Idea for improving AI's use of forts

[Ahriman]

On a more practical level; the idea of improvements that you can build in enemy territory has a number of problems; it wipes the existing improvement, which even pillage can't do, and is particularly nasty to do to great person improvements.

 

[Dunkah]

So you don't like the idea... that is clear.

I don't like the point that we can wage war, fortify ourselves in a hex, conduct army operations and not have any effect on the terrain.

We fight a battle for years in game time move off to another city and voila the terrain is as pristeen as it was the day we improved it.

Diggin in/ fighting should have effects on the terrain we are waging war in.

 

[Ahriman]

I don't like the point that we can wage war, fortify ourselves in a hex, conduct army operations and not have any effect on the terrain.

You stop that tile from being worked. If you want, you can pillage the improvement. So yeah, you affect the terrain.

We fight a battle for years in game time move off to another city and voila the terrain is as pristeen as it was the day we improved it.

But it has lost years of productivity.

Diggin in/ fighting should have effects on the terrain we are waging war in.

There are very few historic examples where large areas of land were devastated in such a way by war that it would take decades to recover production. 30 years war Germany, maybe. But otherwise, what is your model here? Belgian agriculture was pretty productive again not long after WW1 ended.

In previous versions of Civ, improvements were a binary thing, there, or not there. We now have a third option; there, not there, or damaged. This was a big improvement. We now can absolutely model the kind of temporary damage you're talking about, but without needing to completely reconstruct the improvement from scratch.
This is as it should be; military occupation hardly ever puts an area of land back to productivity as if it were wilderness. They might burn some buildings, kill some peasants, cut some railway lines, and so forth, but they seldom destroy towns and roads and farms entirely to the point that they can't recover.

For example:
The idea that Carthage was sown with salt and was unuseable afterwards is a myth. The Romans razed the city, but the farmland recovered quickly and was parceled out to Romans.

*edit*
And again; why should you be able to destroy great person improvements, or to permanently destroy farms or villages or mines?

 

[Dunkah]

Point is that someone had to come along and reclaim that land. In game terms a worker. There are still today large areas of France that still have large Trenchworks dug into them. Parts of the middle east that have large areas of duded shells and landmines that will never be cleaned up.

If the Trenches improvement is not used then the land stays as is. Thus not needing to improve it. The land is only damaged as much as you choose to damage it. So it scales with the length and ferocity of the battle.

Who says that the improvement you are replacing needs to be destroyed. Why not have it put the old improvement into the same state as if the hex was pillaged. 5 turns and you get whatever was there back again?

If it is your intent to take the land you are fighting in then you wouldn't want to destroy an improvement to begin with. Unless you really needed to set up a defensive line due to your main army being locked up somewhere else.

Great Person improvements should be indestructable? Would suck to lose one I agree. But all the more reason to defend it/ put it in a spot that won't be vulnerable. Or you could simply not allow trenches to be built on Great Person Improved tiles?

 

[Ahriman]

Point is that someone had to come along and reclaim that land. In game terms a worker.

If you pillage a tile a worker has to come to repair it to make it operable again. We have this already. Surely you'll agree that some repair work is easier than literally rebuilding everything from scratch, as if it were wilderness that had never been farmed?

There are still today large areas of France that still have large Trenchworks dug into them. Parts of the middle east that have large areas of duded shells and landmines that will never be cleaned up.

Define "large areas". At the scale of a full hex, dozens of kilometers across? That have a significant impact on regional agricultural yields?

If the Trenches improvement is not used then the land stays as is. Thus not needing to improve it. The land is only damaged as much as you choose to damage it.

Sounds like pillaging to me, which you get to chose to do.

Who says that the improvement you are replacing needs to be destroyed.

The game code. You can only have 1 improvement on a tile at a time.

Why not have it put the old improvement into the same state as if the hex was pillaged.

Why not just pillage and then fortify your unit?

If it is your intent to take the land you are fighting in then you wouldn't want to destroy an improvement to begin with.

Then don't pillage.

Unless you really needed to set up a defensive line due to your main army being locked up somewhere else.

Why do you think you should rapidly be able to set up a powerful defensive line in enemy territory, than can survive while your own army is off elsewhere?
If you want a defensive line for holding off the enemy, then do it in your own territory. Territorial control should be meaningful during wartime; you should be much worse off in my territory than in your own.

Great Person improvements should be indestructable?

Yes. If you could lose them then it would be super-not-fun, and you wouldn't bother to use great people for improvements, better to use their abilities or golden age them.

But most importantly
I can't imagine how you think you the AI would use a mechanism like this intelligently (constructing improvements in enemy territory during wartime).
The AI already is bad enough at pillaging, which is comparatively simple.

 

[Thalassicus]

There are very few historic examples where large areas of land were devastated in such a way by war that it would take decades to recover production.

Land mines. Those damned things last forever.

I agree there's no need to represent them in Civ, however. It's a modern invention that wouldn't apply to most of the game, and the current pillaging system is close enough. The AI just needs to pillage more! Then we could make the rewards for pillaging greater, and battlefields would truly look devastated.

 

[Ahriman]

Land mines. Those damned things last forever.

Sure, but the areas with significant numbers of land mines are still pretty small by world standards. There are still some areas in Cambodia, Vietnam, Sudan, Mozambique, Western Sahara, and a few other places with a fair number of mines, but not large areas by world standards - so not covering a sufficient area to reduce a tile by a single yield point.

 

[Sneaks]

Land mines. Those damned things last forever.

I agree there's no need to represent them in Civ, however. It's a modern invention that wouldn't apply to most of the game, and the current pillaging system is close enough. The AI just needs to pillage more! Then we could make the rewards for pillaging greater, and battlefields would truly look devastated.

I believe if we double the value of
<Row>
<Type>TACTICAL_PILLAGE</Type>
<OperationsCanRecruit>false</OperationsCanRecruit>
<OffenseFlavorWeight>100</OffenseFlavorWeight>
<Priority>40</Priority>
</Row>

then the AI should choose pillaging over other combat options in more scenarios..

 

[Seek]

The AI just needs to pillage more! Then we could make the rewards for pillaging greater, and battlefields would truly look devastated.

This would be fantastic - I hope Sneaks' suggesion works!

 

[Ahriman]

Forcing the AI to pillage more (instead of attacking) has the risk of reducing its ability to actually take cities, so we'd need to be careful to test this.

 

[Thalassicus]

I'm not sure changing TACTICAL_PILLAGE values does anything. With settings higher than anything else in the TacticalMoves table (even 999), an enemy unit placed in a field of improvements still ignores them.

 

[Sneaks]

Odd. Not entirely sure how priority levels equal out

 

[tlaurila]

Any chance to bring this conversation back on track from additional fortification-esque things?

I think (what I think!) the point still remains rather untouched:
Current forts aren't used by anybody for anything really. They take too long to build and are too expensive (both in terms of cost and lost yields). So,
- Can we improve forts so that they are worthwhile, bringing a stale element back into the game?
- Can we make the AI use forts somehow usefully? Just because forts don't belong to useful human strategy, doesn't mean AI couldn't be improved by their use. In particular, if forts provided extra ranged cover, AI could use them to foil the most prevalent human tactic: outmaneuver and bombard to bits.

 

[Ahriman]

I think forts are doomed to always be a minor part of the game as long as they are an actual tile improvement, and so replace an existing unit.

But frankly, I think this is fine. I don't think they should be a large part of the game.

Forts *should* have a long build-time. Almost by definition, they are supposed to represent the kind of fortifications that take years to build, and so have to be built before a war rather than during it.

I also think that forts boost strength, and so they make you take less from ranged damage already.

 

[Olleus]

How about this?

Leave fortifying as it is.

Make building forts consume a worker but be built in 1 turn. Have the civ4/Thalassicus's idea of them being mini cities. Give them perhaps a strength equal to that of a size 5 city for that era. Can't bombard, but can be garrisoned by a unit, gives the same healing as a city and has ZoC even without a garrison. Have them cost the same in maintenance as a unit.

Have citadels be the same, but with a stronger defence strength and doing the auto damage it does now.


This way forts will do what forts are meant to do. Being able to hold the line on their own for a short while even without your 'standing army' supporting them; while also being a rallying when your main army gets there. Citadels are the same but uber. The insta build by a worker means that they can be put up quickly in emergencies in war, but the fact that it consumes them and costs maintenance makes it unproductive to spam them.


How the AI could be taught to use this is beyond me though.

 

[tlaurila]

I think forts are doomed to always be a minor part of the game as long as they are an actual tile improvement, and so replace an existing unit.

That just depends on what is the actual cost of a fort. If, for example, fort gave 2 gold base, like a village, but didn't get any of village's boosts, then it might support the historical progression of castles dotting landscape in the middle ages, but vanishing later on.

Almost by definition, they are supposed to represent the kind of fortifications that take years to build, and so have to be built before a war rather than during it.

Same is true for roads. You might have a castle have the same build time as a road, not 3 times more than road (or whatisit).

I also think that forts boost strength, and so they make you take less from ranged damage already.

Yes, yes they do. But if they're not useful, then they don't do this enough. Again, AI fortifying cities with forts that have additional defenses against ranged attacks would incline the human to think more varied strategies than the outmaneuver +bombardments galore. And I think that would be a good thing.

 

[Ahriman]

Leave fortifying as it is.

Yes, fortification works fine.

Make building forts consume a worker but be built in 1 turn. Have the civ4/Thalassicus's idea of them being mini cities.

Being able to plop down a castle in the middle of a war, when the enemy is right at your door, seems absurd. Making them into impassible wars that will take several turns for the opponent to destroy, for the cost of a single worker, would be incredibly powerful. So in the middle of a war, I just bring workers along and all of a sudden my artillery unit that you're about to kill is massively protected.

Being able to hold the line on their own for a short while even without your 'standing army' supporting them; while also being a rallying when your main army gets there.

This is what cities are for. There is no way you should be able to do this for just the cost of a worker. If you want a long-term fortification that protects your unit and is hard to break into, then build a city.

The AI already struggles to capture cities; can you imagine how much worse it would be if you could add extra cities that it needed siege units to take (or melee units taking lots of damage) in the way?

The insta build by a worker means that they can be put up quickly in emergencies in war

Why is this desirable or realistic?

How the AI could be taught to use this is beyond me though.

Which is a dealbreaker.
* * *

If, for example, fort gave 2 gold base, like a village, but didn't get any of village's boosts

Then I would spam them in the early game to put my archers in them and make it even more impossible for the AI to attack me successfully, and then by midgame (before Economics), when I've already built improvements anywhere, I can use my worker downtime (saving them for railroads) to replace them with trading posts.

Same is true for roads. You might have a castle have the same build time as a road, not 3 times more than road (or whatisit).

This makes no sense to me. Castles usually took a decade to build. A small stretch of road didn't.

But if they're not useful, then they don't do this enough.

They are useful already, in rare, situational circumstances.
The problem with forts is not that they don't give a big enough boost, it is that wars don't usually take place in a small predictable area, and if they do, it is near your cities, and you want to be able to work those tiles.

AI fortifying cities with forts that have additional defenses against ranged attacks would incline the human to think more varied strategies than the outmaneuver +bombardments galore.

Fortifying cities and protecting units inside them is already modeled with the Walls/Castle/Arsenal/Military base building line.
If the AI had stronger defensive bonuses, then this would just encourage you to focus *even more* on lots of bombardments, because attacking with melee troops would be even more costly, why ranged attacks would remain costless.

 

[tlaurila]

If the AI had stronger defensive bonuses, then this would just encourage you to focus *even more* on lots of bombardments, because attacking with melee troops would be even more costly, why ranged attacks would remain costless.

No, of course it didn't if forts only provided more bonus against ranged, like the cover promotion.

Fortifying cities and protecting units inside them is already modeled with the Walls/Castle/Arsenal/Military base building line.

That's a good idea If those buildings could provide a defensive bonus to units stationed next to the city to make conquest harder. Something like 10% per building on defense, plus another 10% per building against bombardment.

 

[Ahriman]

No, of course it didn't if forts only provided more bonus against ranged, like the cover promotion.

An improvement that provides protection *only* against ranged attacks. That seems really weak and bizarre.

If those buildings could provide a defensive bonus to units stationed next to the city to make conquest harder

The human would understand such a bonus, the AI wouldn't. So it would tend to increase the human player's ability to defend their cities without boosting the AI's ability much.

 

[tlaurila]

An improvement that provides protection *only* against ranged attacks. That seems really weak and bizarre.

Of course not. Give +50% defense from forts like they do now (works against melee and ranged) AND +50% defense against ranged (like cover). So you have MORE protection bonus against ranged.

The human would understand such a bonus, the AI wouldn't. So it would tend to increase the human player's ability to defend their cities without boosting the AI's ability much.

That's the beauty of it, I don't think the AI even needs any extra understanding. Analyze an average game you played recently, and if it had any conquest, I bet you that you attacked many more AI units next to their cities, than they did yours. Due to AI hammer&gold bonuses, if this game was at any higher difficulties, I bet those cities had plenty defensive structures as well.

Note that this bonus would not give much benefit to masses of ranged units defending a city. Which is the human trump card. Those ranged units would be better protected when attacked, but they get no attack bonuses. It would help the human defend cities, no doubt, but not in the most common form, that the city itself soaks up the hits.

Sure the AI might play stupid and you might coax their units to leave further away from cities to die. But that takes time and effort, and all in all, I bet the game will become harder for the human. Especially if the bonuses given to ranged defense are more significant than against melee attacks, it will squarly play against the main human tactic for obliterating the AI.