Civ 6 AI Battle Royal

So from some casual observations of the steam, and based on the comments in here, and knowing pretty well how the AI behaves in Civ 5 (hundreds of simulated matches). I think I can be pretty certain that Civ 6 simply uses the same mechanisms as the Civ 5 AI. Just updated slightly to include the new mechanics. It may possibly have better tactical AI, but we couldn't really tell.

If so the statement the AI isn't declaring war to prevent the warmonger penalty is most likely inaccurate. In civ 5, the AI doesn't actually look at the value of the warmonger penalty, and so the AI probably doesn't here either. It may be kind of indirectly true, as the programmer may have set the desire for peace rather high, because he thinks they should care about the penalty. But it doesn't have to be directly connected to the penalty.

I have rather mixed feelings overall, it's unfortunate that the AI still appears so weak, but assuming we have at least the same moddability of XML files, modders should be able to make big leaps. Some stuff can't really be changed without the code of course, I don't imagine the upgrading issue can be solved through xml, nor the mutual declaration of war resulting in no attacks, nor some other problems. But it should be possible to make them chase victory conditions a little more, make them more aggressive/opportunistic in warfare. better buildorders, more thought out unit compositions etc.
 
Just remember that early in Civ V live, I think the devs actually wants us to send them save files to analyze and tune the AIs
 
ok, did they mention anything about the version they used on this Battle Royal? I mean is there reason to believe they have a better AI ready tonight, e.g. talking about adjustments being made for day1 patch?
Haven't watched it.
And will be happy anyways, because Civ AI is what Civ AI is... :p
 
Every video and stream, except for the one used at Beachgate, will not be the most recent build. The risk of crashes or other unnoticed things (like what caused Beachgate) is not worth it.

Most of the AI errors the programmer pointed out were likely things that existed 2 weeks ago when he was working on the build...which also happens to be the build used for the stream. Some will be gone, I'm sure.
 
I think there's a lot of over reaction to todays video regarding the AI. What we basically saw was AI Civs reacting to other AI Civs with no random disruption from a human player.
The #1 thing I think most of the sky is falling crowd are failing to remember is that Civ is not a game designed to be played as AI vs AI only.

I think it's totally OK to be worried if even the lead AI developer say's that he does not like that fact - only two days before release.

All those AI problems we could see in the stream were minor ones but them not upgrading their units is such a major one! It's not only a major problem but a huge letdown because it totally destroys my fantasy. Warriors, archers and catapults in atomic era? No way I could have fun when seeing armies of those that late in my game. The AI not upgrading their units is something that I've seen in multiple Let's Plays so this is nothing purely related to AI vs AI. And it has nothing to do with missing resources, there were many catapults and archers in the endgame, too. It's way more likely that this is related to the maintenance costs.

Maybe this is just some number tweaking (and chances are good that this is the case)... However I think this is not an overreaction, just a mixture of disappointment and a reminder of the Civ V release.
 
It's way more likely that this is related to the maintenance costs.
In that case it should not be a problem on higher difficulties where the AI gets a lot of bonuses and might have enough gold for unit maintenance. Maybe the difference between King and Prince is not that big when it comes to yield bonuses (but free Amenities/housing instead? I haven't looked close enough, but it seemed strange for the AI too have some rather large cities without many neighborhoods and too many luxuries) I can also imagine that on Deity, or whatever difficulty, the AI does not need strategy resources any more.
 
I'm not a programmer, though had some hobby coding long time ago.. really how hard could it be to put it into the scripts:
-build units if you are under a power treshhold (the ai always should have strong military even on the expense of economy.. that is what needs to be compensated with more ai bonuses on higher levels)
- don't build builders if you already have 3-4 unused
-don't leave good/moderate land unsettled if you already has a settler
-don't settle on other continent when you have good land next to you.
-repair pillaged tiles as soon as possible (might not always be optimal, but still better than not doing it)
-focus on a victory condition and don't lose that focus easily. (I got bored of the stream but i heard england could had won, but just haven't started to work on the final space project part even if they could had.)
 
In that case it should not be a problem on higher difficulties where the AI gets a lot of bonuses
That's not what I ment. They may have enough gold but they seem to think that it is better to have a weak unit that costs nearly nothing (or nothing, I don't know) than a good unit that costs a lot of gold.
What I think is that the "military weight" of those old units is too high.
 
That's not what I ment. They may have enough gold but they seem to think that it is better to have a weak unit that costs nearly nothing (or nothing, I don't know) than a good unit that costs a lot of gold.
What I think is that the "military weight" of those old units is too high.

But those are things that can be easily tweaked (I assume). Until then, there's tons of stuff to learn and I'm sure it's going to be worth my time.
I just don't get why people question everything now, just because one AI battle was too peaceful and the AI didn't know how to circumvent resource limitations in unit upgrade paths (which is definitely a design flaw ... gaps between units are way too big AND severely limited by resources).
Anyway, I'm 100% sure this won't ruin my fun ... it can limit longevity of the game, yes, but I believe patches can help a lot until then.
 
Yes, for what I've seen and read, the problems are not really a result of a bad AI but more the result of designs decisions.

For example if the AI doesn't declare war because the warmonger penalty is too high, well that's the sign of a smart AI not a bad one, and if that's true, we just need change the warmonger penalty.

In a game as complex as Civilization, you have to design the gameplay with the AI in mind as you can't expect it to be smart and fast if you keep adding tons of specific rules that are here to provide fun to a human, and the real difficulty is to balance the fun/challenge ratio.

Ai not upgrading units is a result of bad game design no strategic resources to upgrade. (no niter)
 
How comes I can't find this video on YouTube? (searching for something like Civilization VI ‘AI Battle Royale’ simply doesn't work)

Does anybody have a link?
 
Thanks a lot! (Unfortunately I may not modify this computer and get on Twitch: "No supported video backend available: Flash is not installed")
On YouTube watching the other Civ videos worked fine .

I followed already the link given in the first post of this thread ('Civilization YouTube channel' == 2kgam.es/CivilizationYT), can see there all (older), but the wanted Civ video. Also a general search doesn't yield a result ...

Is there a waiting time before it becomes available on YouTube?
 
Am at work so won't have time to watch. Questions to those who did:
1. How did it end? (Aztec Cultural victory?)
2. Did anyone get capped? (Unlikely)
3. Did the Aztecs return any capture Russian cities (again)?

(I turned in before it finished as I still have hopes for a few turns this evening so "pre-loaded" some sleep last night...).
 
I wonder why they played this AI Battle Royale "live" - I would have made a few runs of this AI vs. AI beforehand and choose the best that suits needs...
 
But those are things that can be easily tweaked (I assume).
That's exactly what I said.

Chances are good that they can fix this before release, but I could be wrong and they need a few days or weeks. In that case this weird behavior would totally destroy my fantasy.

Of course patches will fix such bugs and as time goes by the game will get better and better. But why do they always plan their exact release date month and month ahead? It's impossible to foresee such a long time period of development. It would have been better to release the game after those bugs are fixed. I don't like this "early access"-mentality of many game publishers (I don't mean the developers themselves). That's why I am a bit upset.

I'm still having hope and I will nevertheless start playing on Friday!

Edit: "Upset" is a bit to strong, I don't know any other word to describe my feelings :crazyeye:
 
Last edited:
Top Bottom