(Suggestion) Strategic AI targets

Users who are viewing this thread

italiano

Grandmaster Knight
I know this has been brought up before (it certainly has in other sections of the site), but I think the script that determines the target for a campaign should place far more emphasis on distance than the strength of the defence of the target settlement. I have the Khergits currently fighting over two castles (one Swadian, one now owned by the Vaegirs) near Suno. I think only "border" settlements should be considered potential targets. It gives the AI a better chance to hold onto its gains, lords don't have to risk travelling across the entire map, campaigns are over quicker allowing for better protection of the kingdom etc.
 
Oh yeah. Would make supplying easier aswell. (IRL. Maybe Supply trains could be used to go between towns to sieging armies?) ANyway, I strongly support distance being more important.
 
Add my support to this as well + another suggestion: make armies defend newly captured castles instead of rushing off to siege a new one.
 
daumor said:
Or at least the lord it is awarded to, eh?

Nah, newly captured castles take some time to get awarded to lords, during this time is when it's at its most vulnerable (and also when the enemies tend to strike) which is why the army or a detachment of it ought to defend it after capturing it while waiting for it to get awarded. Once awarded the garrison will increase a bit more and then it's up to the lord who got it to defend it.
 
There are a couple of reasons not to make distance the sole criteria. Basically, it would mean that border villages would get hit again and again.

In the new AI system, lords will have somewhat different criteria for selecting villages to target. Some go to the easiest target, some choose the richest target, and others choose villages owned by a lord with whom they have a personal grudge. You should be able to find out a lord's rationale by speaking to him. This isn't quite ready yet, though.

Faction armies, however, might be directed more towards the most vulnerable spots.
 
amade said:
daumor said:
Or at least the lord it is awarded to, eh?

Nah, newly captured castles take some time to get awarded to lords, during this time is when it's at its most vulnerable (and also when the enemies tend to strike) which is why the army or a detachment of it ought to defend it after capturing it while waiting for it to get awarded. Once awarded the garrison will increase a bit more and then it's up to the lord who got it to defend it.

But wasn't this dealt with by the "script_order_best_besieger_party_to_guard_center" script? Or you mean after the campaign is over and the center is still getting filled with troops?
 
Sarejo said:
amade said:
daumor said:
Or at least the lord it is awarded to, eh?

Nah, newly captured castles take some time to get awarded to lords, during this time is when it's at its most vulnerable (and also when the enemies tend to strike) which is why the army or a detachment of it ought to defend it after capturing it while waiting for it to get awarded. Once awarded the garrison will increase a bit more and then it's up to the lord who got it to defend it.

But wasn't this dealt with by the "script_order_best_besieger_party_to_guard_center" script? Or you mean after the campaign is over and the center is still getting filled with troops?

In the few times I've followed campaign armies I note that either a few break off to go home or that the whole lot goes off to siege another castle. I did notice that a party always stayed behind near a newly captured castle back in 1.011 but if I'm not mistaken they had "waiting" status instead of "patrolling". Even then, they don't wait too long and are often helpless when the whole enemy campaign army rides around to retake the castle. The guard center script needs a boost IMO, so that a newly captured center don't change hand too frequently.
 
nijis said:
There are a couple of reasons not to make distance the sole criteria. Basically, it would mean that border villages would get hit again and again.

In the new AI system, lords will have somewhat different criteria for selecting villages to target. Some go to the easiest target, some choose the richest target, and others choose villages owned by a lord with whom they have a personal grudge. You should be able to find out a lord's rationale by speaking to him. This isn't quite ready yet, though.

Faction armies, however, might be directed more towards the most vulnerable spots.

I think distance versus vulnerability would be the best mix. Because the most vulnerable spot may not be the most strategically viable spot. But the bet spot strategically might be too well defended...
 
Border villages might be hit more often but couldn't the ai for when to take a castle or town be weighted much more strongly to distance?  Isn't deciding to raid a separate objective from sieging?  I would love to see a gradually expanding border rather than the random conquest of castles scattered all over the map that happens now.  While we are talking about borders would it be possible for the ai to guard border castles much more often than the interior ones?
 
nijis said:
There are a couple of reasons not to make distance the sole criteria. Basically, it would mean that border villages would get hit again and again.

Sorry, I should have said I'm talking about castles (and towns). I have no problem with the liberal targeting of villages, but I think targets for territorial gain should be mainly on the border, and slightly further (roughly the next castle/town beyond the border, allowing the faction to cut off the border castles/towns). While it makes sense that a vulnerable target is chosen, consideration to its defensive capabilities after it is captured also has to be considered. A close but harder target is a far sounder target than one that is easy, but far away and undefendable. In theory it's good that they select targets based on riches, vulnerability and grudges, but if the AI can't hang on to a castle, why take it in the first place?

I just don't see it being worth the many high risks. Thre's risk in travelling through enemy land to reach the target. Unless it's a large campaigning force, this is a risk. There's risk in besieging, which again is a great risk unless the force it very large. There's risk in not being able to defend their own lands. There's great risk in having to travel back through enemy lands to reach home. There's great risk that the farmer parties, caravans and villages constantly raided, significantly lowering any economic benefits. Finally, there's a great risk that the lord won't be able to defend the castle, and that it will be taken soon afterwards. That's a lot of risk for a settlement that will constantly be raided, besieged, and likely lost very quickly. A player wouldn't make this mistake, so why should the AI?

Besterker Pride said:
While we are talking about borders would it be possible for the ai to guard border castles much more often than the interior ones?

Lord patrols or the garrisons? It doesn't bother me that lords are out and about, but border garrisons probably should be larger relative to others. I think the main problem though is a response to a settlement being besieged. It seems to me that the castle will always be taken, and rarely is the siege lifted. I'd like to see garrisons increased (or defensive bonus to the garrison - better?) to a point where any castle is hard to take by force, often forcing armies to starve them out, leaving themselves vulnerable to a force raised to lift the siege. I also think the first priority of lords should be to defend their own lands rather than invade others. If a campaigning army invades, I think the defending army should aim to drive them off, rather than conduct their own invasion.
 
nijis said:
There are a couple of reasons not to make distance the sole criteria. Basically, it would mean that border villages would get hit again and again.

What's wrong with that? I mean hell it's fairly realistic to boot. Borders might get hit over and over but overtime those zones of conflict would shift towards a side. Great Marshall campaigns would be more successful at pushing it one direction and has less odds of random lords running off after targets if it's on the border.

Sounds like a win-win proper way they would fight kinda thing to me. Heck if distance was also a priority to picking targets you'd have less random wars across the map too.
 
Borders being war torn sems the right direction. When you become a vassal you get teh poorest village in the faction. If border settlements are getting pillaged constantly then it would add a bit of a climb to the mid game,  which would be nice. and you'd never be bored...
 
Back
Top Bottom