Is it possible that something with the mod is not working and playing well with the changes that were made in 1.029? I saw another bug in the Diplomacy screen where it was not possible to Declare War and the time listed on the line was something like #value rather than a number of turns.

I've played a few games with 1.0.3 and it's by far the best I've ever seen this game play. Diplomacy was difficult at times, but certainly workable (I even had a few alliances that lasted the entire game). Yes, if I take off to a huge lead early on, diplomacy is a little scarce but I dont mind that at all....it makes you think about reckless expansion a bit more.



Now as far as the AI running itself into the ground? Yeah, it's pretty bad about that. I've seen some AIs be fine, but others are still doing what was reported about...building too many crappy ships. And yes, it becomes a cycle because once it has them, it cant upgrade them since there is no money. So the AI stagnates and falls by the wayside.



1.0.25 requires careful economic balance and I dont think the AIs are capable of it without a fair amount of cheating. That said, in 'stock' (ie, no mod), the AIs tend to perform a little better since they are less hostile and thus they trade a lot more which brings in LARGE amounts of money and science. With the mod, there is still trade, but it doesn't seem to be as prevalent.

Uncle_Joe wrote:
I've played a few games with 1.0.3 and it's by far the best I've ever seen this game play.

Thanks! That's really nice to hear.

Uncle_Joe wrote:


Diplomacy was difficult at times, but certainly workable (I even had a few alliances that lasted the entire game). Yes, if I take off to a huge lead early on, diplomacy is a little scarce but I dont mind that at all....it makes you think about reckless expansion a bit more.

Good news here: The devs are thinking about implementing some more conditions to play around with, when it comes to Diplomatic-Agreements (which include war).

This should allow modders to implement even more sophisticated and less Attitude-dependant behaviour of the AI, when it comes to war-declarations.

I will certainly use it once available to realize some of my plans for it that currently cannot be realized.

Uncle_Joe wrote:


1.0.25 requires careful economic balance and I dont think the AIs are capable of it without a fair amount of cheating.

Why do you think so? The AI currently is not capable of doing it, yes, but if you can balance your economy as a player, the AI could do the same by just following the same rules that the player does.

It's 3 simple rules, that, if followed would prevent it from happening:

1st: If income is negative don't build more ships.

2nd: Never increase tax to a point where your people would be in the state of rebellion.

3rd: If your tax is at the highest non-rebellion-level and you still have a negative income build Ind=>Dust.

Well, I think the econ system is tight enough now (at least in the early game) that it requires a lot of more subjective judgement calls. For example, I can sometimes skate by without much of a fleet in certain situations which allows me to invest my money in developing me colonies faster for a quick burst of economic growth. But knowing when I can get away with that is not an easy think to recognize.



Another example would be research. Sometimes I can bypass a lot of military research for a while, but again, it's not always a good idea in many situations. But I can estimate how much time I have before the wheels come off pretty well by now. But again, it takes some judgement calls and some risk that wouldnt be smart in many situations.



Overall, all of these things are GREAT for gameplay, but they are difficult for the AI to be able to 'sense'. So my guess is that it simply be easier to have the AIs get a little more in the way of resources (money, science, whatever) to compensate for the lack of situations judgement ability.

Well, that's what higher difficulty-levels exist for. ^^

And higher difficulty basically add 'cheats' which is what I trying to say above. I believe that even on 'Normal' level, the AI will require some (at least minor) degree of cheating in order to actually be competitive. And hopefully the higher levels will add additional cheats gradually, not in huge chunks.



To me there is nothing worse in an AI than going from 'it has no chance' to 'it cheats so badly that it feels like you arent even playing the same game anymore'.

I think one of the long-term goals is to improve the AI so much that it needs cheats less and less and that higher difficulty-levels only increase their boni in small steps.

Hmm, AI building tons of useless units and running itself to the ground unless you give it a huge bonus... Where have I seen that before? Oh, yes, in Alpha Centauri! And Sword of the Stars 1! And Civilization 4! Well, now that I think of it, almost all 4X games seem to have this problem...

Did the new patch do anything to help the bankruptcy situation from a modding point of view?

I will check it out once I'm at home. I'm missing quite a bunch of things in the patch-notes that I had expected to be done. But especially for XML-changes we often had the case, that those things were not mentioned in the patch-notes.



I'm still hoping that bankruptcy-issue is resolved without even the need of modding! ^^



Edit: I just saw there's a huge list of XML-changes posted in this thread. That makes me very confident, that I can now finally implement the smart-diplomatic-behaviour, I've wanted to.

I uploaded Version 1.1.0.



All-AIL-Mods-Mod.zip



Unfortunately there's no new options to configure the diplomatic-behaviour yet. So, now I'm off to try if the patch solved the bankrupcy/rebellion issue.

Well it looks like it still manages to run itself into the ground. Granted the game I'm playing, the Sowers only had 4 systems and they were cut off. But they were bankrupt (and then they DoW'ed me...). They had 2 tiny fleets that were outdated and their tech was waaaaaay behind.



Other than that, I've been pretty much in the lead the whole game. The AIs are all content to let me be in peace again and they are all trying to ally with me. And it looks like it will be easy to carve them all up one at a time again.



I'm not really sure what the solution is though, but I figured I'd report the AI bankruptcy again (1.1 version of the Mod and Normal level).

That's sad to hear. Especially since Meedoc claimed he had run tests for a week and never had seen it happen. I've not gotten far enough in my game yet to see it and how my changes work out.

I'll keep a close look on this stuff.

I'll test it too, see if it's better now.

Well. Can confirm that the Bankrupcy/Rebellion-Issue is still around. :\

Ail wrote:
Have I forgotten to upload the version where the AI would try to not get involved into more than one war each (unless they really hate someone)?

No you did upload it, when I tested the mod, I played with 1.0.2. In my game both factions (Sowers and Amoeba) were already at war, 1 war for Amoeba and 2 for the Sowers, besides they were both far away from me. But yes they both hated me so maybe that's why they declared war.

I removed everything related to the bankruptcy/rebellion-issue from the mod because I thought it would work now.

But you may be right about the diplomacy-issue.

There have been changes to waiting-times for some things.



I've also noticed that there was unexpectedly much peace in my game. My first thought was it has to do with me removing score-tension but there might also have been another issue preventing the AI from going to war.

Test game: investigate the AI.



Huge ovoid galaxy, all parameters to default except wormholes (high) and constellations (many), 8 players (each faction represented), normal difficulty. I play as UE and I don't use any mods or cheats or anything.



After 25 turns:



- United Empire: 4 systems, 19 population. I'm in a decent constellation but I'm already facing a lot of pirates chasing my colony ships. I encountered one neighbour (Sophons). As I have enough room and good systems to expand, I won't declare war to anyone during early game. 0 ship (I don't count colonization ships, if any).



- Amoeba: 3 systems, 14 population. The only 2 other systems he has in his constellation are full of gas giants and barren planets. He has 59% approval with tax rate at 20%. He already has a negative dust income. 6 ships built so far. I think things will go from bad to worse for that dude.



- Pilgrims: 2 systems, 14 population, 3 ships. He has a fair constellation and is doing fine. But why is he building ships before anything else? And I mean military ships, not colonization ships or scouts. He has 59% approval with tax rate at 40%.



- Hissho: 5 systems, 19 population, 5 ships. 60% approval with tax rate at 30%. He's expanding quickly and he's doing fine. He has a very good starting constellation, which helps.



- Sophons: 3 systems, 15 population, 3 ships. 55% approval with tax rate at 25%. I wonder why he chose to colonize 2 systems which are clearly not that good to start with. He has plenty more valuable systems at his disposal.



- Cravers: 5 systems, 19 population, 7 ships. 61% approval with tax rate at 30%. Just like Hissho, he's expanding quickly and he's doing fine.



- Horatio: 4 systems, 17 population, 4 ships. 61% approval with tax rate at 20%. He's isolated in a corner of the galaxy but he has a lot of valuable systems to colonize.



- Sowers: 7 systems, 20 population, 8 ships. 65% approval with tax rate at 30%. Just like the Horatio, he's isolated in a corner of the galaxy. I don't know why he built those ships, but his dust income is already negative.



That's it for now. I'll check again at turn 50 to see how the AI is performing.



Edit: forgot the save, here it is

[ATTACH]4480[/ATTACH]

After 50 turns:



- United Empire: 8 systems, 49 population, 9 cp. Everything's going fine, I'm building Magnetic Generators on my systems. I taught the pirates and the Sophons a lesson about space battles and I have more fids than everyone else.



- Amoeba: 6 systems, 34 population, 3 cp. 51% approval with tax rate at 15%. He tried to colonize the systems in the center, which was a good idea. Unfortunately for him, the pirates and the Hissho destroyed all his ships and now the Hissho are invading his systems in the center. He is far behind everyone and will be crushed sooner or later.



- Pilgrims: 4 systems, 37 population, 18 cp. 56% approval with tax rate at 35%. Despite a very bad start, he has some chance to come back in the game, if no one comes to claim his systems of course. The AI is doing well with the Pilgrims considering its bad start.



- Hissho: 6 systems, 56 population, 28 cp. 60% approval with tax rate at 15%. The major flaw of the AI appears first with the Hissho: even with 15% tax rate, and all his colonies with ind-->dust, he has a negative dust income and not so much approval. I bet his economy will soon stop working.



- Sophons: 8 systems, 54 population, 24 cp. 56% approval with tax rate at 20%. His major colonies are either on strike or unhappy. Just like the Hissho, he will soon stall and besides he keeps building ships. But one thing is interesting: he's currently building Colonial Rights on the colonies which have the lowest approval. Maybe there is hope for him.



- Cravers: 6 systems, 60 population, 21 cp. 48% approval with tax rate at 20%. He doesn't have a lot of approval and has a negative dust income. Anyway, he seems to think that the solution is to build more ships... and bigger ones!



- Horatio: 6 systems, 62 population, 32 cp. 49% approval with tax rate at 30%. All his major colonies are unhappy and he has a negative dust income. If the tax rate is set at 40%, all his colonies are on strike.



- Sowers: 9 systems, 48 population, 22 cp. 20% approval with tax rate at 40%. He has 0 dust and +2 income. All his colonies (not the outposts) are on strike, the rest is unhappy. He is on the edge of a rebellion state and 5 systems out of 9 are building ships.



I guess it demonstrates what we already knew, but I'll do another report at turn 100. Not all is bad however, I think the improvements the AI builds are correct except maybe stealth constructions which I don't find useful. I just don't see the point to build so many ships. Except the Cravers, no one actually did a lot of military research so they'll all end with tons of useless weak ships in 20 turns.



Save: UNITED EMPIRE - Turn 50.zip

I am curious how you can tell the relationship between taxes and happiness for an AI. Are you just jailbreaking, loading the position as a human player, and looking at the system happiness values? This is not correct. The position, when played by an AI, has a significant approval bonus depending on difficulty level. If you are not adding that in, then your report that all these AI colonies are on strike seems incorrect.