Ahh, you're right. I play on Giant, not Large. I always get Giant, Large, and Huge mixed up.
Realism Invictus
- Thread starter Walter Hawkwood
- Start date
I'm on the latest SVN and I've been trying to generate maps with Totestra on Giant size and for some reason they all end up like this, almost fully flat except for some hills or forest on the area certain civs spawn and with rivers streaking from the upper left to the bottom right.
Spoiler Broken map :
Are you using the "Regenerate Map" function? That function has the pictured effect on my maps sometimes, but it never happens when going to the main menu and starting a brand new custom game.
Walter Hawkwood
RI Curator
No, of course AI can have more than one attack stack, but as long as they have separate objectives. AI will not inter-operate two or more stacks when attacking from a single direction to stay under logistics penalties, as reasonable human players often do.
What he said. If I were to make a Reformation-centric scenario, Protestantism and maybe even Calvinism would get the same treatment. Or in a Middle-East-centric one, Shi'a and Ismailite. Sandbox doesn't need more religions.
Given the time passed, I wouldn't bet on it. TBH, even the more realistic and much more modest goal of making Civ 4 run with Colonization exe was never seriously tried by the community, and it would already have been a step above what we have.
Thanks, fixed. This one is a stark example of what wrong indentation can do in Python, as the fix consisted of literally deleting one tab.
Fixed, kinda - by removing the symbols altogether. The gamefont symbols are well and truly cursed. I cannot understand how the same (not similar, the actual same) string being parsed for the log and for being displayed on the main screen has a different symbol in it, offset by one, and only if it is separatism-related. Just. No.
The shared war modifier goes away almost instantly after no longer in a shared war, so not even then.
The tech progress seems to generally be too quick these days, and that is something that will need to be thoroughly tested and balanced. Around late November, I'll freeze any new balance/AI-related changes to run lots of test games and adjust the overall balance. For now, I'm not starting yet, as I assume more AI-related stuff is incoming.
No, I see it even in all-AI (and thus Noble-level) autoplays. While it might be exacerbated by difficulty, it is true across the board.
It is tedious, and it is harder on AIs than on humans.
Thanks, I'll investigate, though Totestra is a rather convoluted script, and coded by a guy with some rather exotic coding ethics (though a good coder no doubt).
I wonder what you mean by that
Walter Hawkwood
RI Curator
Let's just say converting it to work with RI was an experience I wouldn't want to repeat.I wonder what you mean by that
giorgio1234
Chieftain
Yes, I never liked this feature of python. For someone who is used to writing code in c or java it is quite tricky.
Yes, I tried to fix the problem my own, and I got the font issue. It is possible to get the yellow and red fist by increasing the index by one (but not the orange one, it gives an error).Fixed, kinda - by removing the symbols altogether. The gamefont symbols are well and truly cursed. I cannot understand how the same (not similar, the actual same) string being parsed for the log and for being displayed on the main screen has a different symbol in it, offset by one, and only if it is separatism-related. Just. No.
But I agree with your solution: if there is something I don't understand, I take it out.
Takofloppa
Chieftain
- Joined
- Sep 6, 2024
- Messages
- 18
It would probably be rather tricky to implement, as split stacks are more prone to all forms of exploitation by players - so while the current situation is not perfect, it's still much easier to manage by AI even with all logistical penalties, and I would rather leave it that way.
AspiringScholar
King
There look to be some excellent quality of life improvements in this current revision! I can't sit down and write a fully thorough message right now, but aside from all of the healthy balance tweaks, the Shipyards stipulation on barbarian presence seems like a particularly creative solution to the problem AllTheLand brought up about near-instant development of new world civs after the spawn.
Two small bugs from SVN 5430.
1) New collateral damage clarification has a typo: max damage and max number of targets are mixed up though it is correct in civilopedia.
2) Some slaves that I got had a bigger select radius than the tile radius. If it helps I played as Russia and fought barbs and South China.
1) New collateral damage clarification has a typo: max damage and max number of targets are mixed up though it is correct in civilopedia.
2) Some slaves that I got had a bigger select radius than the tile radius. If it helps I played as Russia and fought barbs and South China.
Actually, I just learned this is being tried currently. Look at this thread.
Takofloppa
Chieftain
- Joined
- Sep 6, 2024
- Messages
- 18
It's not really a problem of splitting these stacks itself, but more about making them cooperate competently - the example of 1UPT in Civ5/6 clearly shows it's not a trivial problem to solve and without some clever solution you are ending with mindless carpet easy to defeat from a player perspective. Civ4 stacks (like them or not) have the advantage that they are easy to digest by computer both from a planning and moving perspective. Even Civ7 is slowly changing its approach by introducing commanders that merge multiple units in one tile, which I'm sure was done to make some tasks easier for AI.
AllTheLand
Warlord
- Joined
- Sep 18, 2021
- Messages
- 115
Observations from my game on a small RI generator map:
- The num-cities maintenance values need to be tied to the number of land tiles, not to map size. The map was generated with a low sea level (I didn't even think about it before starting the game, it was just the default from the previous time I had used it) and as a result the balance in this regard is out-of-whack.
- I'm playing with Pericles. The creative trait is near useless apart from the very early game to get a city to be able to use its full cross (it does reduce a few building costs, but not much, and it helps marginally in culture wars, but I still got my iron-city flipped by cultural revolts right when I needed iron the most to fight off an invasion), and populist is costing me 24gpt in 438. It seems almost like an entire difficulty level harder compared to playing with a proper leader.
- The way the cost of civics scales with the number of cities is problematic in my opinion. It does double-duty with maintenance and the tech research penalty penalizing additional cities. I understand the general goal of avoiding a mad rush to create as many cities as possible in the early game, but this creates most of its burden after the expansion phase is over when expansion is blocked firstly by other civs holding the land.
- There is some hint that says bronze and iron are not entirely interchangeable. That's true, but really, without iron, you are stuck to classical units that are just bad against late classical and early middle ages iron units. I was able to repel armies of 7 Str. swordsmen and 8 Str. cavalry because the AI sent too small stacks, dispersed units (the AI can be baited into attacking an isolated unit and creating itself an isolated unit in the process), and because the Greeks have a very good 5 Str. unique unit that has enough bonuses to have good attacking odds on plains even against 7 Str. and 8 Str. units. If I had a civ with more vanilla units, or if the AI let me less opportunities to chip away at its forces, I would have collapsed. Bronze is not a replacement at all for iron.
- The AI asking prices for some resources is insane. A proper trade deal will have both sides benefit from it, but the AI has no problem asking even more than the receiving civ would get as a benefit. For example, for a +1 happiness +1 health spices, I get asked 53 gold per turn for 12 cities. The one population extra I would get from it would barely be able to break even once all cities hit the new limit, but I would be then completely dependent on the AI to not break the agreement. Just raising the culture slider is quite cheaper. I wonder if the AI is accounting the +1 health from the hindu temple and the +1 health of the zoroastrian temple and thinking it's offering a +1 happiness +3 health good?
- The AI is way too restrictive on what resources it' willing to trade. The #1 civ in my game has been giving me open borders for a while, but it stubbornly refuses to allow to trade some of the happiness resources it has because "we are not so close" (-5 relationship mostly driven by the massive religious penalty). If real life shows anything, it's that even countries that are quite hostile to each other will still trade unless it turns to war or the conflict is extremely bitter, especially for non-strategic resources.
- Half of the civs don't even have any resources I don't have available for trade... And they actually do quite little ressource trade between each other too. Several significant luxury resources are available only once or twice on the entire map (that started with 6 civs but got more from barbarians settling and a case of breakaway civ), so there is very little oppportunity for them to be traded.
- The AI is extremely stubborn when it comes to signing peace. It will demande highly unrealistic terms even if its armies got defeated several times in a row. I would lean towards reducing the war-weariness bonuses of the AI and making sure the AI is more willing to give up a war when it suffers from high war-weariness.
- I think I didn't report it from my previous game, but the doctrin unlocked by Heraldic mostly obsoletes the need for the flanking promotions (one or two can still increase retreat chances, but you get 50% retreat chances and immunity to first strikes with the 3 promotions), however you still need flanking 3 to be able to unlock the +1 movement radius promotion... It was a sad discovery.
- The num-cities maintenance values need to be tied to the number of land tiles, not to map size. The map was generated with a low sea level (I didn't even think about it before starting the game, it was just the default from the previous time I had used it) and as a result the balance in this regard is out-of-whack.
- I'm playing with Pericles. The creative trait is near useless apart from the very early game to get a city to be able to use its full cross (it does reduce a few building costs, but not much, and it helps marginally in culture wars, but I still got my iron-city flipped by cultural revolts right when I needed iron the most to fight off an invasion), and populist is costing me 24gpt in 438. It seems almost like an entire difficulty level harder compared to playing with a proper leader.
- The way the cost of civics scales with the number of cities is problematic in my opinion. It does double-duty with maintenance and the tech research penalty penalizing additional cities. I understand the general goal of avoiding a mad rush to create as many cities as possible in the early game, but this creates most of its burden after the expansion phase is over when expansion is blocked firstly by other civs holding the land.
- There is some hint that says bronze and iron are not entirely interchangeable. That's true, but really, without iron, you are stuck to classical units that are just bad against late classical and early middle ages iron units. I was able to repel armies of 7 Str. swordsmen and 8 Str. cavalry because the AI sent too small stacks, dispersed units (the AI can be baited into attacking an isolated unit and creating itself an isolated unit in the process), and because the Greeks have a very good 5 Str. unique unit that has enough bonuses to have good attacking odds on plains even against 7 Str. and 8 Str. units. If I had a civ with more vanilla units, or if the AI let me less opportunities to chip away at its forces, I would have collapsed. Bronze is not a replacement at all for iron.
- The AI asking prices for some resources is insane. A proper trade deal will have both sides benefit from it, but the AI has no problem asking even more than the receiving civ would get as a benefit. For example, for a +1 happiness +1 health spices, I get asked 53 gold per turn for 12 cities. The one population extra I would get from it would barely be able to break even once all cities hit the new limit, but I would be then completely dependent on the AI to not break the agreement. Just raising the culture slider is quite cheaper. I wonder if the AI is accounting the +1 health from the hindu temple and the +1 health of the zoroastrian temple and thinking it's offering a +1 happiness +3 health good?
- The AI is way too restrictive on what resources it' willing to trade. The #1 civ in my game has been giving me open borders for a while, but it stubbornly refuses to allow to trade some of the happiness resources it has because "we are not so close" (-5 relationship mostly driven by the massive religious penalty). If real life shows anything, it's that even countries that are quite hostile to each other will still trade unless it turns to war or the conflict is extremely bitter, especially for non-strategic resources.
- Half of the civs don't even have any resources I don't have available for trade... And they actually do quite little ressource trade between each other too. Several significant luxury resources are available only once or twice on the entire map (that started with 6 civs but got more from barbarians settling and a case of breakaway civ), so there is very little oppportunity for them to be traded.
- The AI is extremely stubborn when it comes to signing peace. It will demande highly unrealistic terms even if its armies got defeated several times in a row. I would lean towards reducing the war-weariness bonuses of the AI and making sure the AI is more willing to give up a war when it suffers from high war-weariness.
- I think I didn't report it from my previous game, but the doctrin unlocked by Heraldic mostly obsoletes the need for the flanking promotions (one or two can still increase retreat chances, but you get 50% retreat chances and immunity to first strikes with the 3 promotions), however you still need flanking 3 to be able to unlock the +1 movement radius promotion... It was a sad discovery.
Last edited:
Takofloppa
Chieftain
- Joined
- Sep 6, 2024
- Messages
- 18
Interestingly I had exactly the opposite experience when I was playing last time some weeks ago - AI was pretty happy to make peace after barely touching their attacking stack, so I assume it's very situational. They also immediately agreed for open borders which was kind of weird, but possibly that was fixed in some latest SVN versions from what I read in the changelog.
AspiringScholar
King
"Fixed a long-standing bug with ranged attacks. Unit with most strength would be chosen to defend, even if it was already reduced to bombardment limit, disallowing further bombardment of the stack. Now if there are some units with strength over the limit in a stack yet, they will be properly chosen"
Amazing!
This one was rather frustrating.
Amazing!
This one was rather frustrating.trashmunster
Chieftain
- Joined
- Feb 20, 2024
- Messages
- 91
Yeah the goal is to create the most bland, most evened-out experience, with maximum predictability, Where all the "fun" is to optimize yields, count numbers and outproduce dumb AI with its production bonus. Because that's what this game about, everything must work same always.
AspiringScholar
King
Are you alright? That seemed to come from nowhere, and your other posts seem to have a different tone.
Similar threads
