A Better AI.

[Jaybe]

Originally posted by Tattila the Hun
I wonder if something like this could be fixed...
(pic)

Those ruskies have been sitting there for couple of turns, at war with Isabella, bombarded the defences to zero, and attacked with cannons for collateral. And the infantry, they do nothing. There are six Longbowmen in that city, there were seven in the last turn. There are ten or so russian inf in that stack, and about similar amount of egyptians, vassals of russia.

Not that I'm complaining, that city made a nice addition to my small empire.

They obviously didn't WANT to take the city, at least not then. Maybe they wanted to starve it down, first?

[DrSpike]

Originally posted by MikeH


They did release the SDK including all the AI code for exactly this reason, that expert players could improve the game once they've spent time mastering the finished product. So Firaxis as well as Blake

Yes, I accept this argument. It was courageous of them to release these tools.

[DrSpike]

Originally posted by Blake
I just love how everyone has a unique opinion on the start point stuff, makes decision making so easy .

Hehe, you can't expect everyone to agree. You should see the fact there is a discussion as a testament to the value people place on improvements.

Ultimately it's your baby and your choice - if there is a lack of community consensus on something do what you think is best.

[DrSpike]

Originally posted by DaveV

Like the pop-rush bug in vanilla? Blake's dll is a great addition to the vanilla game, and big thanks to him for still supporting it (even though Firaxis isn't). But I'd agree that this mod shouldn't add new features, as opposed to bug fixes. The distinction between the two depends greatly on your viewpoint, of course.

This is spot on - there is a very strong case for bug fixes like this being included (in fact I asked about it specifically above, and presume from the quoted post it is in already).

To muddy things slightly, I'd also argue there is a pretty strong case for applying the 2.08 CS change to 1.61 vanilla.

And whilst I accept the arguments against introducing major gameplay changes the map change (like synchronisation and bugfixes) doesn't really fall into that category and should be evaluated on its own merits.

[Tattila the Hun]

I wonder if something could be done fix starts like this...



Augustus did quite well, despite.

[Adagio]

Originally posted by Tattila the Hun
I wonder if something could be done fix starts like this...

Reminds me of my first game on Monarch, except that I had room for 3 cities, but with 50% of the land being dessert that turned out to be one hell of a challenge... I didn't win

[Harrier UK]

Originally posted by Mergle


Yes. I'd prefer a pure AI-only mod. But the point is, Blake's AUI changes broke the setup code (more), so he has to fix it. And fixing it to broken-as-originally-written seems silly.

I agree with Mergle.

As the setup code uses some of the AI code - it is a valid contender for change, if as Blake says, improving the AI part of the code has a detrimental effect on the setup part.

Analogy - You invent a charriot with square wheels and find the horse cannot pull it very well.

So you change the square wheels to round ones. Only now the horse finds it to easy to pull and runs amock - tipping the occupents on the ground.

What do you do - go back to the square wheels and live with the bad design.

No - you also adjust the horse part of the equation - by changing the steering/braking methods etc.

Its an act of balancing not a fundemental game play change.

So IMO go for it Blake.

[Blake]

Originally posted by Forwarn45
Just wanted to add another vote against changing non-buggy game mechanics. I'd prefer this mod be to improve the AI and otherwise remain as close to vanilla civ as possible. I don't think the AI improvements have "broken" the game at all.

I don't think anyone, least of all me, has voiced support for changing non-buggy game mechanics . In fact even the buggy ones are mostly left alone.

Of course it's a fine line between bug and feature...

By the way, I posted this on the CFC thread but I think I forgot to say the same here, so please read it if you don't follow the CFC thread.
Or at least this paragraph:

I can make two promises: Firstly, this doesn't set a precedent. It is a special, and rare case where AI code is coupled with gameplay code, in earlier cases things got screwed up, I have then fixed that and taken the liberty of making some more improvements since I'm not about to return it to the lesser of two brokens.

This. Doesn't. Set. A. Precedent.

Originally posted by joncnunn

2.
On the imporoved AI city placement, I really like it, but the one major thing it's missing is considering if the new location is close enough to the rest of their empire to be defended. See the latest AU game where the Cyprus AI in many(most?) of them pouched an excelent city site too far from their empire (and too close to the humans) to defend, promptly resulting in the human taking the city from them. Maybe the alogrithm could go something like (on standard map; scale for different sizes) if another player's culture boundary is within 6 tiles of proposed city site and on this landmass and I don't have culture already on this landmass within 6 tiles of this proposed city site, then don't build here.

You'll be happy to know this changes in the next version.
As well as making a heavy penalty against founding in culturally oppressive areas (defined as "You don't have a ****ing chance of budging that culture"), I've also made a change to another aspect. The "AI city found value" will usually be a number between 1000-6000, but under some cases it ends up with a found value of 1, which means the site has no redeeming qualities but it is possible to put a city there. I have removed that so 1 is now 0 an the AI wont put a city there. We'll see how that goes. It will reduce the AI settler spam significantly.

I've also made changes so the AI founds it's second city usually 4 or 5 tiles from it's capital, a much more defensive placement and much closer to what humans do. This means that as a human you don't need to get the settler out quite so fast, but beware, it give the AI a significantly stronger start since the city will be linked into the trade network much earlier and such.

Originally posted by Tattila the Hun
I wonder if something could be done fix starts like this...



Augustus did quite well, despite.

This case should be much rarer due to the start point changes I've made.

[Solver]

Blake, with all the military improvements you've made, do you think that your AI will be able to reasonably achieve domination, for aggressive personalities?

[Blake]

New Build! 12/8 on sourceforge

- automated build governor for human player cities totally reworked to be useful. It will never build units unless there is no other choice, including processes. (It will prefer building gold/research/culture to building a unit). It whips buildings when appropriate.
- Drama tech should be higher priority for cultural victory AIs
- stack attack setting should work for human player
- AI stacks should now correctly completely attack, no matter the user settings on stack attack
- rare religions will spread better under cultural victory
- AI research choices changed, including wonders that are already built are not counted in value of a tech, scaling with game speed better, better at picking cheap techs, and more
- big changes to working plots: plots that are valued on their overall usefullness
- AI will turn off science when in the late stages of cultural victory
- AI will grow its cities properly when under cultural victory
- city governor no longer takes into account any multipliers (production/commerce), some other changes, significant change
- starting points changed to fix problems with city founding, including mad seafood starts, starts too close together, and more
- bug in 2.08 pop rush code where a building does not finish the next turn, but requires 2 more turns
- BetterAI tag now has the build date in it
- chipotle war info now displays even if you have not met a civ (ctrl-z case)

The AI should be a whole lot more competent at cultural victories with this one.
Hopefully there will be lots of feedback on the startpoint change stuff.

[Nacht]

Does the AI ever run a labor civic other than Slavery?

I just concluded a spacerace victory on Prince and everybody, including peaceful natures like Gandhi, Hatshepsut and Roosevelt were still running Slavery.

Being behind in tech I skipped Democracy. I expected to be forced to research it later because of unhappiness problems, that never happened.

I temporarily replaced the leadername popup code to show how they were valuing those civics:



This is late game, when all the tech leaders are buzzy building spaceship parts. The smaller civs seem to value emancipion somewhat higher compared to slavery but still prefer slavery, the sole exception being the Barbarians .

Somewhat later we get to vote on the issue and to my suprise:



Political correctness when voting in the UN, so lets look at what they really think:



Almost unchanged. While now changing to slavery would mean massive unhappiness.
So I rigged the next round of voting forcing everybody to vote no and next turn:





When they still prefer slavery that late in the game while all the others are running emancipation, emancipation is taken out of the game for all practical purposes. This basicly removes a gameplay element, I always liked the fact that slavery became old fasioned and unsustainable in the later game.

[Kuciwalker]

Originally posted by Blake
The AI should be a whole lot more competent at cultural victories with this one.

I'm worried that many of those changes will make it too easy for the human to figure out which AI's are going for cultural victory early on.

[Blake]

I wrote some code so that Emancipation is massively devalued when using 100% culture setting for cultural victories, with Caste being massively increased in value. A bracket in the wrong place and Emancipation was being capped at 20/total instead of 20/city. Fixed, thanks.

[Nacht]

I get an unhandled exception in CvCityAI::AI_stealPlots().

            if (pLoopPlot != NULL)
            {
                pWorkingCity = static_cast(pLoopPlot->getWorkingCity());
                if (pWorkingCity != this)
                {
                    iOtherImportance = pWorkingCity->AI_getCityImportance(true, false); // exception!!
                    if (iImportance > iOtherImportance)
                    {
                        pLoopPlot->setWorkingCityOverride(this);
                    }
                }
            }

Changing the conditional to

if (pWorkingCity != NULL && pWorkingCity != this)

got rid of it.

[Blake]

Ugh how is it that a non-null tile within a city radius isn't getting worked by the city? I'd like to know, mostly out of sheer morbid curiosity. Got the save or know the circumstances?