Redesigning the total war battle system from scratch

[Darth Wong]

I'd like to see a real AI; instead of the retarded Flank them = Win battle automatically one presently in the game.

It's not unreasonable that you win if you can flank the enemy. That should work. The problem is that it's so easy to flank the enemy; the AI is too stupid to respond to it or avoid opening himself up for such a maneuver.

[ray245]

I'd like to see a real AI; instead of the retarded Flank them = Win battle automatically one presently in the game.

Every total war fan knows the fact that the AI sucks. However, I am talking about flaws in the gameplay, flaws that is noticeable and annoying even when you are playing a multiplayer match.

[MKSheppard]

It's not unreasonable that you win if you can flank the enemy. That should work. The problem is that it's so easy to flank the enemy; the AI is too stupid to respond to it or avoid opening himself up for such a maneuver.

It's not that hard to note the use of flanking manouvers by a human player, since the best troops for flanking are your general's bodyguards, elite cavlary etc.

I think a good way to simulate an AI learning would be after the first couple of battles you play in which your cavalry/bodyguards get excess kills via flanking, the AI starts to put cavalry in it's stacks, and in battle, deploys them around it's main body's flanks to act as spoilers for any attack..

[PainRack]

I have no complaints about the system(well, other than naval combat, but I appreciate how that would lag down my hard disk space further).

What I do have to complain is the stupid AI.

The enemy AI has virtually no intelligence whatsoever once you outnumber the enemy. They would just twist and face a unit with no threat, opening themselves up to be charged and destroyed by units in close.


System wise, LOS is the main problem. The system doesn't really have any good benefits to keep hidden, and LOS is too difficult to maintain anyway.

[AniThyng]

Another point to note, why do they have to restrict our army to 20 units? When the size of our army is restricted, there is no point in outnumbering an enemy army.

If the AI is dumb, at the least it has an advantage by using a much larger army, which makes the game more challenging.

Sure there is a point - your units that didn't get to participate are there to reinforce the depleted army that did participate.

I suppose the game could be rebalanced so that it should be infeasible to make full stack armies all over the place, and the "average" stack should be about half or 2/3rds the engine set maximum.

[Fire Fly]

Larger armies is always a plus; the 20 stack armies are relatively find for the ancient and medieval period but for ETW, it just seem so pathetically small.

I would also like to see a forced march option for the campaign screen: when you select an army stack to march, two or three concentric circles will highlight around the army stack, a green one, a light green and a red one. The green circle signifies how far the army can march without being exhausted if it should ever enter combat, a light green one for a slightly exhausted army while the red circle signifies a more tired army but an army that moved further.



To simulate supply lines, one could assign a city or even a fort as a supply depot for a given army stack or stacks and if the army stacks move beyond a certain radius of this city, there will be small decreases in the available supplies should the stack enter combat and it might take an extra turn to fully resupply the stack. If the stack stays out of reach of it's supply depot, it will suffer gradual loss in available supplies. This would mean that if the stack engages in battle twice in a player's turn, it might lose all of it's ammunition and can no longer engage. Upon assignment, the city or fort will take n turns to accumulate enough supplies for it to properly distribute and to make them worth while targets to attack and defend. Or something along this line; I haven't completely thought this through yet but I think it could easily work within the present Total War engine.

On using your generals to engage in combat, I agree with the idea that if the general dies, then the player loses control of the battle and the AI takes over. A simple solution that doesn't require many contraptions to make it work.

[ray245]

Another point to make is, the outcome of a battle has no real impact on the campaign map.

What happens when you lost one full stack army? You can simply retrain another army like no bodies business. The lost an army should at the least put fear into an enemy's citizens. Or, the lost of one major army should prompt the AI to seek for peace, and cities flocking to your domain.

[GuppyShark]

I consider full stack armies to be the basic unit of maneuver.

The AI is capable enough that if you march three full stack armies up to an enemy force, you can control one while allowing the AI to handle the other two, especially in MTW2 which provides a rudimentary "aggressive/defensive/shootout" command system for the allied forces. You can outnumber the enemy, you just currently have to surrender control.

My main gripe with the Total War series is how frustratingly sieges are modelled.

[ray245]

I consider full stack armies to be the basic unit of maneuver.

The AI is capable enough that if you march three full stack armies up to an enemy force, you can control one while allowing the AI to handle the other two, especially in MTW2 which provides a rudimentary "aggressive/defensive/shootout" command system for the allied forces. You can outnumber the enemy, you just currently have to surrender control.

My main gripe with the Total War series is how frustratingly sieges are modelled.

The thing is, you can give coordinate other stacks to support your own stack. And the AI loves to waste your units lives.

[wautd]

Instead of the generic hills, grass and some trees I'd like to see a much more detailed battlefield (a swamp area that would slow down heavy cavalry, dense forest that'll give a bonus to skirmishes, ...).

[ray245]

Instead of the generic hills, grass and some trees I'd like to see a much more detailed battlefield (a swamp area that would slow down heavy cavalry, dense forest that'll give a bonus to skirmishes, ...).

From what I know about ETW, muddy fields will slow down your units, and your artillery can get stuck in the mud.

[Samuel]

Another point to make is, the outcome of a battle has no real impact on the campaign map.

What happens when you lost one full stack army? You can simply retrain another army like no bodies business. The lost an army should at the least put fear into an enemy's citizens. Or, the lost of one major army should prompt the AI to seek for peace, and cities flocking to your domain.

Cannae? The Romans lost a whole army... and retrained another from scratch. Thanas can give you the details, but I think Roman casulties were 80000? At this time they only had Italy, so they didn't have such a large manpower reserve.

Of course, this could be unique to the Romans, but the Persians lost three major armies to Alexander (the Strap's and the two main ones).

[Thanas]

^Yes, Samuel is correct. Though Cannae was obviously a huge loss, considering the romans started to draft slaves and elderly.

Polybius writes 70.000 slain, 10.000 captured (80.000), Livius names 50.000 slain or captured.

Then again, people forget that at that time they were already fighting in Spain, so if necessary, they could have pulled legions from there, but they did not.

[ray245]

Another point to make is, the outcome of a battle has no real impact on the campaign map.

What happens when you lost one full stack army? You can simply retrain another army like no bodies business. The lost an army should at the least put fear into an enemy's citizens. Or, the lost of one major army should prompt the AI to seek for peace, and cities flocking to your domain.

Cannae? The Romans lost a whole army... and retrained another from scratch. Thanas can give you the details, but I think Roman casulties were 80000? At this time they only had Italy, so they didn't have such a large manpower reserve.

Of course, this could be unique to the Romans, but the Persians lost three major armies to Alexander (the Strap's and the two main ones).

The problem is, in RTW, we have to siege every single city to take down that faction, and end the war. If not, the AI would just ignore the loss and their population stays remained as loyal as ever, to their leader.

[RogueIce]

Another point to make is, the outcome of a battle has no real impact on the campaign map.

What happens when you lost one full stack army? You can simply retrain another army like no bodies business. The lost an army should at the least put fear into an enemy's citizens. Or, the lost of one major army should prompt the AI to seek for peace, and cities flocking to your domain.

Cannae? The Romans lost a whole army... and retrained another from scratch. Thanas can give you the details, but I think Roman casulties were 80000? At this time they only had Italy, so they didn't have such a large manpower reserve.

Of course, this could be unique to the Romans, but the Persians lost three major armies to Alexander (the Strap's and the two main ones).

The problem is, in RTW, we have to siege every single city to take down that faction, and end the war. If not, the AI would just ignore the loss and their population stays remained as loyal as ever, to their leader.

Yes, but that's not really the combat system, but more the strategic AI, and most probably the utterly broken diplomacy. Such as the infamous "Freedom is not given up so lightly..." response when you demand they become a protectorate, despite the fact they're down to one city with only a few peasents and militia defending while you have three full stack armies next to them.

[Thanas]

Diplomacy allegedly will be fixed in Empire, but that is what they promised for MTW2 as well.

[Artemas]

Another point to make is, the outcome of a battle has no real impact on the campaign map.

What happens when you lost one full stack army? You can simply retrain another army like no bodies business. The lost an army should at the least put fear into an enemy's citizens. Or, the lost of one major army should prompt the AI to seek for peace, and cities flocking to your domain.

Mention of Cannae has already been made, but as another example, look to the Franco-Prussian War. While French Regulars had been captured or immobilised, the country still managed to raise 5 new armies, in 5 weeks totalling 500,000 men.

Second, the AI demanding peace really has nothing to do with the battle system. There are diplomacy mods out for M2TW that fix this in part. Like capturing a castle, then demanding a town in return for a cease-fire. Or post-war settlement trading. Another problem it that most countries in the game are only represented by a half-dozen major settlements, giving away one is likely to cripple that faction for the rest of the game.

The real problem is that annihilating an army happens so freguently. If the complete destruction of an army was a rare thing indeed, then some major unrest in cities, or susceptibilty to demands might work better, but as it is Cannae is the rule (as the player), rather than the exception. So, until they figure out how to vastly improve the survivability of armies, then the social and political effects of complete destruction of an army should be put on hold.


Really, the only major problem with the battle system is the stupid AI. Everything else is just trying to fix the symptoms.

[RogueIce]

Mention of Cannae has already been made, but as another example, look to the Franco-Prussian War. While French Regulars had been captured or immobilised, the country still managed to raise 5 new armies, in 5 weeks totalling 500,000 men.

Second, the AI demanding peace really has nothing to do with the battle system. There are diplomacy mods out for M2TW that fix this in part. Like capturing a castle, then demanding a town in return for a cease-fire. Or post-war settlement trading. Another problem it that most countries in the game are only represented by a half-dozen major settlements, giving away one is likely to cripple that faction for the rest of the game.

The real problem is that annihilating an army happens so freguently. If the complete destruction of an army was a rare thing indeed, then some major unrest in cities, or susceptibilty to demands might work better, but as it is Cannae is the rule (as the player), rather than the exception. So, until they figure out how to vastly improve the survivability of armies, then the social and political effects of complete destruction of an army should be put on hold.

The only way I could think of to do that offhand is what someone else suggested earlier: make cavalry more expensive or somehow more difficult to produce. I'm no TW combat expert mind you, but whenever I've made those "pretty much all army wiped out" battles (whether through full kills or just enough so they rout and disappear from the map) it's been from cavalry. My foot troops just can't catch up with fleeing soldiers.

Not sure how else to really tweak that, though.

[Thanas]

The problem is you only really need your general or 1-2 cavalry troops to flank or destroy a fleeing army.

[Artemas]

Mention of Cannae has already been made, but as another example, look to the Franco-Prussian War. While French Regulars had been captured or immobilised, the country still managed to raise 5 new armies, in 5 weeks totalling 500,000 men.

Second, the AI demanding peace really has nothing to do with the battle system. There are diplomacy mods out for M2TW that fix this in part. Like capturing a castle, then demanding a town in return for a cease-fire. Or post-war settlement trading. Another problem it that most countries in the game are only represented by a half-dozen major settlements, giving away one is likely to cripple that faction for the rest of the game.

The real problem is that annihilating an army happens so freguently. If the complete destruction of an army was a rare thing indeed, then some major unrest in cities, or susceptibilty to demands might work better, but as it is Cannae is the rule (as the player), rather than the exception. So, until they figure out how to vastly improve the survivability of armies, then the social and political effects of complete destruction of an army should be put on hold.

The only way I could think of to do that offhand is what someone else suggested earlier: make cavalry more expensive or somehow more difficult to produce. I'm no TW combat expert mind you, but whenever I've made those "pretty much all army wiped out" battles (whether through full kills or just enough so they rout and disappear from the map) it's been from cavalry. My foot troops just can't catch up with fleeing soldiers.

Not sure how else to really tweak that, though.

Except making cavalry more expensive is just tacking a symtpom, and not the core issue (stupid AI). Limiting cavalry applies to both the player as well as the AI, and since the AI cant properly USE cavalry, you get the same problems, but with less horses.

But lets assume that the AI can not be suitably improved, what options then are available to us?

-Limit the units that players abuse, and the AI cannot use (cavalry more expensive)

-Prevent the after rout mopping up

-Allow routing units to rally more often, even after the entire army has begun to rout. Units can still rally, but will try to make their way off the field.

-Put a wounded soldier recoup rate to armies that allows a defeated army to retain 30-80% of its force. So even if you decisively defeat a particular army, a not-insignificant portion of it will survive. (armies already regain a number of men that were casualties, but this only applies to victorious armies, and seems to be semi-random)

Ideally, the AI would be improved to such a point where the computer maintains a reserve force that can be used to stall the enemy from chasing down routers, allowing them enough time to escape.
Or an "orderly withdrawal" option that allows units to slowly move backwards?

[Artemas]

I was just thinking about how the more freeform map and freedom to act has resulted in the campaign AI suffering from idiocy, while the constrained choices it had to make in the old risk-style map resulted in more effective stategy, and got to thinking on terrain generation.

Often, I will station my army on some hilly or mountainous terrain, only for the terrain advantage be given to an attacker when battle is joined. I wonder if it would be possible to, perhaps, set up a colour coded "advantage index", where the terrain advantage can be roughly gauged from the campaign map. Where red is assaulting across a river, grey is even, yellow is a slight height advantage, and green is "first, you must navigate through my rabid weasel traps". Perhaps have a colour coding for both attack and defence for a particular tile (IE, attacking might be red, but defending is yellow, etc).

I'm not sure if this would be possible given the terrain generation, but I would be glad to see it.

[GuppyShark]

-Prevent the after rout mopping up

My vote would be for this option. Honestly, I'm just getting sick of doing it. You feel that if you don't, the enemy will have more troops for the next pitched engagement but it stopped being fun a long time ago.

-Allow routing units to rally more often, even after the entire army has begun to rout. Units can still rally, but will try to make their way off the field.

That would at least make the post battle mop up interesting.

-Put a wounded soldier recoup rate to armies that allows a defeated army to retain 30-80% of its force. So even if you decisively defeat a particular army, a not-insignificant portion of it will survive. (armies already regain a number of men that were casualties, but this only applies to victorious armies, and seems to be semi-random)

Defeated forces do regenerate troops, but that assumes enough escaped to make a retreat (a force that "melts away" will obviously not). I've had defeated forces regenerate wounded before.

[Artemas]

Defeated forces do regenerate troops, but that assumes enough escaped to make a retreat (a force that "melts away" will obviously not). I've had defeated forces regenerate wounded before.

I've not noticed this, even when its me losing the battle. But greatly improving the ratio, would result in less substantial defeats, and at least limit "100% enemy casualty syndrome".

But then, if field battles are only rarely decisive, then the game devolves into a game of siege chess. Which isn't really fun.

What to do?

[Artemas]

BECAUSE I CAN'T EDIT: Maybe allow a full stack army to have 26 units, but only have 20 of them on the battle map. This allows the AI to effectively have a reserve force, without a massive intelligence upgrade. Once the enemy units start routing, you would still have to be cautious because of the potential 700-800 men coming onto the field.

This, by the way, was what happened in the original MTW. Only 16 units could be on the field, but there might be 2 or 3 armies worth as reinforcment. Definately reduced the effectiveness of mopping-up.

[Stark]

The rout phase could easily just be a single button; 'release'. Once the enemy is broken you can just release the army to chase them down (or at least a few units like cavalry) without having to micro it to get MAXIMUM ROUTAGE. It would also be mitigated if the enemy lost a proportion of the routed troops to desertion based on quality and morale.

But then this thread is a giant collision between gamers and simulationists, so I doubt either of these would be acceptable.