The QUEST against ALL-ELITE armies

[Jambo]

After a good few campaigns there's a common theme that unfolds as the human plays.

1. The start involves struggling to raise effective armies and balancing one's tight finances between building troops and infrastructure. Mounting any kind of offensive campaign requires a good deal of thought and having more than one neighbouring enemy can be fatal.

2. THe mid game is somewhere between 1 above and 3 below.

3. The late game involves more money than one knows what to do with, the ability to field several ALL-elite armies, and all the while affording to build any infrastructure that's considered worthwhile at this stage. In fact, having such staggering earnings means that the subtle differences in unit price and upkeep, which seemed to make a noticeable difference at the start, no longer have any relevance (e.g. if you have 50K florins, a knight unit costing 900 is no different to a town militia costing 300). Typically, at this stage one is also at war on several fronts and able to cope admirably.

Somewhere the good balance that existed at the start, which made the human have to think strategically about what was more critical to build and who to attack, gets lost as the game progresses. I've certainly experienced the end game enough to know there's little strategy when choosing what units to build; I just triple click on all the best units I can from each castle I own.

So how can this be combatted without drastically altering the early game too much? Well here are a few thoughts for discussion:

a) Proportionally increasing the cost of units, such that the mid range get a modest hike and the elites get a larger hike?

b) Proportionally increasing the upkeep of units in the same way?

c) Maybe something different altogether like increasing the effect of corruption as the empire grows.

Either way, clearly all these will affect the AI as well as the human, but which one would you consider the most appropriate and why? Or, is there another option that's more appropriate?

Of them all, (c) probably affects the AI least since the AI is far less likely to forge a large empire. However, I'm also interested in the other two. (a) would make building units more difficult and taking care of them more important, but making them too pricy might inadvertantly discourage the AI from building them at all. (b) would punish someone for having too many sitting around doing nothing, but might make someone use them more suicidally, since the high upkeep would encourage them to be used and destroyed rather than kept and maintained. Plus if the cost is still minimal, then building a new unit isn't a problem.

Regards

[Lusted]

I went for C in my LTC mod, as well as reducing the recruitment pools and replensihment rate for elite units.

[econ21]

I hoped the unit caps would help deal with this problem, but they are not really binding enough to do so. If they were tighter (like the one merchant per settlement cap), they could help.

However, I am not sure the all-elite armies issue is a big problem for SP as the player can just use some self-discipline and the AI does not go over board on elites in my experience (and when it does, it might be terrific fun to fight - some of the early AI Danish armies I have fought have been seriously hardcore).

I normally hate "self-discipline" as a solution to a game problem. This is because I don't have any self-discipline. For example, in MTW, it would really hurt me not to spam a sea-wide trade network, even though the AI never does and the lucrative money breaks the game.

However, in the case of army composition, it is easier to apply self-discipline because you can look at historical army compositions and get an external standard that is not arbitrary.

For example, on the RTR forums, there is a wonderful thread giving illustrative historical armies for each faction. Playing with a genuine pre-Marian army is so much more fun that just spamming principes and funditores.

It's a little harder in M2TW, as armies were much less uniform than a Roman legion and there really is no one historical army composition for each faction. But some simple rules of thumb are not hard to conceive. For example, only take one Scots Guard if French etc.

To be honest, I have not found the "all-elites" problem a big issue in my SP England game, as England does not really seem to have notable elites. Knights seems much of a muchness, regardless of era. Armoured swordsmen seem the best infantry, but are not that "elite". There are various longbowmen, but they all do the job. The spears are weak, but you may as well take a few as meatshields against an AI cav charge. You end up with an army that is not a million miles away from a historical English army. Historically, the Hundred Years War English army was fairly "elite" or at least semi-professional - they did not bother to ship loads of militia to France. The French tended to load up more on feudal foot, as it was easily available. I can see the "all-elites" being a bigger problem for factions with truly elite and rare units like Varangians, JHI, etc.

I'd be interested to hear people's suggestions on what would constitute historical army compositions for M2TW factions. It's one thing we debated a little for the HRE PBM we have started in the Throne Room. This is what we came up with for house rules to try to make our HRE armies look historical:

Spoiler Alert, click show to read: 

The following rules apply for field armies of 15+ or more units.

Generals - max 2 units
Knights - cavalry or foot, max 8 units inc. generals
[The class of knights is therefore: Dismtd Feudal knights; Dismtd Imperial knights; Dismtd Gothic knights; Mailed knights; Feudal knights; Imperial knights; Teutonic knights; General’s bodyguard; Gothic knights; plus any mercenary knights included those great dismounted knights you get in the Holy Land.)

Total cavalry - maximum 8 units, inc mounted knights and generals
[Non-knightly cavalry includes: Mounted crossbowmen ; Reiters; Merchant cavalry; Mounted sergeants]

Artillery - maximum 2 units (5 in a siege - if caught in a field battle immediately withdraw excess of over two)
Foot missiles - maximum 6 units including artillery
[Foot missiles include: Peasant archers; Peasant crossbowmen; Crossbow militia; Pavisse crossbowmen; Arquebusiers; Handgunners ]

Total elite heavy infantry - max 6
[Elite infantry comprises Zweihander; Forlorn Hope; Landsknechts; dismounted knights and equivalent mercs - e.g. Galllowglass?]

Other spears & feudal foot - unlimited
[This includes: Peasants; Town militia; Halberd militia; Spear militia; Sergeant spearmen; Armoured spearmen; Crusader sergeants; Pike militia]

For armies of size 7-14, the above limits are halved.

No more than half an army can be mercenary. Crusader mercenaries (crusader sergeants, crusader knights, pilgrims, fanatics) can count as natives.

[grinningman]

The problem is that you earn too much cash as your empire expands. Civ4 tried to solve this problem by charging a city maintenance cost that increased as you built more cities. Would it be possible to introduce a similar upkeep cost per province that increases with total empire size, representing the cost of a bureaucracy?

[Lusted]

The problem is that you earn too much cash as your empire expands. Civ4 tried to solve this problem by charging a city maintenance cost that increased as you built more cities. Would it be possible to introduce a similar upkeep cost per province that increases with total empire size, representing the cost of a bureaucracy?

Possibly through scripting. In my LTC mod i just made corrpution and other settlement factors more important so it ws not as easy to make lots of cash.

[grinningman]

Yeah, that's one way to balance it. But since the cost per province doesn't increase as you capture more provinces, you'll always eventually get to a point where you have too much money as your empire grows.

Scaling the individual province upkeep cost with the number of provinces in your empire means that capturing more provinces isn't always a good thing, like it is now. It sounds like there will be lots of interesting possibilities with scripting...

[Jambo]

Lusted, how much did you modify the SIF_corruption pip modifier by? And is it just affected by distance to capital or is it also related to number of settlements?

Yeah, I also agree that the vanilla unit caps and replenishment rates have absolutely no impact on what units you want to build. The fact that so many are available in 3's straight from a castle upgrade almost negates the need to build the corresponding barracks which will enable another 3. I'm considering a major overhaul of the recruitment system with regards to castles to encourage the building of the stables and barracks line of buildings more. I'm sure the AI wastes plenty of money on these buildings when in reality, in vanilla, they're not even necessary half the time.

Lusted, when you were modifying the castle recruitment pools were there any units specific to the castle upgrades that would be lost if all unit entries were simply deleted? Or are all units at least replicated somewhere in the barracks/stables upgrade lines?

[Lusted]

Jambo i removed all units from the castles/alls buildings and moved them to barrakcs/stables so the ai has to build them.

I just increased corruption, and the distance to capital penalty, as well as religious unrest and tweaked some other things.

[hrvojej]

a) Proportionally increasing the cost of units, such that the mid range get a modest hike and the elites get a larger hike?

b) Proportionally increasing the upkeep of units in the same way?

Unfortunately this solution, as well as lowering the ability to recruit elites and/or cavalry, hurts the AI recruitment more than it's going to hurt me as a player. I'll wait to get the units I want in my army, but the AI won't, it will recruit whatever it can and waste money on garbage. So I'll just end up fighting lots of low-class armies, which is no fun. I therefore increased the availability of good troops in higher tier buildings, and eliminated some of the garbage units altogether from those buildings too, while I balance my own armies through self-restraint (that is, based on my ideas of combined arms). For now it seems to result in a much more entertaining game.

[Booga]

Personally I don't find money or all-elite armies to be too much of an issue, but I tend to rapidly expand rather than turtle. The burden of assimilating a large amount of new territory is a financial drain that also ties down a lot of troops, and seems very historically accurate too.

Right now I'm playing as the Holy Roman Empire and I've wiped most other factions off the map - I control all of Western and Central Europe, to the Balkans and Lithuania, and Africa to the edge of Egypt. Most turns I have enough money to start building in all my cities, but not every turn, and that doesn't include what I need for army recruitment. I still have the Holy Land and the steppe to crack, which will be difficult because I'll need an army of priests to assist me in addition to my own armies. Speaking of armies, they're far from just Reiters and Gothic Knights - they vary widely based on what area of the map they're deployed in. I try to establish troop production regions, with Italy a base for militia types, further north for heavy cavalry and infantry, and my heartland in Germany for artillery, but I move so quickly that it's impractical to do anything but build a few units and ship them off as reinforcements. So just through circumstance, I have lots of low-tier units supported by a few elites, and I fight a lot of pitched battles - if I waited until I had overwhelming supremacy in every single fight I'd probably still be mucking about in France. BTW, it's only about 1330 in my game, so I have not spent any time turtling. So for me, the key seems to be to be aggressive, and the game ends up finding an equilibrium.

[Jambo]

Unfortunately this solution, as well as lowering the ability to recruit elites and/or cavalry, hurts the AI recruitment more than it's going to hurt me as a player. I'll wait to get the units I want in my army, but the AI won't, it will recruit whatever it can and waste money on garbage. So I'll just end up fighting lots of low-class armies, which is no fun. I therefore increased the availability of good troops in higher tier buildings, and eliminated some of the garbage units altogether from those buildings too, while I balance my own armies through self-restraint (that is, based on my ideas of combined arms). For now it seems to result in a much more entertaining game.

In terms of (a) then yes I agree. Increasing the cost of units can and probably will deter the AI from building them, period. However, I'm not sure the AI factors in upkeep when it builds units. If the AI has less elite units which it invariably does, then (b) would hurt the player far more than the AI.

I've never been one for self-restraint unfortunately. I get bored extremely quickly if I find myself beginning to compromise my strategies to accommodate a lacklustre AI. I'd much rather pump the AI up and really have to fight tooth and nail to win.

The idea of reducing the number of low tier units in higher tier buildings interests me though. Quite a tricky and time-consuming one to mod though given the size and complexity of the EDB file.

[hrvojej]

In terms of (a) then yes I agree. Increasing the cost of units can and probably will deter the AI from building them, period. However, I'm not sure the AI factors in upkeep when it builds units. If the AI has less elite units which it invariably does, then (b) would hurt the player far more than the AI.

I've never been one for self-restraint unfortunately. I get bored extremely quickly if I find myself beginning to compromise my strategies to accommodate a lacklustre AI. I'd much rather pump the AI up and really have to fight tooth and nail to win.

The idea of reducing the number of low tier units in higher tier buildings interests me though. Quite a tricky and time-consuming one to mod though given the size and complexity of the EDB file.

Upkeep will eat up AI finances as well, meaning it will have less money, and recruit cheaper units. I don't know 100% that this is how things work, but I believe that they in fact do.

I am not accommodating the lackluster AI, I just want to field armies that feature combined arms. It's more fun for me to have a well balanced diverse army than to have an all-gothic knight army. Of course I want better quality units in my armies, but for me it's boring to have a single unit type army in the first place.

Yep, it took me one whole evening to mod EDB, but I think it was worth the time. I really want to play and enjoy the game according to what feels right for me.

[Von Nanega]

I think the best way to "slow" down unit production is to have a time set for recruiting elite type units. I you just captured a castle in the Holy Land that was a Muslim factions troop production center, it seems to be unlikely that a significant number of dismounted foot knights would be available for recruitment. Just the month before the castle was putting out islamic troop types! I have made my house rule be that untill I have owned the territory for ten turns I will not produce elite units. That makes things harder as to retrain, I have to sail them home, or wait ten turns!

[Jambo]

Upkeep will eat up AI finances as well, meaning it will have less money, and recruit cheaper units. I don't know 100% that this is how things work, but I believe that they in fact do.

True, but to combat that I've made a campaign script to give the AI a cash boost every turn. :)

I'm loathe to go into the unrest and public order modifiers since they just irritate me if they become to involved. Doubling or trebling the corruption modifier however, seems like a neat way to combat the wealth of large empires.

[Wardo]

Shogun and Medieval 1 dealt admirably with this problem.

Sort of. Since it had a "risk province map" going into war meant potentially facing the entire armed forces of your opponent, or a great part of it. So you didn't attack or declared war if you couldn't take the counter-attack, sea-invasions or the multiple stacks.

Since RTW, with the change to the 3D map, it doesn't work like this anymore. The Armies are spread all over the territory and you usually face no more than 2 stacks at a time. Sometimes, later in the game or in the case of the Mongols/Timurids, more stacks roam around together, but it's never a win-or-loose situation. You can beat the enemy and take a town, no matter, he will rebuild granted there is still something left. You can also loose your men, you will have time to rebuild aswell and re-capture any lost territory.

I consider the move to the new map very stupid. As beautiful as it is and as cool as it is to move around with the stacks, the strategic element was destroyed and it became a loooooooong game of attrition. Note well, the point here is not about the game lasting long, but that the game becomes reduced to a prolonged battle of attrition where strategy is irrelevant as long as you can win the tactical battles (diplomacy, what for? Decision, what for?). Since you are guaranteed to win most tactical battles, unless you fight severely outnumbered on purpose, only the initial years offer any strategic gameplay value where your decisions matter. To attack or not to attack later on is meaningless.

We seriously need to rethink the strategic map. The strategic element of the old map needs to return. When you declare war on somebody, you must face consequences.

[Philbert]

To be honest, I don't see what you describe exactly as a problem. I love the latter stages of the game, where I rule supreme and are the scourge of the world. Because of the struggling in the beginning, I also have a sense that I have earned it.

This is also the reason I am mostly a medium/medium player...

[Garnier]

Shogun and Medieval 1 dealt admirably with this problem.

Sort of. Since it had a "risk province map" going into war meant potentially facing the entire armed forces of your opponent, or a great part of it. So you didn't attack or declared war if you couldn't take the counter-attack, sea-invasions or the multiple stacks.

Since RTW, with the change to the 3D map, it doesn't work like this anymore. The Armies are spread all over the territory and you usually face no more than 2 stacks at a time. Sometimes, later in the game or in the case of the Mongols/Timurids, more stacks roam around together, but it's never a win-or-loose situation. You can beat the enemy and take a town, no matter, he will rebuild granted there is still something left. You can also loose your men, you will have time to rebuild aswell and re-capture any lost territory.

I consider the move to the new map very stupid. As beautiful as it is and as cool as it is to move around with the stacks, the strategic element was destroyed and it became a loooooooong game of attrition. Note well, the point here is not about the game lasting long, but that the game becomes reduced to a prolonged battle of attrition where strategy is irrelevant as long as you can win the tactical battles (diplomacy, what for? Decision, what for?). Since you are guaranteed to win most tactical battles, unless you fight severely outnumbered on purpose, only the initial years offer any strategic gameplay value where your decisions matter. To attack or not to attack later on is meaningless.

We seriously need to rethink the strategic map. The strategic element of the old map needs to return. When you declare war on somebody, you must face consequences.

I agree 100% with that post. The AI is simply too easy in too many ways, for this reason: The game was made too complicated for the AI to handle as well as the player. Giving the AI piles of money each turn seems to get them to recruit stacks of town militia, and I can't think of any way to fix this, Ive tried doubling the cost of the cheap militia units and making their recruitment pools slow to replenish and not very big, but they just ended up having LESS stacks then before, but the existing stacks were made up of town militia primarily.

I think I will be going back to SPQR TW again until a mod comes that solidly fixes the difficulty problem and makes the game interesting.

I think the City/castle system was poorly implemented especially as the AI is concerned.

[Lord_hazard]

Shogun and Medieval 1 dealt admirably with this problem.

Sort of. Since it had a "risk province map" going into war meant potentially facing the entire armed forces of your opponent, or a great part of it. So you didn't attack or declared war if you couldn't take the counter-attack, sea-invasions or the multiple stacks.

Since RTW, with the change to the 3D map, it doesn't work like this anymore. The Armies are spread all over the territory and you usually face no more than 2 stacks at a time. Sometimes, later in the game or in the case of the Mongols/Timurids, more stacks roam around together, but it's never a win-or-loose situation. You can beat the enemy and take a town, no matter, he will rebuild granted there is still something left. You can also loose your men, you will have time to rebuild aswell and re-capture any lost territory.

I consider the move to the new map very stupid. As beautiful as it is and as cool as it is to move around with the stacks, the strategic element was destroyed and it became a loooooooong game of attrition. Note well, the point here is not about the game lasting long, but that the game becomes reduced to a prolonged battle of attrition where strategy is irrelevant as long as you can win the tactical battles (diplomacy, what for? Decision, what for?). Since you are guaranteed to win most tactical battles, unless you fight severely outnumbered on purpose, only the initial years offer any strategic gameplay value where your decisions matter. To attack or not to attack later on is meaningless.

We seriously need to rethink the strategic map. The strategic element of the old map needs to return. When you declare war on somebody, you must face consequences.

Totally agree, when i play RTW and M2TW i feel that something is different like theres something missing, i didnt get the totalwar feeling when playing RTW and M2TW as when i played Shogun and MTW, and now i think i know why.
But to deal with the elite army issue, something ive been against from day one, maybe a simple limit to how many you can have in an army? Because having a limit similar to the one with the merchants wont eliminate elite armies simple make it so you can only have one elite army, which is one too many. Units like the scots guard should have a limit of one unit per army, etc.

[Malkut]

In trying to make the end-game (by which I mean the part of the game during which you control most of the world and have crushed anyone who ever had any chance of stopping you and are an invincible military power in the final stages of achieving your goals) more challenging, I believe you're ignoring another important part of the game which is way to easy: the AFTER-game.

Someone should write a script where, after you've achieved your campaign goals, finished the game, watched the ending, read the credits, and turned off your computer, a group of very large, hairy men with axes and funny accents break down your door and attack your home, and you have to defeat them with, uh, tactics, or something.

Also, enemy units should be able to fly, cavalry should be vulnerable to tall grass, and all computer factions should use the Holy Hand Grenade of Antioch against you at least once a battle.

[Lusted]

These good enough ai stacks for you?(all taken from LTC 2.0 with edited recruitment pools, no units recruitable from walls/castle building, and edited unit csots).













There are still some militia stacks running around but you cannot prevent that if the ai mostly has cities as all it can produce in cities is militia. This is the main reason behind milans milita spam.

[ArchieGremlin]

Personally I find that the problem is not that I build elite armies but that the AI fields trash.

Econ's house rules are very similar to my rules of thumb for constructing armies. They give me a well balanced force that's ready to take on anything the AI can throw at it. i.e. roughly 1/4 missile, 1/4 cav and 1/2 line infantry.

I just wish the AI would follow some sort of similar guidelines to limit the mix of troops in a stack. At the moment, it nearly always fields armies that are hopelessly out of whack. I'm sure you've all seen this sort of thing:

1. Milanese armies containing nothing but crossbowmen and siege weapons.

2. Danish armies composed of 1 unit of cav, 1 unit of archers and 12 units of heavy infantry.

3. Egyptian armies consisting of nothing but cavalry and a few peasants

Once the AI gets these armies in the field it then tends to walk around until it loses all cohesion. This begs the human player to defeat each part of the army in detail.

These battles are by far the most common in the few campaigns I've played. It's got to stage where I feel that I've failed if the AI escapes with more than 20 men or the kill ratio is worse than 10:1.

And yet... if I set up a custom battle and give the AI a decent army it can actually do a lot of damage. I usually win but only after taking a lot of losses. (Partly because I have no experience of fighting fair battles.) In these sorts of engagements the AI can kill half my force.

To come back to the original post, I think the reason it's tougher at the start of the campaign is that it's the only time that the AI armies contain a reasonable mixture of unit types. Once you destroy them the AI just replaces them with whatever's cheap and available.

The AI needs to be re-designed so it does what people do: field well balanced stacks containing the finest quality troops available. Unless it does that, it can't hope to compete 'cos even a mediocre army beats a rubbish one.

FWIW My campaigns have been vanilla 1.1.

It would be really nice if CA could allow for AI plugins that would allow the modders full control of both the battle and campaign AIs. I'm sure Lusted and the other mods could produce some truly stunning AIs. As Lusted pointed out elsewhere, we've got more time than the devs.

cheers,
Archie

[Lusted]

I'm sure Lusted and the other mods could produce some truly stunning AIs. As Lusted pointed out elsewhere, we've got more time than the devs.

See post above to see what i've been able to make the ai do.

[ArchieGremlin]

These good enough ai stacks for you?(all taken from LTC 2.0 with edited recruitment pools, no units recruitable from walls/castle building, and edited unit csots).

They look pretty good. Nice one. :~)

By the way, I didn't see this until after I posted my last reply. I hope no one thought I was having a go at you.

Hmmm. This leaves me with a tough question. Do I go for a LTC campaign now or wait for the patch. Sadly, I haven't got time for both. :~(

cheers,
Archie

[Garnier]

I just thought of something.

I will try making cities only able to recruit 2 units per turn, but castles be able to produce 4,5,6,7,8 depending on the level of development. This should induce them to recruit castle units should it not? I will test it I guess.

[Lusted]

That does help a bit, but if the ai only controls cities it can only produce miltia. And if it controls mostly cities it will produce mostly militia with 1 or 2 castle units mixed in. But if it controls a good mix it produces very nicely balanced armies.

[Garnier]

Since the AI doesnt convert cities to castles and vice versa (at least not that I know of) I have used a houserule that I dont do it either.

The city/castle system is simply badly implemented... ugh

Btw, as to those AI army pics you showed, yes they will get to those kinds of armies eventually, but I am always too many steps ahead of them, despite letting the AI control construction.

Btw, the AI uses military build policy on almost all of its provinces. A few cities use balanced, but thats it.

I will try to play a campaign of LTC 2.0 I guess... Do you think it would severely unbalance things to remove siege weapons and make castles be able to recruit lots of units per turn?

[Slyspy]

Shogun and Medieval 1 dealt admirably with this problem.

Sort of. Since it had a "risk province map" going into war meant potentially facing the entire armed forces of your opponent, or a great part of it. So you didn't attack or declared war if you couldn't take the counter-attack, sea-invasions or the multiple stacks.

Since RTW, with the change to the 3D map, it doesn't work like this anymore. The Armies are spread all over the territory and you usually face no more than 2 stacks at a time. Sometimes, later in the game or in the case of the Mongols/Timurids, more stacks roam around together, but it's never a win-or-loose situation. You can beat the enemy and take a town, no matter, he will rebuild granted there is still something left. You can also loose your men, you will have time to rebuild aswell and re-capture any lost territory.

I consider the move to the new map very stupid. As beautiful as it is and as cool as it is to move around with the stacks, the strategic element was destroyed and it became a loooooooong game of attrition. Note well, the point here is not about the game lasting long, but that the game becomes reduced to a prolonged battle of attrition where strategy is irrelevant as long as you can win the tactical battles (diplomacy, what for? Decision, what for?). Since you are guaranteed to win most tactical battles, unless you fight severely outnumbered on purpose, only the initial years offer any strategic gameplay value where your decisions matter. To attack or not to attack later on is meaningless.

We seriously need to rethink the strategic map. The strategic element of the old map needs to return. When you declare war on somebody, you must face consequences.

Thats it! Thats what my problem with RTW is. It really is one long war of attrition, which the player will inevitably dominate, even if he doesn't meet the objective. What is the point of playing if the end is already known and the path there is a drag?

[Garnier]

Yeah they do need to bring back the risk-style map in the next game for a good change. Though its not likely to happen as this is supposedly an improvement.

I suppose they could make it so that in a battle, all armies within a certain radius can arrive over time, depending on how far away they are.

[antisocialmunky]

The risk style map was fun but that doesn't mean it is better. I think that it's implementation was better than the 3D map implementation. If they put some decent effort into the campaign map AI in the next iteration of Total War so that it'll recognize bottle necks and be able to concentrate force effectively, then I think that the 3D map will be much more interesting than the risk map.

In fact, that's the the main problem with TW, crappy AI. There's been a definite trend of AI not keeping up with the rest of the game design whether it be the sheer quanity of units and weaker RPS system in MTW or the shift to the 3D engine in RTW and M2:TW. It has always been the weakest link in a great franchise.

[Carl]

The problem is the movment rates. In reality an army could march the length on England in a month easy. In game it's a 10+ year slog. If armies could move 5 times as far each turn you could get a lot more armies to trouble spots a lot faster and so could the enemy, thus aiding defence. Whilst offence would be limited by the need to stop to seige every city. Especially if it was made so that an attack on a city prevented and army from moving furthar that turn, that way you couldn't use spies to make multipiule conquests per army per turn.