AI Can't Play Dural

[Pudding]

The computer finally spawned a Dural in my last game, and then promptly died within 100 turns on Marathon. I reloaded and used Worldbuilder to see what happened and, to my amazement, the Dural player had not built a single improvement around their one city. So I deleted the monsters that were about to overrun them, played out a few hundred more turns, and then decided to check on them. Lo and behold, still no improvements.

What the hell? The D'tesh that have spawned also die, and I thought it might be their build mechanic, but a Scions AI did just fine. So I have no idea what's up. Any idea why this is, and if so, any way I can make the computer not be so thick?

 

[Camper Joe]

The computer finally spawned a Dural in my last game, and then promptly died within 100 turns on Marathon. I reloaded and used Worldbuilder to see what happened and, to my amazement, the Dural player had not built a single improvement around their one city. So I deleted the monsters that were about to overrun them, played out a few hundred more turns, and then decided to check on them. Lo and behold, still no improvements.

What the hell? The D'tesh that have spawned also die, and I thought it might be their build mechanic, but a Scions AI did just fine. So I have no idea what's up. Any idea why this is, and if so, any way I can make the computer not be so thick?

I have seen the Dural fail miserably before, and it doesn't make much sense to me. Isn't the AI's basic bare minimum scripting for building improvements at least present? Now, I have seen the D'tesh fall many times because they rely on their 2 strength warriors. They do not know that the arcane line holds priority for them- and I don't think AI can efficiently use magic, as I have not seen it done so far. Kinda makes for an interesting game, when you're the only one that can pull spectres and wraiths outta your a**. Haha.

 

[isthmus]

It's likely that either the workers that the dural are making are getting eaten straight away (pretty common) or that they focus on getting education related techs and might miss out on the basics. I'm not sure what their tech priorities are though.
It takes so long for a worker to make the improvement, that the worker gets eaten before it can finish.
Then again, they could just be unlucky, pillaged to hell by orthus/zarcaz/hill giants and never got the momentum to fend these guys off their lands. I've seen it happen to other civs too in this version.

 

[Camper Joe]

Then again, they could just be unlucky, pillaged to hell by orthus/zarcaz/hill giants and never got the momentum to fend these guys off their lands. I've seen it happen to other civs too in this version.

Zarcaz has happened to me all too many times. That seems likely- but it seems that the Dural are almost always doing absolutely dreadful.

 

[admtanaka]

Sometimes, I must wonder if the AI would do better if all their workers started with hardy I and II and were forbidden from building improvements for which they didn't have the speed-increasing tech.

 

[Kalina]

Sometimes, I must wonder if the AI would do better if all their workers started with hardy I and II and were forbidden from building improvements for which they didn't have the speed-increasing tech.

Hardy I+II for workers will cause AI to suicide them against barbs Orbis has tried it and it doesn't work well.

 

[Humakty]

Is underdevelopment specifically tied to Dural ? My game showed Ais were underdevelopped as a whole.

 

[Evalis]

What happens is the ai workers will each pick a tile on their own to construct something. Often this will be a lumbermill (a 30 turn mission). If this worker isn't killed by then, it still takes them an ungodly ammount of time to complete it. They choose their improvements based on how much a benefit that resource will provide, not how long it will take. This also slows their scientific growth tremendously, and slows their construction speed. If you happen to have timid animals off, or play with barbarian hordes, the computer will never complete an improvement. I've never found the dural to play any worse than other computers. Although the losaljafar are always doing well (probably due to peace with animals).

 

[Valkrionn]

Ljos aren't at peace with animals.

Also, next version the Mills will be unavailable at start.

 

[Rapier]

Frankly I'm not sure there's any advantage to being able to build the improvements before you have the tech to make them faster - as far as I can tell all this does is disadvantage the computer that picks randomly or poorly.

There are a few times when getting access to a resource earlier than the tech could is beneficial (usually happiness resources that need plantation) but even then I find mostly waiting is better unless my workers have nothing t do.

 

[shoggi]

If i'm not wrong you don't actually get the ressource if you don't have the appropriate tech. So you may build the plantation early to make the plot better, but you don't get the ressource.
It's the same thing with the metals. You discover Copper and can build mines on it one tech before you actually get the ressource.

 

[admtanaka]

Frankly I'm not sure there's any advantage to being able to build the improvements before you have the tech to make them faster - as far as I can tell all this does is disadvantage the computer that picks randomly or poorly.

There are a few times when getting access to a resource earlier than the tech could is beneficial (usually happiness resources that need plantation) but even then I find mostly waiting is better unless my workers have nothing t do.

I agree with this, and I've always been curious why the decision was made to allow construction of improvements without the formerly-requisite technologies. If anything, it takes away part of the strategy from picking techs early in the game and seems to be hampering the AI (which was already struggling before having more sub-optimal choices from which to pick).

I very rarely make use of the ability. Sometimes, as another poster noted, I will improve a plantation resource before I tech calendar, but only when my start is rather commerce poor. Calendar is usually a relatively-prioritized tech anyway. Even this probably isn't optimal, because the resource isn't actually used until you get the required technology - it's just 20 worker turns for 5 or so commerce.

 

[lordrune]

An AI Malakim in a previous game (current public patch) had the same issue. 200 turns in, he hadn't built one improvement. His city wasn't on a river, so maybe Varn Gosem was upset by the lack of food. The city had five warriors in it, and he had no workers - I don't know if he had tried to build any at any stage.

 

[Jabie]

Steal the proffesions from Colonization (or was it the AoD / AofD modmod) for Pioneers, Miners, etc. Whenever the Dural produce a Worker it should get a random profession. That profession then doubles the speed of building that type of improvement. Pioneers build roads at 2x speed, Miners build mines at 2x speed, Plantation workers build Plantations at 2x speed, etc.

Fits in with their ethos and shouldn't be too difficult to add to RiFe.

 

[Korias]

Steal the proffesions from Colonization (or was it the AoD / AofD modmod) for Pioneers, Miners, etc. Whenever the Dural produce a Worker it should get a random profession. That profession then doubles the speed of building that type of improvement. Pioneers build roads at 2x speed, Miners build mines at 2x speed, Plantation workers build Plantations at 2x speed, etc.

Fits in with their ethos and shouldn't be too difficult to add to RiFe.

Considering that the Dural are getting a Greek flavor...

What is your profession?

OT: I like this idea, but not sure how it would work code-wise... Itwould need to change its workrate but how would it know what the worker is building?

 

[Valkrionn]

Considering that the Dural are getting a Greek flavor...

What is your profession?

OT: I like this idea, but not sure how it would work code-wise... Itwould need to change its workrate but how would it know what the worker is building?

Not a bad idea, but more likely to go to the Grigori. We've decided to make them good with improvements, Dural good with buildings; Differentiates them.

As for how to code it... Simple. PromotionBuilds. Already in the test version, and have been for some time.

Simply have the promotion remove the original build, and grant a new, better one, that completes faster. No need to mess with workrate at all.

 

[Randomness]

Hmm... it would be interesting if more profesions could be gained through experience.

 

[Valkrionn]

Hmm... it would be interesting if more profesions could be gained through experience.

They will be. Only one implemented for now is "Fortification Specialist", requiring Hardy3... Slightly faster workrate, allows you to build a Castle rather than Fort, and blocks construction of a farm.

 

[hbar]

They will be. Only one implemented for now is "Fortification Specialist", requiring Hardy3... Slightly faster workrate, allows you to build a Castle rather than Fort, and blocks construction of a farm.

If you have a stack of workers, and one has fortification specialist (or any of the presumeably many specialist promotions waiting in the wings) and you want to build a farm, will the fort guy just stand back and watch or will the farm button be greyed out?

 

[Valkrionn]

Haven't tested this. Should just have the one guy stand back and watch, though.