Just Stroll on Down to the Enemy Capital?

[Malagant]

I believe the problem is more the ease with which supply depots can be placed in uncontrolled territory. Many of us believe that a depot should require a friendly military presence when not in friendly territory.

[carnifex]

So, what exactly is happening that is the problem here? Is it that AI enemies tend to ignore cities on the way to a capital? Or is it that AI enemies don't surrender fast enough after losing a large battle? (Or both?)

IMHO the problem is that while on the offensive, the AI focuses on the capital to the exclusion of other factors. In the Napoleonic Wars, the enemy capital was just something you took while trying to defeat the enemy's ability to fight, and in the age of nearly self-sustaining armies this meant actually destroying the enemy force.

On offense, the AI should do what a human player would do. Concentrate and beat the main enemy stack in a pitched battle, then pursue them into their own territory, sieging and taking provinces on the way to the capital. The enemy army should always be the main focus of any thrust.

[EarlPembroke]

I think there are 2 issues being discussed, and they are tied together:

1) The thread started with: the AI doesn't try to eliminate your depots, letting you string supply too easily. Personally, I've had cossacks, etc. disrupt my supply line, so I don't think this alone is as much the issue.

2) To me it's more a matter of the AI doesn't defend its territory, but rather sends its army out to conquer the other land. In my current game, Austria and Turkey have been at war several times. Their largest armies often do not meet on the field of battle. Rather Austria sends its largest army off to siege Turkish cities and leaves a smaller force behind to defend. The huge Turkish army descends on Vienna and sieges it, taking it. Austria's largest army keeps sieging a Turkish city until Austria is forced to capitulate. Or until Turkey is beaten by Russia and thus surrenders to Austria too.

I think Austria should try to seek out the Turkish army (lower morale) and face it with a more equal force to prevent it from taking Vienna so easily. As a human player I can do this, but the AI doen't do it much. I also hope that making a tweak here would lead to more offers of limited surrender. (Unless all those messages I see about AI nations surrendering to each other include some limiteds. The AI doesn't send me offers of limited surrender; instead it waits until it is fully beaten then has a full surrender.)

#2 could be due to the enemy capital being weighted too heavily in the AI's priority. Of course that's probably tied to the effect it has on surrender. If the AI was more focused on your army's movements (including supply line) it would have a more realistic feel.

[Naomi]

I doubt AI's ability to co-ordinate the actions of allies' troops. I saw France kept defeating them piecemeal while they were just scattered everywhere.

[Malagant]

ORIGINAL: EarlPembroke

I think there are 2 issues being discussed, and they are tied together:

1) The thread started with: the AI doesn't try to eliminate your depots, letting you string supply too easily. Personally, I've had cossacks, etc. disrupt my supply line, so I don't think this alone is as much the issue.

2) To me it's more a matter of the AI doesn't defend its territory, but rather sends its army out to conquer the other land. In my current game, Austria and Turkey have been at war several times. Their largest armies often do not meet on the field of battle. Rather Austria sends its largest army off to siege Turkish cities and leaves a smaller force behind to defend. The huge Turkish army descends on Vienna and sieges it, taking it. Austria's largest army keeps sieging a Turkish city until Austria is forced to capitulate. Or until Turkey is beaten by Russia and thus surrenders to Austria too.

I think Austria should try to seek out the Turkish army (lower morale) and face it with a more equal force to prevent it from taking Vienna so easily. As a human player I can do this, but the AI doen't do it much. I also hope that making a tweak here would lead to more offers of limited surrender. (Unless all those messages I see about AI nations surrendering to each other include some limiteds. The AI doesn't send me offers of limited surrender; instead it waits until it is fully beaten then has a full surrender.)

#2 could be due to the enemy capital being weighted too heavily in the AI's priority. Of course that's probably tied to the effect it has on surrender. If the AI was more focused on your army's movements (including supply line) it would have a more realistic feel.

Napoleon agreed with you:

"There are in Europe many good generals, but they see too many things at once. I see one thing, namely the enemy's main body. I try to crush it, confident that secondary matters will then settle themselves." -NB

[mogami]

Hi, I began a 1796 scenario (I've started playing this later start to allow France to get Nappy) I was minding my own business when I observed a large build up by Turkey on Austrias border. Not wanting Austria to get beat up further and seeing Turkey in second place with over 400 points I moved the Baltic fleet to the Black Sea combined it with my Black Sea fleet and moved 200,000 troops to the Turkish Border.
Now it seems I should have made an Alliance with Austria because after I declared war on Turkey I moved down and fought a series of battles pushing the Turks back to their capital but before I could follow Austria declared war on me and strung the depots out towards Moscow. (Held by a Korps with 40,000 men )
I had retained 5 infantry and 2 Cav divisions as independant units to guard my supply lines so I just send the 2 cav to the rear Austrian depots and destroyed them. Turned around my Army in Turkey and using the other indepedent infantry to prevent Austrian retreats destroyed the Austrian Army in Smolensk.
Of course the entire exercise to date has been a failure because I didn't prevent the continued collapse of Austria but instead killed the main field Army myself.
The Turks are all gathered at Constantinople and I am no closer to ending that war then I was when it began.
However the Russian Army did gain a lot of morale.
Now I have around 50,000 Turk POW and 60,000 Austrian POW building developments at Kharkov (guarded by 30,000 Cossacks) I don't have problems with POW and I use the Cossacks for suppressing disgruntled Russians and guarding POWs.

[ericbabe]

Teach the AI to move out to cut lines of retreat
Probably difficult to implement. Not sure. What do you think, Eric?

AI's popping units out of cities to destroy depots is a good idea, and easily doable. Other players on other threads are complaining that units like cossacks are running around cutting their depot chains -- the AI already does target supply lines to some extent, though things might always be improved with some tweaking.


Eric

[jchastain]

ORIGINAL: ericbabe

Teach the AI to move out to cut lines of retreat
Probably difficult to implement. Not sure. What do you think, Eric?

AI's popping units out of cities to destroy depots is a good idea, and easily doable. Other players on other threads are complaining that units like cossacks are running around cutting their depot chains -- the AI already does target supply lines to some extent, though things might always be improved with some tweaking.

Eric

Let's try and be clear on this one.

Small bands of troops running around in enemy territory wreaking havoc = bad

Small groups of troops preventing enemy supply lines from running through their own territory = good

Control of a province should matter. When you own a province, you should be able to count on indiginous troops and police forces to prevent a militarily insignificant roaming band from disrupting everything. Having to chase down every single division on the board in some bizarre game of keepaway isn't fun. This is a real example of where a land based intercept would help immensely. Or perhaps a force that isn't large enough to siege the local city would not be eligible to free prisoners? But one half filled division shouldn't be able to disrupt an entire nation. And when you don't own a province, you shouldn't be able to roll unguarded supply wagons through unmolested. There must be a reason to siege the city and control the province or at least to leave troops in place to guard the supply lines.

[mogami]

Hi, A full strength division is 10,000 men. This is not a military insignificant force. Napoleons first Army only totaled 45,000 men. I often send a division from one of my province (with depot) into enemy territory to impede their movement, supply and communication lines. I will send any where from 1 to 3 such divisions. (if I am going to send more then 30,000 troops I send a HQ with a leader)

When the mararders are groups of 3k or 4k it is one thing when they are full strength divisions it is another.

[jchastain]

ORIGINAL: Mogami

Hi, A full strength division is 10,000 men. This is not a military insignificant force. Napoleons first Army only totaled 45,000 men. I often send a division from one of my province (with depot) into enemy territory to impede their movement, supply and communication lines. I will send any where from 1 to 3 such divisions. (if I am going to send more then 30,000 troops I send a HQ with a leader)

When the mararders are groups of 3k or 4k it is one thing when they are full strength divisions it is another.

Rather than getting sucked into the debate on whether a division could or would ever have functioned in such a way, the expected responses of homeland forces represented by the militias within cities if it ever did happen, and the weight of any outcomes upon the overall strategic situation and therefore whether or not it should have any place within a grand strategy title, I'll just say that it is no fun to have it happen in a game and leave it at that. [8D]

[Latour_Maubourg]

Nappie went twice to Vienna to find the main Austrian army north of the city not caring about the capital. The Ruskies didn't mind Moscow being burnt in 1812 and withdrew east, north and south (technically St. Petersburg was the capital and not Moscow, but that's another story). So what is the significance of the capital? Which brings us to the point when should a country surrender? A) after the capital has been taken B)after the main army has been defeated? Well A is easier to program then B, right? But a lot of us agree to the fact that the capital isn't the main goal. So what to do? My humble suggestion is: a country should surrender if the total amount of troops is below a certain percentage compared to the total population or surrender if the total amount of troops falls below for example 40% of total troops or whatever percentage you suggest.

just my 2 cents.....


L-M

[Malagant]

I like the current system of NML, but I think the hit to NML for having your capitol held is a bit too steep, making the A/I (and humans!) ignore everything else.

[Naomi]

ORIGINAL: Malagant

I like the current system of NML, but I think the hit to NML for having your capitol held is a bit too steep, making the A/I (and humans!) ignore everything else.

Maybe gold bullion and art treasure shouldn't 've been stored only in capitals; (the mere worry of) losing them seemed to weigh very heavily.

[EarlPembroke]

ORIGINAL: Malagant

ORIGINAL: EarlPembroke

2) To me it's more a matter of the AI doesn't defend its territory, but rather sends its army out to conquer the other land. In my current game, Austria and Turkey have been at war several times. Their largest armies often do not meet on the field of battle. Rather Austria sends its largest army off to siege Turkish cities and leaves a smaller force behind to defend. The huge Turkish army descends on Vienna and sieges it, taking it. Austria's largest army keeps sieging a Turkish city until Austria is forced to capitulate. Or until Turkey is beaten by Russia and thus surrenders to Austria too.

I think Austria should try to seek out the Turkish army (lower morale) and face it with a more equal force to prevent it from taking Vienna so easily. As a human player I can do this, but the AI doen't do it much. I also hope that making a tweak here would lead to more offers of limited surrender. (Unless all those messages I see about AI nations surrendering to each other include some limiteds. The AI doesn't send me offers of limited surrender; instead it waits until it is fully beaten then has a full surrender.)

Napoleon agreed with you:

"There are in Europe many good generals, but they see too many things at once. I see one thing, namely the enemy's main body. I try to crush it, confident that secondary matters will then settle themselves." -NB

Thanks - good to hear it from a real expert on Napoleonic warfare. [:D]

[EarlPembroke]

ORIGINAL: Malagant

I like the current system of NML, but I think the hit to NML for having your capitol held is a bit too steep, making the A/I (and humans!) ignore everything else.

Agreed - I like the NML method. I think there should be a trigger that makes attempts at limited surrender - seems underutilized to me.

I wonder if the capital weighting is what leads the enemy to march headlong to the other nation's capital? But your own capital should weigh as much as the enemy's - that might make AI nations leave some troops to guard the homeland if they attempt an invasion. Somehow the large armies need to seek each other out, I think (esp. in the AI-AI wars as humans can do this on their own). Per the quote from Napoleon above, enemy armies might maneuver for better position / choose location to fight, but at least one was seeking out the other's main body to defeat it. Think we have a bit too much emphasis on sieging the enemy's lands as it stands now. Might be hard to code: "When you have an advantage, go seek out the other nation's main army to destroy it and win the war." [;)]