[Disclaimer: Stardock does not endorse or support this mod. This thread has merely been stickied so users can more easily direct comments or questions to the authors of this popular modification.]
Some time ago the modding community started poking around in the TA tech trees to see why the AI was reluctant to develop certain techs. We found that some weapons techs were categorised as PD, instead of Military, meaning that unless missile weapons were deployed, those techs would never be researched by the AI. Obviously, an AI that can't fight back isn't much of a challenge.
The idea behind this mod was originally just to get the TA AI on the right track as far as weapons are concerned. However as suggestions have been put forward by the community, more improvements have been made to the tech trees. Some features have been driven by my own need to fix annoyances.
For example there is the AI's passion for spamming factories or labs onto every tile, crashing their early economy. These basic improvements would get upgraded and then cost the AI even more in terms of miantenance and production cost, forcing the AI to run inefficiently with the production slider at far less than 100%.
So my approach has been to deprive human and AI alike of early structures that can cause economic meltdown, and in their place have Super Projects which allow for some single-planet focus.
N.B. Use the external link above or choose one of the versions below. The successive versions are not patches, they contain all the features of that version of the mod. Extract the ZIP folder to your mods folder - if one doesn't already exist, create it. If your mod select screen in-game is broken, see here to fix it.
For a video overview of mod installation, see here.
For gameplay examples, check the new YouTube channel here.
Recent versions
v5G:
v6:
v6A:
v6B:
For obvious reasons, updates to the mod are on hiatus until the community update is completed.
For a full version history of the mod, see here.
For download links of older versions, see here.
The bug about the barter stations not upgrading should be easily fixed. Just put <S_UpgradeTarget>Barter Station</S_UpgradeTarget> for Merchant Emporium in the planet Improvements file.
The Iconians aren't researching Organic Hulls in my games, as well, even as I continually bump up it's AI value. For my next game, I'm going to go a little bit overboard and see what happens.
Test Games . . .
Sounds like you guys are doing it the hard way. I don't actually play out my test games. At first, I was using the cheats to let them auto-play, but that breaks down between year 2230 and 2231. The game just can't handle it. Then I had a good idea. I used the cheats to create a pop-up minor, then used the cheats to take over said minor. After that I could just hit the turn button over and over again. Takes about 10 minutes to cycle through a year after 3-4 years. I also use the cheats to reveal the whole galaxy and remove the fog of war. That way I can see how the races colonize, trade, and go to war.
Looks like you started taking over races to see their tech advancement, qrtxian. You can go one better. If you take over the race, quicksave, then load the quicksave, it will be just like you were always that race. That means you can go to the Events log in the Civ Manager and it will list the order in which they researched their techs. No guesswork, no overall picture, but an actual tech-by-tech list of what the AI priortizes.
Doing it this way, you can knock out a test game in one or two sessions, rather than taking a week or so to play out a game. It also keeps your development and trading from influencing the game. Also, I usually use the cheats to grant my minor all techs, then auto-build a tricked out dreadnought. That keeps the AIs from frelling with me while the game plays out. It's about a close as possible as I can get to removing the player from the game.
Iconian Test 1, year 5
I played out another 8 months or so of my previous test just for fun.
By this point I declared Test 1 conclusive and moved on.
Iconian Test 2, through year 4
I took a different path than TOV regarding Xeno Coms and the UT. Comparing the Iconian techs to the TechTree showed that Xeno Coms had cost increased from 25 to 100, and UT from 50 to 55. Rather than try to force the Iconians to research the overly expensive techs, I removed the Category and Cost tags from the Iconian tech tree. How would they react to normal category and price?
I changed their Econ tech categories to Economics, then bumped up the AI values. Int Barter - 20, Merchant Emps and Complexes - 15.
For Organic Materials, I changed the Category to Miniaturization, since that's the bonus it gives, leaving the AI value at 15. I did that in Test I and they research it.
Of course, I corrected the category for all the other techs. I've mostly left the AI values alone so I can see how they research the techs as they are.
Also, I reduced Space Mining 1,2, and 3 AI values from 30 to 10 in the TechTree.xml. All the AIs spend too much effort on those techs.
Don't have access to my game. That's just from memory. Other research seems fairly normal for them. Lots of colonization techs, Fleet Battle Tactics and Fleet Defenses (these techs are too high priority for all AIs, I think), and Planetary Invasion techs.
I conclude that the Iconians have enough going against them that they don't need an overly expensive Xeno Comms blocking them from critical techs like Diplo, Trade, and their econ techs. My next step will be to boost AI value of Interstellar Barter to ensure they research it earlier. Well, first I'm going to let this one play out at least another year or two to see how the war with the Drengin goes.
Final, non-Iconian notes from Test 2
That's it, for now.
PS - qrtxian, post in the Kushana Initiative!
You said that they would research Organic Materials after changing the category to Miniaturization. Maybe, if you did the same to the other techs it's not researching, it might finally research them. I'm going to try that in my next test game.
True. I categorized Organic Hulls and Self-Healing Hulls as Hull and they don't get researched. Thing is, the Iconians aren't researching Expert Engineering or any of their other Hull techs. They're only 1-1 on researching Interstellar Construction in my 2 test games.
I'm wondering if not researching hulls is a symptom of their poor overall performance. In my two games they've been at the bottom of the pack and their research levels are always very low. My working hypothesis is that their weak economy is stunting their development which, in turn, prevents them from proceeding to Hull techs. It's like you noticed with Xeno Comms - if it costs too much the AI won't research it no matter how high the value. Either that or maybe they're so weak Hulls never become a priority.
I'm going to proceed by upping the AI value of the economic techs and fixing the Barter Station upgrade like you suggested. I'd like to see how they handle themselves given a functioning economy.
Just wondering, what cheats did you use to do this. Having the AI autoplay would help me a lot in the custom tech trees I'm working on.
I test games with the following steps:
-start a new game in cheat mode
-use the reveal map cheat (ctrl+u)
-create a new minor (ctrl+shift+d)
-take control of that minor by selecting their planet with the mouse, and hitting ctrl+shift+z
-go to the foreign policy tab and leave the u.p.
-destroy your colony
-click on a nearby sun, and hit ctrl+z
-hit the next button, and the game will play out as if you had hit ctrl+z at the start, but it seems to go further before hitting the overload stage without the u.p. interruptions. You can check in on a race by hitting ctrl+z again to interrupt the autoturn (but only at certain times). Take control of the race with ctrl+shift+z to see what it's building.
here's a status report on the Drengin: contradicting my earlier post, I now think it isn't necessary to change their ai away from 7.
In my last two test games with all 10 original races on a large galaxy, the Drengin using AI 7 came to dominate the galaxy with their allies the Yor in the first game, and the Thalans in the second. Unfortunately I'm not sure what the critical change was because I changed more than one thing at a time. The main things I did were to alter their tech tree to emphasize their unique manufacturing and research buildings, increased the aivalue on ultimate shock troops a bit too. Probably more importantly, I also cranked up their aggression from the default value of 70 to 80. It makes sense for them to declare war early and often to get the most mileage from the Super Dominator ability. They still aren't great colonizers, but now they can conquer people effectively enough to make up for it. (caveat, it's also possible that the last two test games were outliers). This was probably supposed to be the way AI 7 works- it's either going to expand by killing you or not at all. But if it doesn't win some early wars, it's screwed, so increasing aggression helps.
By the way, I tried uploading my tech tree changes to i-mod productions, but couldn't get it to accept my upload, just got an error in German.
What he said, except I don't destroy my minor. I find that taking over the minor and clicking the turn button allows a more detailed view of the game. For instance, you'll get the pop-ups when races build their super projects and galactic achievements.
To enable cheats, you just have to add " cheat" to the target of your shortcut. Like so:
"C:\Program Files\Stardock Games\GalCiv2\Twilight\GC2TwilightOfTheArnor.exe" cheat
You could either copy and paste that line, or just add the [space]cheat to the end.
I have this awesome keymap of the cheats that I got from the wiki, but don't see it around there now. If you need it, I could upload it for you.
Edit - About moving this discussion elsewhere, I'd sure hate to lose everything that's been posted here. Perhaps our kind host MarvinKosh would be kind enough to amend the topic name or something.
Tried a test game with that method myself. I'm up to year three and will be continuing, but I've made a few conclusions:
I suspect the Iconian's problem with researching hulls is specific to AI 10. That's the same one used by the Torians, who are known for their swarms of tiny ships. Have you tested them on 11 or 8?
I wonder what AIs 6 and 9 do? Odd that 5, 7, 8, 10 and 11 would be used but not 6 or 9.
Started running another test with the Iconians when I had a big realization. After two years they were once again in last place. Thing is, they started out pretty strong, but by the two year mark they hit stagnation and fell behind. I took them over and started looking around. I quickly noticed that half their colonies had absolutely zero production - the Aquatic and Toxic worlds they had colonized. Then it hit me - Super Adapter isn't a Super Ability at all, but a Super Liability. Who can afford to have half of their first ten colonies at zero production? They were stuck paying maintenance on worlds that produced nothing and couldn't be developed. The AI just isn't smart enough to use Super Adapter. No wonder they fell behind.
To test it out, I ran a new game with no Super Abilities. The Iconians started off slow, and they're still toward the bottom, but they are doing a lot better. They finally researched their Econ techs and researched through Expert Engineering to Advanced Hulls. Actually, with their current position and momentum they are solid contenders.
Bullet point notes and responses:
An unrelated but interesting note: using the cheat to take over AIs is starting to make me wonder whether some of the AI's more ordinary "mistakes" may not be mistakes at all.
For instance, often AIs seem to pursue wars in less-than vigorous fashion. They don't use a majority of their fleets, or they don't invade planets even after destroying the defenders. However, looking at a couple of the test games where an AI was doing this, and studying their fleet positions, rally points, and active spies, suggests that the AIs consider their biggest threats to be races other then the ones they're fighting. Their behavior seems to mimic mine in situations where I only fought a race to keep it down, while actually planning for war with a neutral or supposedly friendly civ. I don't know whether I'm giving it more credit then it deserves, but it's hard to shake the feeling that sometimes, the AI isn't trying its hardest for a reason. Particularly when all of the races I've seen doing this are ones that use AI personality 8, which always seems to be the subtlest in terms of its long-term strategy.
Worth consideration.
That's a fascinating discovery! It's like the AI is thinking farther ahead then we give it credit for. I'll be eager to here more from your tests.
I have been able to get the Iconians to research their fist economics tech, but they stubbornly refuse to go beyond that. It might be because I haven't been playing long enough, though. Same for the techs after Organic Hulls.
My current game played on my custom epic map I made, as I was sick of smashing ctrl+N just so I would get an even playing field map for all the races.
Every square is consisted of either one pq10 habitable or pq 8 I think and the far outer rim parts of the galaxy are radioactive and barren or dead worlds.
Screenie shows every race is expanding evenly by turn 92.
http://d.imagehost.org/view/0540/GC2TA_Orion_Empire_1293498424
8 Ai Races in my map are..
Drengin (red) generic ai
Korath (orange) generic ai
Yor (purple) generic ai
Thalan (dark green)
Torian (yellow)
Terren (dark blue)
Iconian (white)
Krynn (light blue)
me (bright green) tiny blob
Current Galaxy settings -minus preset custom made map design..
Biggest custom map created in editor used.
Alliances,Influence,mega,super,techstealing - habitable planets abundant, number of plantes rare,stars abundant, anomalies abundant,asteroids rare,tech normal,minors 8,extreme planets uncomman,random events random.
Difficulty Painfull
ai cpu max
I am using the marklar mod with some of my own race abilities tweaked like +10 speed for all etc.
https://forums.elementalgame.com/382002/page/2/#2665126
I've declared the above-mentioned test game conclusive. It's late year 7 and the game is a three-way fight between the Drath, Altarians (turned evil by a random event) and a Torian-Iconian alliance. The Drengin are still hanging on but aren't really a factor. Some of the minors are also still alive, and 1 (the Lentzlandians) has managed to avoid war all game.
General race info:
Drengin: Weak, having been used on personality 7. Not much of a power, but managed to survive by militarizing and thugging minors for cash. Failed to research their manufacturing and research techs.
Altarians: Currently the strongest faction but are being worn down. All behavior is fine and I consider them entirely fixed.
Arceans: Done in by their failure to research Planetary Conquest until too late. They got into a war with the Altarians early on and it was fairly even - the Altarians had better ships but a lot less of them - but the Arceans slowly lost planets and couldn't retake them. Eventually they did get them but it was too late. I may want to devalue Planetary Fortification as they got it very fast and built Space Cannons everywhere, but only ever had one Navigation Center, which is much more important. On the other hand. they were falling behind and apparently considered weapons research a priority. Might need some tweaking to AI values, but are otherwise done.
Torians: No problems.
Yor: Researched techs they've been failing to get in other games, but they never colonized and so were very weak. First civ eliminated. That said, their research priorities were good and they should work fine with different AI.
Korx: A failure. Never did research Cut-Throat Trading et al and didn't ever develop weapons - their most advanced warship was a fighter with the "Space Weapons" era particle beam (not the kind that comes after Laser V). They bought a lot of ships from the Drath, but too late.
Drath: Weakest of the surviving major powers (not counting the Drengin) but in terms of techs and behavior they're fine.
Thalans: Research seemed fine. However, their military was second-rate and they lost to the Torians. No actual issues, though.
Iconians: Did very well. They are the secondary partner in their alliance with the Torians but are ranked 3rd, above the Drath (who are powerful in their own right). Finally, I got them to research Organic Hulls and Materials, though not Self-Healing Hulls. As with the Korx unique techs, I suspect Iconian pursuit of the OM branch is based on how well they're doing. Also did well in terms of manufacturing and economy techs. They also successfully went beyond Advanced Beam Weapons. Their one glaring error was failing to research Expert Engineering; with tech trading off, they never built anything larger then a heavy fighter. That still needs fixing, but otherwise, I may have actually gotten the Iconians functioning as intended.
Also worth noting was that weapons research went far higher then normal. The Altarians have gotten all the way up to Disruptors II, though no other AI has anything better then early techs in the third weapons tier (phasors/graviton drivers).
In short, the Altarians, Drath, Torians, Thalans, and Yor are pretty much done. The Iconians and Arceans need tweaking but are still at their best ever on TOA. The Drengin and Korx still need some work. I intend to starting investigating the other civs (Terran, Krynn, Korath) in my next test, although for now I'm playing DA.
Incidentally, as far as the previously noted observations on AI strategy, they seem to be associated with the races using personality 8. This fits into my established theory that 8 is the "strategist", 10 is the "expander", and 11 is the "tactician" among the AIs.
Interesting post, qrtxian. I would add that ai 7 is the "transport/spore-ship rusher", or something along those lines. They have to conquer another race relatively early in the game to compensate for their failure to colonize aggressively. But if they can do that, they can be very tough. The Drengin can do very well on personality 7 provided you turn their aggression very high (I'm using 80) and increase the aivalue of their manufacturing/research techs to 25. Since making those changes, I've seen them win out as part of a victorious military alliance in three of my last ten-or-so games. These test games were all on Large maps against nine other races using their default personalities (w/ no mega events, no surrenders, tech trading and tech stealing enabled, and no tech brokering).
I'm trying the same thing with Yor, but so far they still seem to be struggling on 7.
Thanks for the report. My focus on the Iconians has paid off. I'd say they're ready for release, but I'm doing some polishing. I'll respond by race, then go general.
Drengin: Do alright. I only made minor changes for them:
They get Diabolical Research pretty early with AIV 35, but never go further. I've seen them research their Slave Pits line, but not often. They're typically pretty anemic overall, but they scrap very well when somebody picks a fight with them. Certainly in need of work.
Arceans: How do you pronounce this? I used to think Are-Key-ans, but not I'm thinking more like oceans - Are-shins. Anyway, they tend to do well in my games and all I fixed was Xeno Geology and Planetary Fortification. They do take their sweet time getting to the navigation techs, but I'd look to increasing the AI value of the nav techs first. I've modded out the Arcean speed penalty, so it hasn't been hurting them as badly in my games.
I see Weather Control Zeniths and Space Cannons on all their planets as a sign of strength. The Cannon and WCZ improvements are coded so that they prioritize them. Researching either later only makes it so the older colonies don't get one.
Torians: Do fine. I tried AI 10 again and it sucks. Set them as Thalan AI (8) and they are monsters. They don't research their healing pools or advanced research building techs, though.
Korx: Odd. Are your Korx perpetual weaklings? Poor position or something? They are typically mid-to-top tier players in my games and always research their techs. I have them set as Korx AI (11). The Tolmekians are set as Human AI (8) and also research them. They do take a little time, and won't get around to it as a priority if they're not doing well. A slight boost to AI value would surely do the trick. Maybe 5 points all around.
Iconians: My babies. I'm always so proud when they hold their own. Their economy always struggles since they have not a single econ boosting tech and only have the Barter Station buildings. That said, I have them researching their techs pretty much as they should. Expert Engineering and the Organic Materials line are down solid. The problem with those was category and very low AI values. The key is setting AI values in line with the other techs.
I boosted Interstellar Bartering so it's now a priority, though they won't research past it until quite a bit later on. With all the other changes, they stopped researching Interstellar Refining, so I boosted it a bit. They usually don't go for Molecular Fabrication until very late game, so I may boost that a bit too. Otherwise, I'm very happy.
General: I noticed through all this that about half the races never research Xeno Comms and Universal Translator. Since they're the gateway to diplomacy, trade, tech exchange, treaties and alliances it seems like a major disadvantage.
That's it for now. Gotta run. Of all the things Frogboy could fix in his latest patch, AI 7 and 10 are highest on my list. It would almost be like a whole new game with 2 more working AIs.
I just noticed that the Torian's Cultural Enlightenment, which leads to Healing Pools and other morale techs, have a really low AI value (5). This prevents them from researching them, at least, not early. I've bumped up the value considerable, and they should research these techs on their own in my next test game.
Also, I've always pronounced the Arcean's name as ar - SEE - ans
That explains it. That's another one I noticed as a big "what?" that they didn't research. Actually, most of the yellow techs are pretty well ignored. That's something I intend to look into.
Also, all this testing with tech trading off confirms that the AI seems to pick the same weapon type for all races. It's not tech trading that gives them all the same weapon. Interestingly, AI 10 researched two different types of weapons in my last game. Too bad it sucks at development. It would add some needed diversity.
I may try your fix for the Drengin, deweyjohn. I'd like to see them go to war. I guess the same treatment should work for the Korath, too.
It's interesting, because the Korx were originally one of the most successful races when I started with TOA. Lately, they have indeed been doing badly.
However, I've just started up a normal non-test game against mostly custom opponents: Altarians, Torians, Krynn, 2 Thalan civs, 2 Korx, 1 Terran, and 1 Arcean. Both Korx civs are doing quite well and have gotten their techs fine, although AFAIK they have yet to build the Fortified Freighter Command. This pretty much confirms that the Korx's previous behavior is due to their poor position. Hard to tell what the problem with it was, although I would guess either position or no tech trading - my customization of the Korx has little to no research bonus.
As for the others, the Altarians and custom Arcean opponent haven't yet been met but seem to be doing well, judging from reports of wonders and so forth being built. The Altarians appear to be ranked 3rd and the Arceans are ranked 6th. Also successful are the Torians (1) and Krynn (2). One Thalan civ is unknown, but the other is a total disaster - only four planets colonized and almost no development. Almost certainly not because of a bug with the AI, though, but simple luck of the draw.
I don't have any trouble with the Torians researching their morale and research buildings. They've developed Universities, which is around par for the current time period. They don't have Healing Pools but did get Cultural Enlightenment - I suspect they're ignoring it because someone gave them the normal Xeno Entertainment techs.
This is my first real test of the Krynn, and they seem to have no problems. They're up to The Way in terms of research, and have built a Super Spy Training Center on their HW. They also are researching down the Terror Star path (!), although slowly. That could be interesting, given their highly aggression behavior... So far, anyway, the few changes I made to the Krynn at the beginning of this project(fixing some odd categories and tweaking a few values) have been enough. I'd say they're fine.
Edit: I also say "Ar-SEE-ans".
Hello again! I found that the Torian's morale techs above healing pools also have low AI value. I've bumped those up as well.
Also note that in the Drath tech tree, healing pools are categorized as "Medical" instead of "Entertainment", explaining why they never research it.
Medical instead of Entertainment?! Will the stupid categories ever end?
Sigh.
As a short note, I went through the entire tech tree and made subtle bumps in AI value for all the culture and government techs. Now I've got Star Republics and Star Democracies popping up, and the Torians have even built their political capital. Incidentally, the Torians researched their Cultural Enlightenment followed by Healing Pools.
It's like a whole new game even without having two more AIs. So cool to see them work around the tech tree and actually develop a bit.
Aaaand, my change-log just broke onto page four. TechTree takes up just over a page and Iconian TechTree just under. Good times.
I just tested my changes, and the Torians are now researching their moral techs! So far, they are up to Wellness Falls.
Tolmekian, I would be interested in seeing your change log one of these days. Four pages is quite impressive.
There are many great features available to you once you register, including:
Sign in or Create Account
