RHQ Artificially Intelligent AI Mod RHQ AI Mod 2.10

[notque]

I'm getting close to what feels like a release of the next version. In the next 10-15 minutes I will run more autoplays on twitch.

If they go well, I will release v 1.0.8

 

[Sytharin]

Now that I'm later into the antiquity age, I am seeing the exact same thingin my game. Everyone is 2 over the cap which is great but Catherine is **4** over the cap with zero happiness and unhappy/angry towns ...

This time her gamble did pay off, she turned it around big time in exploration age and is scaling bigly, she has a ~100 lead in culture and science, and her happiness problems are a thing of the past. She wasn't so lucky the last 2 tries at this XD

 

[bumpyglint]

Thanks a lot again for your incredible work; I'll try the mod as soon as I'll finish my actual game (in order to not change it midway), unluckily I need to work more than expected this week but I hope I'll be able to try it soon!!

 

[Leyrann]

I'm getting close to what feels like a release of the next version. In the next 10-15 minutes I will run more autoplays on twitch.

If they go well, I will release v 1.0.8

Do you have an estimate for when you'll be able to release it?

I wanted to start a new game but I'm not sure whether I want to wait for the new version or play with the old one.

If you don't answer within an hour or two (I imagine you went to bed sometime shortly after the message I quoted) I'll just assume I'm better off downloading the old version for my next game.

 

[notque]

Do you have an estimate for when you'll be able to release it?

I wanted to start a new game but I'm not sure whether I want to wait for the new version or play with the old one.

If you don't answer within an hour or two (I imagine you went to bed sometime shortly after the message I quoted) I'll just assume I'm better off downloading the old version for my next game.

Sorry, no.

I am not releasing it until I fix the City Attack tree. If I fix it today, it will be released today. If I fix it in a week, it will be released in a week.

 

[notque]

This time her gamble did pay off, she turned it around big time in exploration age and is scaling bigly, she has a ~100 lead in culture and science, and her happiness problems are a thing of the past. She wasn't so lucky the last 2 tries at this XD

This is what I often see. The AI that go over the cap end up leading the game.

I understand it looks wrong, especially with the way Civilization 7 tells you to play. I am only concerned about putting the AI in the best chances to succeed within the rules.

All evidence I've seen from in game action is, going aggressively over the cap is better for the AI.

Even having every city revolting and unhappy, is a net benefit. I know it doesn't make sense. The evidence all says the same thing. I see no reason to change it.

 

[notque]

I'm getting close to what feels like a release of the next version. In the next 10-15 minutes I will run more autoplays on twitch.

If they go well, I will release v 1.0.8

Update, it did not go well. I am not releasing the next version until the City Attack Tree is fixed. If that's today, it will release today. If that's in a week, it will release in a week.

 

[notque]

To Address a topic I've been asked many times - Ongoing support for the mod.

Once I get the hard stuff done (behavior trees), I am handing the mod off to Roman Holiday.

I am only here as long as it takes to get the basic, core elements working and function. Once that's good, I will give the goahead for Roman to take it, and release it. He managed the Civ 6 AI mod, and did a great job.

And then I will come back only when there is new DLC, or updates that require me. I did the same thing in Civ 6.

So when that hand off happens, I will make the post, Roman will release his version of the mod, and I will go back to hibernation.

 

[Morgawr]

This is what I often see. The AI that go over the cap end up leading the game.

I understand it looks wrong, especially with the way Civilization 7 tells you to play. I am only concerned about putting the AI in the best chances to succeed within the rules.

All evidence I've seen from in game action is, going aggressively over the cap is better for the AI.

Even having every city revolting and unhappy, is a net benefit. I know it doesn't make sense. The evidence all says the same thing. I see no reason to change it.

Yes I agree, 1-2 over the cap seems to be optimal for long term and 3-4 in short term (while waiting for settlement cap bonuses) and >4 very occasional short term (eg captured city). For antiquity at least.

IMO as long as they don’t completely IGNORE it (which I DONT think they’re doing) and go 6+ over and have -30 in all towns (= angry = pillage every turn = lose gold / prod every turn), it is all good. That’s why the devs call it a “soft” cap.

 

[Sytharin]

This is what I often see. The AI that go over the cap end up leading the game.

I understand it looks wrong, especially with the way Civilization 7 tells you to play. I am only concerned about putting the AI in the best chances to succeed within the rules.

All evidence I've seen from in game action is, going aggressively over the cap is better for the AI.

Even having every city revolting and unhappy, is a net benefit. I know it doesn't make sense. The evidence all says the same thing. I see no reason to change it.

I'm definitely impressed so far, AI making gambits is highly enjoyable

 

[Lebo44]

@notque will you be streaming Today? I found these autoplays fun to watch, especially when you see a progress in AI behaviour

 

[Squibie]

haha yeah, its fun to hang out

 

[notque]

@notque will you be streaming Today? I found these autoplays fun to watch, especially when you see a progress in AI behaviour

Potentially, although I really want to talk through the problem with other people. I was thinking about doing it on the AI Modders discord if I can find anyone who will do it with me.

It's a tough challenge I'm working on. People underestimate how hard it is to work within an AI system with no documentation, that I have to reverse engineer out of guesses, and checking if those guesses react appropriately.

 

[notque]

Everything else outside of the City Attack tree I have working better. I must get the City Attack tree working. That is the major revelation that I require to release the next version.

Because that will be a wow moment for people. And I would like my AI children to defeat these cheating humans who use momentos my ai children don't have access to.

 

[Squibie]

Ye not picking mementos was my first choice for dealing with this lousy AI. I think its really meant for multiplayer so people have more diverse options. But maybe with your mod I will need to use them haha.

 

[Morgawr]

Everything else outside of the City Attack tree I have working better. I must get the City Attack tree working. That is the major revelation that I require to release the next version.

Because that will be a wow moment for people. And I would like my AI children to defeat these cheating humans who use momentos my ai children don't have access to.

I don’t use mementos since it is in direct contradiction with making AI better at Deity by giving bonuses lol

Your current version AI already ruined my attack plans and I heavily underestimated its defensive capabilities… I am starting a new game now with that in mind and can’t wait to see how you’ll make it even better with the few resources you have!

 

[Leyrann]

Potentially, although I really want to talk through the problem with other people. I was thinking about doing it on the AI Modders discord if I can find anyone who will do it with me.

It's a tough challenge I'm working on. People underestimate how hard it is to work within an AI system with no documentation, that I have to reverse engineer out of guesses, and checking if those guesses react appropriately.

You've been doing a great job so far. I'm not even 100 turns in but the AI is clearly settling significantly more.

And I would like my AI children to defeat these cheating humans who use momentos my ai children don't have access to.

If you can make Deity too hard for me to win, I'll start using them.

(also, I see this mistake everywhere, but it's mementos, not momentos - related to 'memory')

 

[Squibie]

I mean their whole purpose is to give some momentum.

 

[ggalindo001]

Everything else outside of the City Attack tree I have working better. I must get the City Attack tree working. That is the major revelation that I require to release the next version.

Because that will be a wow moment for people. And I would like my AI children to defeat these cheating humans who use momentos my ai children don't have access to.

This is good to see.

I think you have hit the 2 of the 3 core issues that I have with the AI -- Settling, City Attack/Defend/etc. The one that is still frustrating is the AI not declaring war when they are furious (or whatever the final state of bad is) and you have a city or two ripe for the taking. In my current game, Frederick of Prussia had 3 cities virtually surrounding an enclave type city I had, he has been furious with me for 2 ages, I have had one unit stationed in the city, and nothing. It just makes it too easy when you don't have to actively defend against a threat.

 

[Morgawr]

BTW does AI know that embarked strength in Antiquity is only 5? I killed some that way when they were crossing the navigable river to come at me.. didn’t see them too do it much tho, so it’s probably ok

(I have a phobia from early Civ 6 days where the AI would jump all its units the sea trying to path and you could shoot all like sitting ducks)