced. They're overpowered full stacks with little or no skirmishers or light cavalry.

P.S. I hope factions in EB2 are proof from destruction by assassination of all their family members. Also, garrison scripts might be an answer to make taking settlements more difficult, but preventing the AI in using anything for attack. It could be considered historically accurate, as long as all garrison troops are of levy level, as during sieges, the inhabitants of the city WOULD fight alongside the garrison to prevent the highly probable and eventual sacking and slaughter.

Assassination of FMs is something possible in M2TW, you can't make factions "proof" from it. However, it's pretty unlikely than an assassin can kill the FL and FH.

We already have a limited implementation of a garrison script for the starting factional capitals and a handful of Rebel "capitals". Which spawns up to 10 mostly levy-grade units to cover the inability of the AI to garrison appropriately. We're not going to roll that out any further, because it simply makes for dull gameplay when every single siege is a slog. We'd much prefer factions to be defeated by battles in the field, which are thus decisive, than Siege: Total War.