Castles will have a Zone of Control?

正在查看此主题的用户

Tyson_48

Regular
Is there any dev blog or similar to this?
Will finally a castle located on a choke point prevent enemies from marching through?
 
Tyson_48 说:
Is there any dev blog or similar to this?
Will finally a castle located on a choke point prevent enemies from marching through?

Possibly. We know settlements are physical objects and that parties must travel around them to reach the gates.
So if a castle is placed in a mountain pass like Nelag and Sungetche castle are in warband, then it may prevent enemies from using the mountain pass.
 
AmateurHetman 说:
Tyson_48 说:
Is there any dev blog or similar to this?
Will finally a castle located on a choke point prevent enemies from marching through?

Possibly. We know settlements are physical objects and that parties must travel around them to reach the gates.
So if a castle is placed in a mountain pass like Nelag and Sungetche castle are in warband, then it may prevent enemies from using the mountain pass.
Wouldn't then plugging a gap with a fort prevent ANY party from moving through?

Edit: having thought about is further, I think map designers should not block a pass completely with a fort, but allow small paths around it.
Then, if an enemy party tries to wiggle through, teeny tiny archers from the fort would rain down some arrows on them and cause casualties. This would happen at any town or fortress if an enemy gets too close.
The orthodox zone of control concept is too mechanic and doesn't model reality as well as small warriors pooping over your party on the world map.
 
If there is a big enough garrison to leave the town and attack the enemy, then that would serve as the "zone of control". If a Vlandian army is marching past an Empire castle, it could be attacked by the garrison. If the garrison is not big enough to take the Vlandians on in full battle, it could still follow at a distance and then sandwich the Vlandians against a larger Empire force.
 
I definitely think they should have a DoC or else their usefulness would be quite significantly curtailed. A medieval army on campaign would rarely leave an even lightly defended enemy castle in their rear without at least leaving a small force to invest. Personally I'd prefer either an attrition penalty moving through a DoC like in total war or perhaps similar to Sid Meier's Pirates! where a hostile fort would shoot at your ship on the map and do damage when not engaged directly on the battlemap. 
 
I think getting shot with arrows when going very close to a settlement/castle would be good. It would make it a bit costly to go too near these sites and it is realistic; even if not in range of the walls, you might be harassed by skirmishers who retreat to the castle as soon as you send troops to deal with them.
 
Rodrigo Ribaldo 说:
Wouldn't then plugging a gap with a fort prevent ANY party from moving through?

Edit: having thought about is further, I think map designers should not block a pass completely with a fort, but allow small paths around it.
Then, if an enemy party tries to wiggle through, teeny tiny archers from the fort would rain down some arrows on them and cause casualties. This would happen at any town or fortress if an enemy gets too close.
The orthodox zone of control concept is too mechanic and doesn't model reality as well as small warriors pooping over your party on the world map.

If the castle has gates on each side, friendly parties and caravans could pass through the settlement that way maybe.
 
Rodrigo Ribaldo 说:
Then, if an enemy party tries to wiggle through, teeny tiny archers from the fort would rain down some arrows on them and cause casualties. This would happen at any town or fortress if an enemy gets too close.

Why arrows? TW have already demonstrated their animations of siege engines on the world map. The defenders should be catapulting boulders or fire at any trespassers.
 
NPC99 说:
Rodrigo Ribaldo 说:
Then, if an enemy party tries to wiggle through, teeny tiny archers from the fort would rain down some arrows on them and cause casualties. This would happen at any town or fortress if an enemy gets too close.

Why arrows? TW have already demonstrated their animations of siege engines on the world map. The defenders should be catapulting boulders or fire at any trespassers.
That would be a welcome overreaction.

AmateurHetman 说:
Rodrigo Ribaldo 说:
Wouldn't then plugging a gap with a fort prevent ANY party from moving through?

Edit: having thought about is further, I think map designers should not block a pass completely with a fort, but allow small paths around it.
Then, if an enemy party tries to wiggle through, teeny tiny archers from the fort would rain down some arrows on them and cause casualties. This would happen at any town or fortress if an enemy gets too close.
The orthodox zone of control concept is too mechanic and doesn't model reality as well as small warriors pooping over your party on the world map.

If the castle has gates on each side, friendly parties and caravans could pass through the settlement that way maybe.
Yes, that seems likely, now that you said it. If they are not lazy about special pathfinding coding.
 
Rodrigo Ribaldo 说:
NPC99 说:
Why arrows? TW have already demonstrated their animations of siege engines on the world map. The defenders should be catapulting boulders or fire at any trespassers.
That would be a welcome overreaction.
Picture this: you're a single ale trader, travelling with your 6 mules and your 8 bodyguards on horseback. You reach a new town from a different faction to see how prices are there. For some reason the townsguard perceive you as an enemy intruder, they immediately load and fire up their mangonels, trebuchets and scorpions, they start throwing flaming rocks and darts at your party from 400m away. You and your bodyguards die after being hit by 45 flaming rocks, and one particular guard was impaled with his horse by a flaming giant arrow.

Seems reasonable.

Rodrigo Ribaldo 说:
AmateurHetman 说:
If the castle has gates on each side, friendly parties and caravans could pass through the settlement that way maybe.
Yes, that seems likely, now that you said it. If they are not lazy about special pathfinding coding.
Gotta maintain those hope™s high!




Gods, we can only hope™
 
Depending on how they've coded the pathfinding, it might actually be near impossible to get the AI to choose another route besides the absolute quickest one. In warband this was the case, unless you coded your own system from scratch (which would be laggy as all hell because there is no efficient way to access the map data with the exposed modding tools).

This would result in a scenario where the AI just walks past your forts because it's the quickest route and then gets shredded by attrition or arrows or whatever it is.

Also historically the main strategic purpose of forts was to control supply and foraging and prevent an army from being able to safely spread out. If a fort isn't in a pass or choke point, and someone just walks around it slightly, then what's even the point?
 
Designating areas of the campaign map with different terrain types allows us to do more than just match battle scenes to parts of the map, it also allows us to control AI behaviour (for example, limiting the areas where certain bandit parties can roam) and affect party modifiers (such as movement speed).
https://steamcommunity.com/games/261550/announcements/detail/1681414081363392053

While I’m hopeful that this will give modders greater influence on ai pathing, I remain concerned about walled centres. Every video I’ve seen, shows a golden circle with a single access point, irrespective of the direction of travel. If towns/castles only have one gate they will never be choke points in mountain passes etc.
 
NPC99 说:
Designating areas of the campaign map with different terrain types allows us to do more than just match battle scenes to parts of the map, it also allows us to control AI behaviour (for example, limiting the areas where certain bandit parties can roam) and affect party modifiers (such as movement speed).

I am afraid, there's little in that quote that clearly points at path finding. Bandit roaming can simply mean spawn points.

NPC99 说:
While I’m hopeful that this will give modders greater influence on ai pathing

Adding modifiers to the pathfinding algorithm makes pathfinding calculation slower. Which is why developers avoid it. Unless devs will hard code it in to their algorithm, moding it in will either be impossible or will potentially create issues with frame rates.

NPC99 说:
I remain concerned about walled centres. Every video I’ve seen, shows a golden circle with a single access point, irrespective of the direction of travel. If towns/castles only have one gate they will never be choke points in mountain passes etc.

Turning towns or castles in to choke-points would totally screw AI. It would require some sort of strategic AI of the sort that newer was in the MB game and likely newer will be because of the character of the game (where every agent on the map acts on his own). Therefore chance for such choke points appearing in the game is close to zero.
 
NPC99 说:
Designating areas of the campaign map with different terrain types allows us to do more than just match battle scenes to parts of the map, it also allows us to control AI behaviour (for example, limiting the areas where certain bandit parties can roam) and affect party modifiers (such as movement speed).
https://steamcommunity.com/games/261550/announcements/detail/1681414081363392053

While I’m hopeful that this will give modders greater influence on ai pathing, I remain concerned about walled centres. Every video I’ve seen, shows a golden circle with a single access point, irrespective of the direction of travel. If towns/castles only have one gate they will never be choke points in mountain passes etc.
Using special terrain codes to cover the pass around the fort is the cheap way of controlling which parties may pass and is probably why Taleworlds would do it this way (if they do it at all). Handling passes in the pathfinding code is the correct solution, even if it's dirty in the sense of low-level code accessing high-level data.
 
hruza 说:
NPC99 说:
Designating areas of the campaign map with different terrain types allows us to do more than just match battle scenes to parts of the map, it also allows us to control AI behaviour (for example, limiting the areas where certain bandit parties can roam) and affect party modifiers (such as movement speed).

I am afraid, there's little in that quote that clearly points at path finding. Bandit roaming can simply mean spawn points.

Limiting the areas where certain parties can roam is quite specific. In Warband parties roam considerable distances from their spawn points - merely relocating spawn markers would not justify TW’s wording,

NPC99 说:
While I’m hopeful that this will give modders greater influence on ai pathing

Adding modifiers to the pathfinding algorithm makes pathfinding calculation slower. Which is why developers avoid it. Unless devs will hard code it in to their algorithm, moding it in will either be impossible or will potentially create issues with frame rates.

Designating areas of the campaign map with different terrain types allows us to do more than just match battle scenes to parts of the map, it also allows us to control AI behaviour (for example, limiting the areas where certain bandit parties can roam) and affect party modifiers (such as movement speed). We think that modders will find these terrain types to be quite useful as they can be used in a number of inventive ways.

NPC99 说:
I remain concerned about walled centres. Every video I’ve seen, shows a golden circle with a single access point, irrespective of the direction of travel. If towns/castles only have one gate they will never be choke points in mountain passes etc.

Turning towns or castles in to choke-points would totally screw AI. It would require some sort of strategic AI of the sort that newer was in the MB game and likely newer will be because of the character of the game (where every agent on the map acts on his own). Therefore chance for such choke points appearing in the game is close to zero.

Warband’s campaign map ai pathing copes perfectly well with impassable terrain (rivers, mountains, seas). Flagging terrain as impassable or passable by party type shouldn’t be a problem as the pathing for each party will be calculated separately and in such calculations it will either be impassable or passable.

The only terrain types that strained Warband’s ai pathing were too many ford triangles or unelevated mountain triangles.
 
Perhaps we will find answers to our questions on this issue when Taleworlds decides to talk to us about it on the promised part 2 campaign map blog.  :iamamoron:
 
NPC99 说:
Limiting the areas where certain parties can roam is quite specific. In Warband parties roam considerable distances from their spawn points - merely relocating spawn markers would not justify TW’s wording,

Designating areas of the campaign map with different terrain types allows us to do more than just match battle scenes to parts of the map, it also allows us to control AI behaviour (for example, limiting the areas where certain bandit parties can roam) and affect party modifiers (such as movement speed). We think that modders will find these terrain types to be quite useful as they can be used in a number of inventive ways.

As I said, there's little in that quote indicating any relation to pathfinding. AI is much more than just pathfinding. TW’s wording is vague and open to interpretation.

NPC99 说:
Warband’s campaign map ai pathing copes perfectly well with impassable terrain (rivers, mountains, seas). Flagging terrain as impassable or passable by party type shouldn’t be a problem as the pathing for each party will be calculated separately and in such calculations it will either be impassable or passable.

The only terrain types that strained Warband’s ai pathing were too many ford triangles or unelevated mountain triangles.

I said problem for AI, not for pathfinding. What I mean is what will AI do if you block only road possible to a destination? And even in cases there is pathfinding through some longer route around, will AI happily run around half of the map just to get from a village to a town market and back? Correct solution would be to stop trade and send some request up the AI chain to the king or marshal to unblock the road by taking the fortress in the choke point.

Will AI be able to do it? No it won't. MB doesn't have any strategic AI layer of the sort and won't have. Implementing this feature without programing AI to be able to correctly cope with it would only lead to yet another abuse of the "stupid" AI by the player.
 
The algorithms used for node based pathfinding like in warband are things like Dijkstra and A* which can be programmed in all sorts of ways which include avoiding certain types of terrain, but as madvader said it's not always ideal to have such a basic level operation like pathfinding have access to high level variable data like the terrain type. There are certain optimisations they might have put in (including calculating it on a separate thread) which make checking for terrain type on the fly prone to glitches.
 
I like that idea.

Along with it, I suggest having different sieging scenes for that kind of castle being attacked from different sides (or different spawing location in the same scene).
 
后退
顶部 底部