Prolific is quite amazing.  Combine that with +2 pop upon colonizing and EVERY colony begins with 6.5 pop and 3.25 Raw production. 

A double-module colony ship will start with 11 pop.  

And this is WITHOUT colonizer's +100% social production.  

Prolific + Wealthy is good.  

Nah. 1%/level is staying. For the adjacency-dependent hub buildings, I'll probably just increase the level bonuses they provide - so you get like +5% from putting that thing next to a factory, but building two factories next to each other still just gives each +1%. Granular control is key.

The Malevolent bonus doesn't effect the flat social production bonus. Like, at all.

The sum is (((((production+flat)*(1+prod multiplier))+flat manu bonus)*(1+manu %))+flat social manu bonus)*(1+social production multiplier), so literally the only thing that multiplies with the 5SP from colony capitals is % social manufacturing bonuses - of which there's very, very few, just a couple of techs and the presently broken colonizer trait.

I've been keeping an eye on Prolific for a while now tbh. I don't *think* it's unbalanced - that 2 pop amounts to just +1 production per new planet, which is less effective than the colonizer trait for building up a planet quickly, plus it usually tips you into <100% approval and so you don't even get the full +1 initially - but it's certainly one of the better abilities at this point.

Got to say, I love some of the work done here!

Just one note as someone he did a lot of work on the Civ 5 Fan Balance Mods....be careful not to change too much too quickly. It can get very tempting to want to fix every little thing, but sometimes the scalpel approach is the best. Small changes can have big consequences, and it can take a lot of games to see the full consequence.

I've done most of the big corrective stuff I was looking to do at this stage tbh; I think the next couple of patches will be general bug fixing and bringing the AI up to date with the changes that have come in since 0.7  or so.

Harmony crystals.  

100% approval on every planet.  

That requires you to have harmony crystals, which isn't guaranteed. Besides, it requires a full colony ship to get the most benefit... which requires taking population from where they are likely to be much more productive. 

But as I say, I'm a bit concerned with any of the things which give you free population from any source, really. I'm not convinced their OP, but I'm not convinced they aren't yet either. The flat 2 pop is more concerning, since you can send out 0.1 pop and get 2.1 at the other end.

Still generating 50 colonies by turn 50.  At least it's not 100 like before.  Paying for it with Krynn temples and trading for money/turn from the AI.  

Ohohohohhohhohohohhohohoh.

OHOHOHOHOHOHOHOHOHOHOHOHOHOHHO.       

Colonizer is quite amazing, especially in combo with Prolific. Main advantage of colonizer? It allows me to QUICKLY build up my research and economy worlds (literally twice as fast) which generates the income and research for FURTHER expansion.  

Turn 60, 60 colonies.  NO DEFICIT. LET ME REPEAT THAT: NO DEFICIT.  9000 CREDITS IN THE BANK FROM TRADING WITH THE AI.  Colonizer +100% social production allows very, very fast production of Krynn temples and Colonial Hospitals, followed by marketplaces or research labs, which magnifies the advantage of extra population from Prolific.  

I have FOUR TIMES THE POPULATION OF THE NEAREST AI.  MORE THAN HALF OF MY WORLDS ARE FULLY DEVELOPED WITH LEVEL 1 BUILDINGS.  +200% research or economy on those worlds, with average pop of 15, and the POP GROWING AT +0.4/TURN.  AHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!!!!!

THANK YOU FOR MAKING COLONIZER EVEN BETTER THAN BEFORE.  I CAN ALWAYS GET SHIPYARDS BY SHIPPING OVER A CONSTRUCTOR AFTER THE COLONY SHIP.  

AHAHAHAHAHAHAHAHAHAHAHAHA!!!!!!           

Once again, is it possible to program the AI to expand to THIRTY colonies ASAP before stopping? The OPTIMAL strategy even for normal AI is to grab THIRTY colonies ASAP before building up the worlds.  

Quoting marigoldran, reply 208

Ohohohohhohhohohohhohohoh.

OHOHOHOHOHOHOHOHOHOHOHOHOHOHHO.       

Colonizer is quite amazing, especially in combo with Prolific. Main advantage of colonizer? It allows me to QUICKLY build up my research and economy worlds (literally twice as fast) which generates the income and research for FURTHER expansion.  

Turn 60, 60 colonies.  NO DEFICIT. LET ME REPEAT THAT: NO DEFICIT.  9000 CREDITS IN THE BANK FROM TRADING WITH THE AI.  Colonizer +100% social production allows very, very fast production of Krynn temples and Colonial Hospitals, followed by marketplaces or research labs, which magnifies the advantage of extra population from Prolific.  

I have FOUR TIMES THE POPULATION OF THE NEAREST AI.  MORE THAN HALF OF MY WORLDS ARE FULLY DEVELOPED WITH LEVEL 1 BUILDINGS.  +200% research or economy on those worlds, with average pop of 15, and the POP GROWING AT +0.4/TURN.  AHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!!!!!

THANK YOU FOR MAKING COLONIZER EVEN BETTER THAN BEFORE.  I CAN ALWAYS GET SHIPYARDS BY SHIPPING OVER A CONSTRUCTOR AFTER THE COLONY SHIP.  

AHAHAHAHAHAHAHAHAHAHAHAHA!!!!!!           

End of marigoldran's quote

You do realise that you've basically been rehabilitated, right? You're building improvements, splitting your shipyard production between constructors and colony ships, setting population  to do things other than 100% military manufacturing, loading as much pop onto colony ships as possible... This is all pretty much what I was intending. The mechanics now actually favour playing the game in the way most people expect to be better, rather than the ludicrous mass of perverse incentives in vanilla.

Be warned that I'm gonna be nerfing the initial cash pile once the AI is ready for it. The default 3k allows you to rush-buy a lot more than it used to in the early game, and probably needs to be reduced to around 1k tops. Combined with the AI being given much more aggressive spending priorities (the colony capital being switched to social manu means it doesn't need to be so careful of bankrupting itself), this might mean a strategy based on selling stuff to the AI won't be as effective in the near future.

Nope, can't tell it to check the number of worlds it has, as I've told you over and over and over again. I will be pushing it to be less worried about social production now, tho, since the colony capital change means it probably never needs to spend less than 100% on the military-social slider now. It'll only take 6 turns to build most improvements, compared to 3-4 with 100% social, so might as well let it pump out more colonies.

Besides, it's more important right now to teach it to research properly, since it presently is very bad at picking which worlds to make into research planets (it would make only 4-5 in a 40-50 planet empire). I've not got much hope for it being a proper challenge for anyone who's played many 4X games before until SD teach it to specialize, though; there's simply no way it can match the level of output from a player planet using just the big wheel.

Rehabilitated in what way? I only choose strategies that I believe is optimal.  If the environment changes, I will adapt with it.  

GC3 is practice for the markets.  

Yeah, starting money nerf is probably the next step.  That's a good idea. 

SD needs to create control groups for planets.  That's the best and simplest idea.  Will help with micro too.  

funny how GC3 is supposed to be supperior AI wise.

It actually is. That's the problem. The AI is actually playing the game, which is waaaaaaaaay more complex than most of the competition (Total War, for example, just hands out goodies to the AI rather than have it playing by the same rules as the player). The problem is, until it gets good at playing, the net result is weaker than if it just cheated wildly.

It's worse because so much of it is presently scripted rather than reactive. Writing the strategic AI presently involves sitting down and trying to put together a plan for the next 300 turns which will not change regardless of anything that happens between the start of the game and the end. This is actually generally OK for the first hundred turns or so... but after that, the AI rapidly falls apart because it has outside forces that it needs to react to, and yet we can't script much in the way of reactions. There's literally nothing to check an outside power in the strategic AI Enums... or even a way to check the AI's own strength.

Still, I admit I did laugh during one of the dev streams when Paul said 'Brad's the best GC3 player in the office', and then said 'if you're playing against the AI, you're basically playing against Brad'.

Does Brad know the constructor to colony ship trick? Because if I played Brad in GC3 I would crush him.  

Point may have already found but the Dur Armor for ships (one after titanium) is at 18 for mass vice 1.8 ... not sure if intentional or not.

So far more challenging then base game but still the issue of expansion for the AI player.  Created some of my own races to play again and no idea why but they seem to expand more and become a greater threat then the vanilla ones in the game.  They do seem to slow down in the age of war though for some reason.

thanx for the mod ....

Yes, in fact he is interested in teaching the AI to do this as well. It's not an issue of the developers being idiots; it's an issue of being able to script in specific tactics (like this one).

In 1v1 multiplayer, with smaller maps the strats pretty obvious: colonize ASAP to get Malevolence Lvl 5 free Overlord ships on every planet.  Ideology route is pretty simple: Pragmatic (lvl 1 3 Constructors) followed by Malevolence invasion route.  

Then ball the dozen or so overlord ships into a fleet and crush anything your opponent has.  

The only other option is to get Pragmatic No War 50 Turns, so that you have time to colonize and get Malevolence Lvl 5 free Overlord ships on every planet. 

There ARE no other viable strategies.  

Hmmm. Annoyingly, the game determines credits by faction. So it's not a single setting which impacts all races, but is rather determined during race creation. So pre-existing custom races (or third-party ones downloaded from the workshop) will have however many credits they were granted when first created - usually 3k. New races will be given whatever I tell them... but other stuff will need to be re-created. That's bloody irritating.

Also, do try to remember that the AI in GC3 is a reflection of how the guy who wrote it would play if he had to make all his moves without knowing anything that was happening in that match. He needs to try and predict what he ought to be doing on turn 341 without any idea how many planets he has, what the diplomatic situation is, who he's next to, or even where he is in the galaxy. It's a bit like playing by shouting instructions to someone in a different room, who can only understand about 300 words of the same language as you. I'm willing to warrant that anyone on this forum would struggle to beat Brad if you both had to set out exactly how you're going to play from start to finish before you even launch the game.

I'd also point out that Marigoldran's own stab at an AI outline last month was so obviously exploitable in so many ways that, at the time, I genuinely thought he was joking. Being able to play does not translate into being able to design AI to do so, because so many actions that the player automatically takes need to be specified for AI.

Doesn't that seem backward though? It's 2015 when will we have reactive AI that takes into account planets, relations, metrics, position and include it own state memory and etc? I mean as a human player after the opening game I only play reactively, following feedback rules I use, At this point all AI empires behave like roaming zombies.

Is that what the golden standard of 4x should be?

I never said my solution was good.  My point was that my awful solution was BETTER than what's in the base game right now.  

I mean we're moving farther away from GC2:DA...

The best strategic AI is to have no AI at all, if you take into account Sorcerer King.

At this point I would like to have an Overlord empire ala Arcen's AI WAR and call it a day if the devs can't be bothered anymore.

GC2 DA didn't have very good AI either.  

It knew how to play the game optimally. Granted you could game it like any other AI but you got your moneys worth for an opponent.

naselus:

maniakos:

Actually, you may be onto something...

What if the game designers in addition to their first revision of an "initial game AI" also offered a way for the gamers to play the game through a customized extendable AI. Not just as a modding tool but as a planned and built game option where players could pit their own AI's against each other. Sort of like those virtual "robot" battle competitions where player's code the algorithms and then see how they fare against each other. These AI battle royals could be made into regular Lobby competitions and the best AI's could be incorporated into custom factions that would then be available in your standard game as a download. With enough interest the AI would get better and better as gamers tinkered with their own and battled it out, and the whole thing might be quite fun.

From SD you'd need the tools and game structure to support all this but it would allow them to gather the creativity of hundreds of hobbyists to make one heck of a game AI...

Yes.  

Much as I'm enjoying this discussion, could we maybe not have it in this thread, which I'd prefer to be used for feedback on the actual mod?