Would improving AI use of forts improve overall AI gameplay?

[Kyroshill]

We've all seen it..... forts in the BFC placed on otherwise usable tiles.... and even then... these forts often sit right outside the city, and are unguarded.... almost like the AI built you a siege platform.

So I wonder.... if the AI could somehow be properly taught how to use forts (or at least not to spam them) how far into other gameplay would the effects flow?

1. It makes it slightly less easy to use AI fortifications against them.

2. Adding one or two productive tiles to a city's BFC could improve a city's production, commerce, or flexibility between the two.



Are there other advantages? And does any of this matter?

Would be nice to find a simple change in the code the creates a small cascade of improvements..... then again.... could be a pipe dream

 

[Honor]

How about completely removing the AI's fort-building algo? I never build them unless there happens to be juicy choke point nearby, but that is very rare.

 

[Demus]

teaching the AI how to use spring would be a huge leap in this area tbh, gets rid of all those desert based forts.

 

[JonathanStrange]

Ever since I was a young kid, the AI has generously built forts for my troops and my father's troops before me; such are the mysterious ways of the AI.

I sometimes build forts in such a way that they're out of the way (so if they're occupied they provide little use to an enemy), but still close enough to give it's protective defense, esp. once it upgrades to the higher forts. And my Ljosalfars are known to have a forest citadel behind a city placed in such a way it covers more than one city.

 

[Lade]

Its so true, I once managed to bring a small stack outside an Illian city on a fortress, I used it as a springboard to take THEIR city after they suicided 3/4 of their defenders against an archer...

I know this is a little off topic, but should city raider and city garrison promo's have the same effects on forts? I would think forts and city walls would be pretty similar

 

[PaleJackal]

I'd be perfectly okay with the AI building no forts whatsoever.

 

[megamanx06]

i agree with no forts, especially for automated workers. Although, there have been a couple times where I've seen an AI unit on a fort. Just one though, not stacks like they have in cities.

 

[ICanHasName?]

IF the AI is going to continue building forts, it needs do so in a way that doesn't benefit the player. 1) it should never ever build them next to a city. 2) it should only build them on the inside of their empire (so never between a city and a border). 3) It should never build them on a non desert non ice tile unless there are still decent tiles the city isn't working.

Otherwise, the AI just shouldn't build forts because it hurts them.

 

[schlalex]

Make it like in the old Civs. Fortresses have a ZoC, so u cannot move from one adjacent tile to another. That finally would strengthen defensive wars, because the fortresses are outside your BFC, so u can wait, until the enemy attacks your well fortified units while your BFC is safe from pillaging.
I hate the civ concept, that only "offensive" defensive wars are effective.

 

[JonathanStrange]

ZoC could be interesting; I've always liked the idea.

Could the AI use forts in a defensive line? Would they just spam forts in random fashion? If they did use forts, I bet they'd construct a Maginot line against an imaginary attack, nowhere near a probable invasion route. There are likely workarounds for this; say preventing fort spamming by having a maintenance cost after a certain number are built. C

Certain civs, like the Dwarves, would be obvious candidates for use of forts in this manner.

 

[Luckmann]

I'm not sure about the opponent AI, but if they can be seperated, the player's automated workers should definately not build forts on their own. It's ridiculous. They place them everywhere, in the most usable tiles, in city limits. If they just built them where I didn't have coverage, it'd be fine. But they don't.

Blech!

 

[LDiCesare]

The game would be better with no fort whatsoever the way they are currently implemented. They are totally useless for the ai. For players, it's really a niche. The defense bonus sounds gamey and doesn't make sense (build forts farther from the invasion???)
ZoC might be useful, but you'd need quite a few to get any result.
Another idea would be to make forts zero-move units which can attack adjacent tiles with an arrow spell or some such. They could also provide a defense bonus to units in their tile, but at least they'd serve a purpose and need to be destroyed.

 

[Brokenbone]

Guilty of automating workers if I'm not in any kind of pressured/desparate situation. Not a fan of their building forts when they can't seem to think of anything better (i.e., deserts end up full of forts).

Whether this means trying to "improve" the logic like noticing chokepoints, especially with hills, or ensuring they weren't too close to cities or to each other, or just making the automated building invalid, I don't know. If forts became a human-only thing, maybe that would be bad too though (as in, human gets a tool for intelligently defending turf that the AI would be completely denied).

 

[snarko]

The AI is completly denied that tool already. It would be a big improvement to have them stop building pointless, sometimes harmful, forts. Harmful for the civ that build it that is - that is how poorly they are using them.

In my last game I saw the AI replace a town with a fort. On a grassland plot surrounded by hills. With no intention of defending said fort. That's INSANE. They'd be much better off never having built that fort at all.

 

[Kyroshill]

OK.... so if the simple answer is to make it impossible for the AI to build forts, then they will be forced to improve every improvable tile.

So the question remains.... does this have a great or small impact on overall AI gameplay?

Would those extra productive tiles add appreciably to AI cities... and if so.... would the extra production translate into something even better for the AI.... greater than the sum of parts?

 

[WarKirby]

If they can't be taught how to use them then removing them from AI hands seems like a good idea.

I wouldn't want them completely removed, though. I really like forts, and there are situations where using them is a good idea. Especially in chokepoints.

 

[jprc]

KYROSHILL: So the question remains.... does this have a great or small impact on overall AI gameplay?

As forts are only built when you have automated workers... which is always the case for your AI opponents, and sometimes you (your choice):
- It is a VARIABLE impact as you can always make a better use of your land if you micromanage each worker. This impact goes actually from mild to big depending on your level (Prince, Monarch, Deity...). Higher is your level, thinner is your line! Each little thing that helps you and gives you an advantage on the AI can not be neglected.
- It has a HUGE impact on your patience though. Automated worker working with common sense free your mind from daily management routine at the office... No general in a battle should worry about knowing if his socks will be cleaned with the right soap... If you grow frustrated controlling your mad workers, it won't help for making better decisions...
- It has little impact if you are in multiplayer; players wasting time with not automated workers often end up with too little time for monitoring other things and their time is out before they can do all job. Hence, a majority of multiplayers go for automated, or swith to automated and "do not built on top of old improvements" at a later stage of the game (when starts the betrayal between "friends"...)

 

[JonathanStrange]

The impact on the AI's gameplay if it could use forts properly?

Probably very little. The AI already, at the higher levels, produces and supports an insane amount of units and research, so the value of an extra tile here and there isn't that significant overall. In the case of the AI building a fort adjacent to a city and then not garrisoning it, it would help but I can't see it being decisive overall either: we'd miss the little grin from the free fortified siege camp, but c'est la guerre.

(The non-AI's will like not seeing evidence of the AI's cluelessness and thus the illusion that the non-AI is playing a cunning and competent opponent can be maintained a turn or so longer.)

 

[ICanHasName?]

The fort next to the city thing can provide a pretty big advantage, especially when the AI tries to actively defend their cities by suiciding half their defenders into your new fortress.

Also, I have occasionally seen some awful tile wastage, particularly with desert cities that only have a few usable flood plains.

 

[Jabie]

Give forts bonuses. +1 Gold for a Castle, +1 Gold and +1 Hammer for a Citadel, +1 Gold at the Feudalism tech, +1 Hammer if running Aristocracy. It might not solve the problem but it would make Forts less wasteful and it's a few easy XML hacks.

Also what about having Towers morph into Forts after 20 turns under your control. Towers speed up the exploration phase of the game and it seems natural that these places would be turned into defensive locations after a while.