Realism Invictus

[Watermelon]

I focused on this thought more and decided that I was being somewhat childish by walking away because what I wanted wasn't being served to me. So I started delving into how map generation works and modifying Totestra to suit my wants more. Been playing around with it for the past day and am now getting continents that suit my preferences more: generally bigger, leaning wider than longer, and less "noise" creating weird and snaky continents. I'll probably keep messing with it as I play, but it's been a positive change so far.

Happy to share my edits if anyone's curious about how to go about modifying the map generation script towards their own preferences.


My rough thoughts/ideas:

• Research bonus requires a civ to have open borders with you, be in your trading network (or whatever it's called when you can trade resources), and for that civ to have completed research on that tech
• Default bonus +10% per civ
• Once you research Alphabet, additional +5% per civ from other civs with Alphabet
• Once you research Paper, additional +5% per civ
• +5% per civ when running each of Republic, Inclusivity, Merchant Families, and Free Religion civics
• -5% (to min 0) per civ when running each of Tribal Union, Traditional Custom, and Protectionism civics
• Theocracy and Civil Religion only allow the bonus from civs that share your state religion
• Free Religion gives +2% per civ for each religion in any your cities
• Qumran provides +10% per civ (instead of the happiness bonus. I often get it and the volatility of the research rate makes it a double edged sword)
• +100% from Information Networking
• With (tech ???--medieval maybe), a Great Spy can create a Spy Network national wonder, which grants the bonus from all civs within trade network, regardless of open borders
• With (tech ???--something roughly early 20th century), a Great Spy can create a Intelligence Agency national wonder, which grants the bonus from all civs, regardless of trade network or open borders

I'd love to hear your edits to Totestra, because I like it most for its "natural" but dislike the vastly different starting locations that can happen due to some snaky bit of the map or because the giant desert is too giant and leaves little space for settling.

Also some random thoughts on balance:
-Turkish Janissaries certainly come too late with Absolutism, which is on the same column as flintlock musket, while Janissaries only have 10 Str. Historically these guys appeared much earlier, and adopted firearms once they become widespread. Maybe just make them available at Arquebus like the Tercio (Spain and Ottomans being quite the rivals at the time as well)? Or at the latest Administration, which was kind of an Ottoman thing anyways.
-Spanish Mineria just feels sad. It gives one commerce more on an improvement you might get 2 in your entire empire. If we would reeeallly stretch the definition of serfdom, it might be fitting to give the spanish Minerias about 2 Commerce extra under Slavery (Classical Spanish Silver mines) and Serfdom (New World silver mines), because unlike Protectionism you are going to be in these civics for a while.

 

[AllTheLand]

I played around with RI recently. I had an old 3.5x version installed previously but I had graphical issues, so I only started with 3.60. There has been some significant changes since that release, but since I had already started games and the saves are not compatible, I didn't jump to 3.61 or the SVN version yet. But I think most of my experience on 3.60 remains relevant.

I won't go in-depth on my praise of what RI does right, in many aspects it's fantastic, leaps and bounds above BtS with the main downside being how much more time-consuming finishing a RI game can be. The technological progression is much more interesting, you actually get the time to use units of each age, the logistics penalty on per-tile stacks is excellent, the Great Works of art and science are really interesting, there is a lot of care to details making everything fit together, and it is in general much more immersive. It's an awesome mod and the hard work that went into it really shows.

My background is playing some vanilla Civ3 and Civ4 (the Civ3 music in the RI Classical era feels great), and then more recently mostly playing with the AdvCiv mod for BtS. I tend to reload an earlier save if I realize I made a major screw-up, which of course makes it easier to deal with higher difficulty. I was mostly playing at Emperor in AdvCiv. There are some elements I liked in AdvCiv I tend to miss in RI, notably the AI being smarter about war declarations and the way city revolts work. And I think something else which escapes my mind at the moment.

My first game has been with Poland in the Europe scenario on Monarch difficulty, and my second game has been with Germany on a random map generated by RI PlanetGenerator on Immortal difficulty (with picked enemies that have the wonder penalty trait, so not faithful to a full random game). I stopped the first game in the early Middle Ages as I had over double the score of my nearest competitor. My second game is in the middle of the Middle Ages although I don't think I'll go beyond the Renaissance.

I'd like to go in more details later on about what happened in those games, but there are some common themes I noticed:
- World Wonders are somewhat too important. This is perhaps subjective, but some of them give extremely strong bonuses and they are the dominant source of GP points for the early eras, with GP points being quite valuable. One of the clearest example is the Kreml. The -10% maintenance bonus across all cities is by itself overpowered (It's almost effectively -20% once you have both a courtroom and a town hall in a city), and it also gives GP points and if you get Servage it's +1 hammer / + 1 gold per city with a building that's very easy to get.
- The Heroic Epic that gives +100% GP points gives a massive incentive to just put every single GP-giving wonder in the same city. In my first game, I tried building a few wonders in a side city, but its GP generation was so behind my capital that it managed to produce a GP only once (I ended the game into the Middle Ages). The tradeoff between losing some GP generation and increased control of the GP type that gets created is simply not there. The no limit on national wonders per city, which I don't mind by itself, makes that choice even more obvious.
- Despite the high number of buildings available to each city, it's quite possible to reach a situation where most of the cities (except some more recent ones) has all the helpful buildings, and nothing really useful to do as the Civ has already enough military units. I had almost never used the "convert hammers to gold/science" options in previous games, I often had half my cities or more using it here. I think there are benefits in not needing half the game to get a city up to standards (and the "old building upgrades into new building" trick, such as Well into Aqueduct is really clever), but reaching a point where most cities don't have much useful to do with their hammers seems strange. It certainly taught me the value of building cottages, because hammer-prioritization doesn't seem that useful when the limiting factor is gold instead. In AdvCiv I could mostly focus on farms, mines and sawmills, not anymore.
- Number-of-cities maintenance costs are outrageous. I played both games with number-of-cities tech cost scaling, and I think it's a fine option by itself. Beyond the realism justifications that can be given (difficulties of coordinating research and spreading research across a vast empire), it helps to limit the snowball benefits of expansion. However, number-of-cities maintenance create situations where you can have completely unclaimed space fit for settling a new city, and still let it aside for dozens or even hundreds of turns because all your old cities are suddenly going to be losing much more money. So you just wait for the next tech that can improve tax revenue enough to justify the expense of settling open terrain or conquering some territory from a neighbouring civ. This mechanic from the base Civ4 game is conceptually flawed, and in RI it feels even worse. Unfortunately it's not just a matter of tweaking a few numbers. Reducing drastically the number-of-cities maintenance would give too much of an advantage to expansionist strategies and make them too easy once the ball starts rolling. But introducing new mechanics is much more involved than simply tweaking numbers.
- AIs are really stupid when evaluating Open Borders. With Open Border tech transfer, there is limited interest in giving Open Borders to someone who is really behind in techs, but there is a strong interest in keeping Open Borders with a civilization that's ahead, as it propells a civ's own tech research. Very early in the game, civilizations are giving Open Borders relatively easily, but then they seem to be completely set against it. In my first game, I had cancelled most of my open border agreements when I got so ahead in techs that I couldn't get benefit from it in this regard, but later on as I expanded I wanted to make some deals to increase my trade route revenue. For dozens of turns, multiple civilizations that had diplomatic relations with me in the +6 to +10 range (thanks to shared religion) repeatedly refused signing Open Borders, despite the massive tech boon this would have been for them (compared to the mild benefit I would have received). The only way I managed to sign some deals was when I joined a war alongside some of these civs after they demanded me too.
- The "You are overtaking us" mechanic feeds into this. I had a rival civilization cut Open Borders and fall behind in tech as a consequence.
- AIs have a nasty tendency to just make stupid diplomatic demands, such as joining a war (but without paying anything to do so or having existing deep ties with the demanding civilization), converting to a religion that's present in maybe 2 cities out of 20, changing civics, and so on. Then they get upset when the obviously stupid demand is rejected. On the other hand, they almost never seem to come to propose a mutually beneficial trade deal.
- AIs have a nasty tendency to declare war to civilizations that are dozens of tiles away. This was especially obvious in the Europe scenario. Getting embroiled in useless wars sets the AI behind a lot. A "We are far from each other" diplomatic relation bonus ought to be considered. Or some way to reduce typical negative modifiers such as wrong religion when the countries are distant.
- The way AI considers civilization currently at war with its target to evaluate the military balance is deeply problematic, as it will often be possible for the target to achieve a separate peace and then the opportunistic declaration of war appears really foolish. In my second game, I had Aztecs in the South and Poles in the North, and after Aztecs attacked me with a proper army, Poland jumped in. But they never sent a proper attacking army my way (only 3 skirmishers), and after I repelled the Aztec assault and made peace with them, I could easily turn back on Poland and took two cities with very little resistance.
- Inflation is really annoying, and there is nothing realistic or engaing about it. It's just there because the gold revenue from normal tech progression and city expansion increases too quickly compared to the costs. If we take the example of a hypothetical stagnant civilization (no new tech, city expansion, workers improving tiles...), revenue and expenses should be flat over time, but with inflation the expenses explode. Just as with maintenance costs however, properly designing and balancing an alternative method would be quite involved, although both could be addressed at once. I think the keys are reintroducing more building-maintenance like in Civ3 (I noticed that the cathedrals have a 1-gold maintenance cost so it's feasible, although if it applies before percentage modifiers such as the tax office, that's not ideal), having separatism/revenue penalties for peripherical cities, and having higher-tech units incur a higher maintenance cost (is that feasible? Also an additional maintenance cost per unit type would be interesting).
- The loss and capture of cities causing destruction and setbacks is expected, but the damage tend to be so high that losing and then reconquering a city feels really bad. This problem is perhaps even worse in vanilla Civ4, but at least for me it fosters a mentality of not letting even one city fall, unless it was founded quite recently.
- I noticed earlier in this thread that the difference between the displayed epidemic rate and the real epidemic rate was discussed. I found it very puzzling when playing, because in my first game I went to great effort to keep epidemic rate low but despite having it often at 2-3% in a few cities, I only got my first epidemic after a few hundreds of turn. I was even asking myself if the feature was not broken.
- I haven't analyzed it in detail, but some leader traits seem so much better than others. Despite all the many sub-bonuses given by a trait like Conqueror, it still doesn't hold a candle to a trait like Legislator. Clear-cut economic bonuses are helpful for every single strategy and in every single situation, whereas some other traits, and especially military traits, have a very narrow niche. Since economic bonuses also make it possible to afford more units in the first place, I think some significant changes are warranted.
- The Folwark is OP. Most custom tile improvements are very niche, only being available on a small number of tiles, but here it's a strong +1F/+2G on half the tiles until mechanized farms become good enough to beat it.
- I don't like the penalty for resarching techs of the next Age before a certain date, especially as going into the next age earlier already forces you to deal with a +1/+1 unhealthyness/unhappyness penalty. Thankfully I could simply switch it off for my second game with the tick of a box, so I don't have anything to complain about.
- Random map generation for RI is a whole can of worms by itself. Some resources are so important that a map that doesn't place enough of them is going to be very unbalanced (which is made even worse by how reluctant AIs tend to be with resource trading and how little excess resources available for trade there is). In my second game, it didn't feel good to have a "thermal baths" unique tile improvement that I couldn't build even once within a 20 tiles radius around my capital...

I also wanted to ask: are there some ways to have AIs play by themselves and watch what they are doing? I would find it very interesting to make some AI tweaks and look at the results, or simply to observe how different civilization and traits fare in different games. Although I also need to learn how to compile the .dll and I don't know the codebase, I shouldn't have much trouble tweaking the C++ code.

 

[Takofloppa]

I also wanted to ask: are there some ways to have AIs play by themselves and watch what they are doing?

Yes, there's autoplay incorporated in RI, what you need to do:
- find CivilizationIV.ini in My Documents > My Games > beyond the sword, find "CheatCode" and set it to "chipotle". This will allow you to use Ctrl+Z shortcut which reveals the whole map so you can see how everyone is playing
- press Ctrl+Shift+X (I hope I remember correctly) and enter how many turns you want to autoplay

It can't be stopped as far as I know, so it's better to run it in some reasonable intervals.

 

[sazhdapec]

- The way AI considers civilization currently at war with its target to evaluate the military balance is deeply problematic, as it will often be possible for the target to achieve a separate peace and then the opportunistic declaration of war appears really foolish. In my second game, I had Aztecs in the South and Poles in the North, and after Aztecs attacked me with a proper army, Poland jumped in. But they never sent a proper attacking army my way (only 3 skirmishers), and after I repelled the Aztec assault and made peace with them, I could easily turn back on Poland and took two cities with very little resistance.

Aztecs just asked Poland to join a war. And they obliged.

- The Folwark is OP. Most custom tile improvements are very niche, only being available on a small number of tiles, but here it's a strong +1F/+2G on half the tiles until mechanized farms become good enough to beat it.

It IS strong but Serfdom has a critical flaw: low production from craftsmen. Industrial era buildings and especially wonders are so expensive that it's painfully slow to build them. I'm currently playing as Poland and in the early industrial era I just had to switch to a more modern civic to get more hammers.

I also wanted to ask: are there some ways to have AIs play by themselves and watch what they are doing?

Another way to do it is to open WorldBuilder and in a game settings menu set a number of turns for AI autoplay.

 

[Watermelon]

I played around with RI recently. I had an old 3.5x version installed previously but I had graphical issues, so I only started with 3.60. There has been some significant changes since that release, but since I had already started games and the saves are not compatible, I didn't jump to 3.61 or the SVN version yet. But I think most of my experience on 3.60 remains relevant.

I won't go in-depth on my praise of what RI does right, in many aspects it's fantastic, leaps and bounds above BtS with the main downside being how much more time-consuming finishing a RI game can be. The technological progression is much more interesting, you actually get the time to use units of each age, the logistics penalty on per-tile stacks is excellent, the Great Works of art and science are really interesting, there is a lot of care to details making everything fit together, and it is in general much more immersive. It's an awesome mod and the hard work that went into it really shows.

My background is playing some vanilla Civ3 and Civ4 (the Civ3 music in the RI Classical era feels great), and then more recently mostly playing with the AdvCiv mod for BtS. I tend to reload an earlier save if I realize I made a major screw-up, which of course makes it easier to deal with higher difficulty. I was mostly playing at Emperor in AdvCiv. There are some elements I liked in AdvCiv I tend to miss in RI, notably the AI being smarter about war declarations and the way city revolts work. And I think something else which escapes my mind at the moment.

My first game has been with Poland in the Europe scenario on Monarch difficulty, and my second game has been with Germany on a random map generated by RI PlanetGenerator on Immortal difficulty (with picked enemies that have the wonder penalty trait, so not faithful to a full random game). I stopped the first game in the early Middle Ages as I had over double the score of my nearest competitor. My second game is in the middle of the Middle Ages although I don't think I'll go beyond the Renaissance.

I'd like to go in more details later on about what happened in those games, but there are some common themes I noticed:
- World Wonders are somewhat too important. This is perhaps subjective, but some of them give extremely strong bonuses and they are the dominant source of GP points for the early eras, with GP points being quite valuable. One of the clearest example is the Kreml. The -10% maintenance bonus across all cities is by itself overpowered (It's almost effectively -20% once you have both a courtroom and a town hall in a city), and it also gives GP points and if you get Servage it's +1 hammer / + 1 gold per city with a building that's very easy to get.
- The Heroic Epic that gives +100% GP points gives a massive incentive to just put every single GP-giving wonder in the same city. In my first game, I tried building a few wonders in a side city, but its GP generation was so behind my capital that it managed to produce a GP only once (I ended the game into the Middle Ages). The tradeoff between losing some GP generation and increased control of the GP type that gets created is simply not there. The no limit on national wonders per city, which I don't mind by itself, makes that choice even more obvious.
- Despite the high number of buildings available to each city, it's quite possible to reach a situation where most of the cities (except some more recent ones) has all the helpful buildings, and nothing really useful to do as the Civ has already enough military units. I had almost never used the "convert hammers to gold/science" options in previous games, I often had half my cities or more using it here. I think there are benefits in not needing half the game to get a city up to standards (and the "old building upgrades into new building" trick, such as Well into Aqueduct is really clever), but reaching a point where most cities don't have much useful to do with their hammers seems strange. It certainly taught me the value of building cottages, because hammer-prioritization doesn't seem that useful when the limiting factor is gold instead. In AdvCiv I could mostly focus on farms, mines and sawmills, not anymore.
- Number-of-cities maintenance costs are outrageous. I played both games with number-of-cities tech cost scaling, and I think it's a fine option by itself. Beyond the realism justifications that can be given (difficulties of coordinating research and spreading research across a vast empire), it helps to limit the snowball benefits of expansion. However, number-of-cities maintenance create situations where you can have completely unclaimed space fit for settling a new city, and still let it aside for dozens or even hundreds of turns because all your old cities are suddenly going to be losing much more money. So you just wait for the next tech that can improve tax revenue enough to justify the expense of settling open terrain or conquering some territory from a neighbouring civ. This mechanic from the base Civ4 game is conceptually flawed, and in RI it feels even worse. Unfortunately it's not just a matter of tweaking a few numbers. Reducing drastically the number-of-cities maintenance would give too much of an advantage to expansionist strategies and make them too easy once the ball starts rolling. But introducing new mechanics is much more involved than simply tweaking numbers.
- AIs are really stupid when evaluating Open Borders. With Open Border tech transfer, there is limited interest in giving Open Borders to someone who is really behind in techs, but there is a strong interest in keeping Open Borders with a civilization that's ahead, as it propells a civ's own tech research. Very early in the game, civilizations are giving Open Borders relatively easily, but then they seem to be completely set against it. In my first game, I had cancelled most of my open border agreements when I got so ahead in techs that I couldn't get benefit from it in this regard, but later on as I expanded I wanted to make some deals to increase my trade route revenue. For dozens of turns, multiple civilizations that had diplomatic relations with me in the +6 to +10 range (thanks to shared religion) repeatedly refused signing Open Borders, despite the massive tech boon this would have been for them (compared to the mild benefit I would have received). The only way I managed to sign some deals was when I joined a war alongside some of these civs after they demanded me too.
- The "You are overtaking us" mechanic feeds into this. I had a rival civilization cut Open Borders and fall behind in tech as a consequence.
- AIs have a nasty tendency to just make stupid diplomatic demands, such as joining a war (but without paying anything to do so or having existing deep ties with the demanding civilization), converting to a religion that's present in maybe 2 cities out of 20, changing civics, and so on. Then they get upset when the obviously stupid demand is rejected. On the other hand, they almost never seem to come to propose a mutually beneficial trade deal.
- AIs have a nasty tendency to declare war to civilizations that are dozens of tiles away. This was especially obvious in the Europe scenario. Getting embroiled in useless wars sets the AI behind a lot. A "We are far from each other" diplomatic relation bonus ought to be considered. Or some way to reduce typical negative modifiers such as wrong religion when the countries are distant.
- The way AI considers civilization currently at war with its target to evaluate the military balance is deeply problematic, as it will often be possible for the target to achieve a separate peace and then the opportunistic declaration of war appears really foolish. In my second game, I had Aztecs in the South and Poles in the North, and after Aztecs attacked me with a proper army, Poland jumped in. But they never sent a proper attacking army my way (only 3 skirmishers), and after I repelled the Aztec assault and made peace with them, I could easily turn back on Poland and took two cities with very little resistance.
- Inflation is really annoying, and there is nothing realistic or engaing about it. It's just there because the gold revenue from normal tech progression and city expansion increases too quickly compared to the costs. If we take the example of a hypothetical stagnant civilization (no new tech, city expansion, workers improving tiles...), revenue and expenses should be flat over time, but with inflation the expenses explode. Just as with maintenance costs however, properly designing and balancing an alternative method would be quite involved, although both could be addressed at once. I think the keys are reintroducing more building-maintenance like in Civ3 (I noticed that the cathedrals have a 1-gold maintenance cost so it's feasible, although if it applies before percentage modifiers such as the tax office, that's not ideal), having separatism/revenue penalties for peripherical cities, and having higher-tech units incur a higher maintenance cost (is that feasible? Also an additional maintenance cost per unit type would be interesting).
- The loss and capture of cities causing destruction and setbacks is expected, but the damage tend to be so high that losing and then reconquering a city feels really bad. This problem is perhaps even worse in vanilla Civ4, but at least for me it fosters a mentality of not letting even one city fall, unless it was founded quite recently.
- I noticed earlier in this thread that the difference between the displayed epidemic rate and the real epidemic rate was discussed. I found it very puzzling when playing, because in my first game I went to great effort to keep epidemic rate low but despite having it often at 2-3% in a few cities, I only got my first epidemic after a few hundreds of turn. I was even asking myself if the feature was not broken.
- I haven't analyzed it in detail, but some leader traits seem so much better than others. Despite all the many sub-bonuses given by a trait like Conqueror, it still doesn't hold a candle to a trait like Legislator. Clear-cut economic bonuses are helpful for every single strategy and in every single situation, whereas some other traits, and especially military traits, have a very narrow niche. Since economic bonuses also make it possible to afford more units in the first place, I think some significant changes are warranted.
- The Folwark is OP. Most custom tile improvements are very niche, only being available on a small number of tiles, but here it's a strong +1F/+2G on half the tiles until mechanized farms become good enough to beat it.
- I don't like the penalty for resarching techs of the next Age before a certain date, especially as going into the next age earlier already forces you to deal with a +1/+1 unhealthyness/unhappyness penalty. Thankfully I could simply switch it off for my second game with the tick of a box, so I don't have anything to complain about.
- Random map generation for RI is a whole can of worms by itself. Some resources are so important that a map that doesn't place enough of them is going to be very unbalanced (which is made even worse by how reluctant AIs tend to be with resource trading and how little excess resources available for trade there is). In my second game, it didn't feel good to have a "thermal baths" unique tile improvement that I couldn't build even once within a 20 tiles radius around my capital...

I also wanted to ask: are there some ways to have AIs play by themselves and watch what they are doing? I would find it very interesting to make some AI tweaks and look at the results, or simply to observe how different civilization and traits fare in different games. Although I also need to learn how to compile the .dll and I don't know the codebase, I shouldn't have much trouble tweaking the C++ code.

Love your input, especially on the teching, great people and world wonders, which I personally like to get second-hand instead of building them, on a higher difficulty like immortal. Can you give me a rough outline of your early game (ancient and classical) regarding city numbers, tech goals and usual building queue? I find it sadly pretty hard to keep ahead of the AI in tech in order to build the WWs, without being randomly DoW'ed by each of the AIs on the continent, sometimes - as you write as well - quite irritatingly stupid by AIs about 25 turns away from my border. Do you have a few cities you just stack full of units to dissuade them and break out later by early medieval or flintlocks? I suspect I still need to use GP more.

On the city maintenance and inflation: Yeah, it seems unfair and unfitting, but its really needed to balance a longer game duration. You'd just have some civ snowball in classical even more than now. Gold also gets pretty plentiful later on, mainly by the ridiculous trading income and modifiers.

Imho Conqueror beats Legislator squarely, because Legislator doesn't do anything really until maybe late classical, with civic maintenance and cost reduction by courthouses being still rather low (though yes it's pretty nice later on). Conqueror gives you a free City Raider I, and Levies with City Raider II have shockingly good odds against archers on flat cities, gain CR3 in the process and do really well as fodder against hill cities. So focus on levies/shortswordmen and Conqueror nets you easily a second capital. I think both traits are in a good place, though in MP I'd say starting next to a Conqueror human player would be pretty unfair.

Regarding the fully built out cities, idk why not build even more units against the certain surprise attacks by AIs or to just bully the next contender AI a bit? If you are stronger than them, why not preempt surprise attacks by going on the offensive instead? (Like a true Roman) K-Mod AI got me really paranoid, because so far I have only lost games because of enemy units numbers being quite a bit higher than we I have prepared for.

Yes folwarks are indeed strong. Sad that by Fintlock Rifles I have to switch out of Serfdom, because 18 dudes with 10 Str. that ignore building defenses potentially appearing anywhere are a bit too much to handle for me, if I like to use my main army for anything else including defense against another AI. Still Poland is one of my favourites.

 

[Roland Johansen]

- Number-of-cities maintenance costs are outrageous. I played both games with number-of-cities tech cost scaling, and I think it's a fine option by itself. Beyond the realism justifications that can be given (difficulties of coordinating research and spreading research across a vast empire), it helps to limit the snowball benefits of expansion. However, number-of-cities maintenance create situations where you can have completely unclaimed space fit for settling a new city, and still let it aside for dozens or even hundreds of turns because all your old cities are suddenly going to be losing much more money. So you just wait for the next tech that can improve tax revenue enough to justify the expense of settling open terrain or conquering some territory from a neighbouring civ. This mechanic from the base Civ4 game is conceptually flawed, and in RI it feels even worse. Unfortunately it's not just a matter of tweaking a few numbers. Reducing drastically the number-of-cities maintenance would give too much of an advantage to expansionist strategies and make them too easy once the ball starts rolling. But introducing new mechanics is much more involved than simply tweaking numbers.

Basic civilization IV has a number of cities maintenance that goes up till a certain maximum value (which depends on difficulty level). That means that city maintenance costs can get so high that they hurt early to mid game, but at some point you can develop your cities far enough that they can handle it. It means that you still shouldn't add a ton of small cities to your empire, but you can eventually make every city self-sufficient.

This mod has removed that upper limit to those costs, so they keep rising. On a larger map, you can build so many cities that even your most prosperous city in the modern age would go bankrupt.

To change this back, you need to do a dll change. I discussed it with Walter a few dozen pages back. The change is not so complicated, but I struggled a bit with compiling the dll. A search for city maintenance in this thread should find the discussion and may help you if you want to change it yourself.

- The "You are overtaking us" mechanic feeds into this. I had a rival civilization cut Open Borders and fall behind in tech as a consequence.

This was also heavily discussed a while back. It was modified quite a bit by Walter in a few svn updates. The diplomatic modifier should be more reasonable now.

- Inflation is really annoying, and there is nothing realistic or engaing about it. It's just there because the gold revenue from normal tech progression and city expansion increases too quickly compared to the costs. If we take the example of a hypothetical stagnant civilization (no new tech, city expansion, workers improving tiles...), revenue and expenses should be flat over time, but with inflation the expenses explode. Just as with maintenance costs however, properly designing and balancing an alternative method would be quite involved, although both could be addressed at once. I think the keys are reintroducing more building-maintenance like in Civ3 (I noticed that the cathedrals have a 1-gold maintenance cost so it's feasible, although if it applies before percentage modifiers such as the tax office, that's not ideal), having separatism/revenue penalties for peripherical cities, and having higher-tech units incur a higher maintenance cost (is that feasible? Also an additional maintenance cost per unit type would be interesting).

Inflation has nothing to do with the real world notion of inflation. It is just a mechanical cost increase over time.

You could introduced building costs, but I don't know if the AI considers those when deciding to construct a building.

Higher maintenance costs already exist for more modern units i this mod. I have increased unit costs and done away with inflation. This gives me costs that are more related to gameplay and less to 'time passing'. You can do a replace all in some xml files, but if you have some very detailed cost progression in mind for units, then it will take more work.

- The loss and capture of cities causing destruction and setbacks is expected, but the damage tend to be so high that losing and then reconquering a city feels really bad. This problem is perhaps even worse in vanilla Civ4, but at least for me it fosters a mentality of not letting even one city fall, unless it was founded quite recently.

I personally disable city razing for this reason as I think it is weird to see cities being completely razed after the classical era. But cities still do indeed get a lot of damage by being captured. Many buildings have a chance of being destroyed, which you could change, I think... Or was that just wonders... I don't know. I never had such a problem with cities getting damaged by war, so I haven't looked into it.

- I noticed earlier in this thread that the difference between the displayed epidemic rate and the real epidemic rate was discussed. I found it very puzzling when playing, because in my first game I went to great effort to keep epidemic rate low but despite having it often at 2-3% in a few cities, I only got my first epidemic after a few hundreds of turn. I was even asking myself if the feature was not broken.

Yeah, I noticed a while back that the chance for epidemics was incorrectly shown in cities and looked up the chance calculation in the dll.That's the discussion with Walter about it that you probably saw. Walter changed the display in a recent svn update and now the displayed chances are correct (and a lot lower) per level of epidemic.

 

[AllTheLand]

Thanks @Takofloppa and @sazhdapec for the tips regarding AI autoplay. I'll check it out.

Some further general feedback I forgot the first time around:
- Conquered cities have a separatism modifier to avoid them revolting away immediately. I think newly settled cities should also get one, as it's not rare for them to suffer from unhealthiness or other troubles early on before you can build a few basic things to allow them to benefit from your resources. Since I often raze AI cities to build "better placed" cities instead, the difference here was clear.
- The barbarian civs settling down are way too advanced technologically. They would be expected to be slightly behind civilizations that have been around for much longer, but instead a civilization that settled from barbarians in my second game ended up being the worldwide tech leader for a very long time (until it got destroyed in war by its neighbours). I caught a glimpse of this being discussed in this thread recently, and I would definitely like to see some change here.

Aztecs just asked Poland to join a war. And they obliged.

Considering how often the AI spams such demand, I can buy it. But the AI accepting, and then not doing anything to actually fight the war, was really foolish. It's all the more ridiculous considering that my civ was the only land neighbour Poland had, and I have the "AI plays to win" option ticked on in that second game (not that I really noticed any difference from having it on, but I don't have enough experience with the mod for that).

It IS strong but Serfdom has a critical flaw: low production from craftsmen. Industrial era buildings and especially wonders are so expensive that it's painfully slow to build them. I'm currently playing as Poland and in the early industrial era I just had to switch to a more modern civic to get more hammers.

I have not played any game that far out so it's hard for me to tell. Still, having most of the Middle Ages and Renaissance with a massive food and gold boost is quite significant. Even without Serfdom, the Folwark is still +1/+1 over a standard farm is still very appreciable.

On the city maintenance and inflation: Yeah, it seems unfair and unfitting, but its really needed to balance a longer game duration. You'd just have some civ snowball in classical even more than now. Gold also gets pretty plentiful later on, mainly by the ridiculous trading income and modifiers.

Yeah I absolutely get that there needs to be increased sources of spending to counterbalance increased revenue, and that there needs to be measures to limit the effectiveness of territorial expansion. I just dislike the incentive pattern that's created by the way the mechanics we currently have work. Sitting doing nothing to avoid upkeep costs exploding is not very fun, and I find it to also go against immersion. That's why I'm thinking of mechanics such as decreased revenue from distant cities (possibly tied in with separatism mechanics), increased upkeep of more advanced military units (if that's possible), building upkeep costs... to avoid players using a 100% science rate and still having a gold benefit. Obviously the design, balancing and testing would not be easy.

Love your input, especially on the teching, great people and world wonders, which I personally like to get second-hand instead of building them, on a higher difficulty like immortal. Can you give me a rough outline of your early game (ancient and classical) regarding city numbers, tech goals and usual building queue? I find it sadly pretty hard to keep ahead of the AI in tech in order to build the WWs, without being randomly DoW'ed by each of the AIs on the continent, sometimes - as you write as well - quite irritatingly stupid by AIs about 25 turns away from my border. Do you have a few cities you just stack full of units to dissuade them and break out later by early medieval or flintlocks? I suspect I still need to use GP more.

As I said, this second game I played is not representative of a full random game because I picked AI leaders with the wonder building penalty. Although later on with barbarian settling into new civs, new AIs had no restriction on this.

Since I only played two games, I don't have any well-defined strategy, and I did some reload in my latest game when I noticed I had chosen the wrong tech path. I will say however that any start without limestone, wood and marble accessible in the first "ring" of cities around your capital is more or less "unplayable" to compete for wonders. The set map for the Europe scenario is full of resources, but getting a good map from the random map generators was difficult. The production speed boosts of limestone and marble are mandatory to get wonders in the ancient times, and the building enabled by wood is such a cheap and good boost to production that you really want your civilization to have access to it.

You cannot really be ahead in tech in the ancient eras compared to the AI in high difficulties (I still am not ahead in my Immortal game despite having had an obscene GDP advantage for a very long time and having been first for a few techs), but if your capital city has a solid hammer output, you can often afford to discover the tech 10 or 15 turns later and still get the wonder. Trying to research things first leads to you falling behind much more easily than if you take heavy advantage of the Open Borders tech boost. Obviously, military techs have to be underprioritized and any tech that gives money has to be prioritized.

Of course, if you get 3 AI civs declaring war on you, these plans are going to go wrong. There is a lot of risk early on.

I think that going without Great Wonders at all can just as well lead to very dominant positions, it's just that I like having Great People and you basically have to do wonders to get good amount of those. And when you reach the "all the cities have all the possible buildings" stage, having active great wonders and bonus from great people gives a higher ceiling to your civ's output.

Imho Conqueror beats Legislator squarely, because Legislator doesn't do anything really until maybe late classical, with civic maintenance and cost reduction by courthouses being still rather low (though yes it's pretty nice later on). Conqueror gives you a free City Raider I, and Levies with City Raider II have shockingly good odds against archers on flat cities, gain CR3 in the process and do really well as fodder against hill cities. So focus on levies/shortswordmen and Conqueror nets you easily a second capital. I think both traits are in a good place, though in MP I'd say starting next to a Conqueror human player would be pretty unfair.

I can't speak too much about the very early game, yes conqueror very early on might make it easier to take out or cripple a neighbouring civ. But when there is a lot of empty land around, going into a war with another civ instead of expanding into empty land is just setting yourself behind in my opinion. In free-for-all scenarios with more than two competitors, it's rarely a good play to go for war if that's not required by immediate expansion plans unless you are trying to take advantage of a glaring weakness (such as a big tech advantage over the other civ).

Then once you have more than a few cities, there is little point in expanding more at all until your revenue increases if you don't want to cripple your research, so even if you have the military might to take cities, your are better off not doing it. The gold you save or gain from traits such as Legislator, Financial or others directly allows you to research techs faster, leading to more opportunities to fight with technologically superior units, and with a higher number of units. RI makes the number of units less overwhelming of an advantage thanks to the logistics system (even more so with Conqueror boosting rural logistics), but in the end if you have enough warm bodies to throw at the problem, it will work. With the increased cost of training units based on how many you already have, it becomes much more feasible in RI to just send units to their deaths instead of caring to minimize your losses. As long as you take care of veterans, green units can be sacrificed, in the end it's going to make the training of their replacements faster anyway.

Personally, I would add something like "+100% free unit upkeep" to traits like Conqueror to have some appreciable economic help too.

In my second game I'm in the year 650 and cutting civics expense by half would save me 150 gold a turn. That would allow me to turn my science slider up by 20 percentage points.

Regarding the fully built out cities, idk why not build even more units against the certain surprise attacks by AIs or to just bully the next contender AI a bit? If you are stronger than them, why not preempt surprise attacks by going on the offensive instead? (Like a true Roman) K-Mod AI got me really paranoid, because so far I have only lost games because of enemy units numbers being quite a bit higher than we I have prepared for.

If I build more units, I increase my units upkeep expenses (with inflation, each unit costs me about 1.4 gold per turn currently in my Germany game). If my cities just went into building units non-stop, I would completely collapse my economy and my teching rate. Also with separatism on, having a bigger army means increasing separatism...

As long as my border cities that the enemy might attack are somewhat defended, it's easy enough switch into war economy when attacked and have a flood of bodies stream towards the front. Surprise DoW in the antiquity era did catch me off-guard in my Poland game (Ukraine got angry I was "getting ahead") and I had to reload to prepare more for it. But while I'm greedy and under-do army early on, after having reached the late classical or early medieval era, it's not as much of a strain to keep delaying defenses ready. I should perhaps mention that I am playing my second game in 1.5 speed though (the first one was vanilla 1.0 speed), which means that producing new units to reinforce is somewhat faster. I also have paved roads everywhere.

The Aztecs have attacked me several times by now, and although some of their invasion columns were scary (especially when they had the 7 Str. warriors and even 8 Str. archers in there), in the end skirmishers and recruits could whittle them down and at some point I'd have enough units massed nearby to overwhelm them.

Even if I lost some peripheral cities early in the war, I have such a big advantage in hammer output over the Aztecs at this stage I could flood them.

I suppose I could have wars where the goal is not to conquer land, only to kill enemy units or raze some cities I don't intend to capture later on, but I don't feel playing as the USA.

Yes folwarks are indeed strong. Sad that by Fintlock Rifles I have to switch out of Serfdom, because 18 dudes with 10 Str. that ignore building defenses potentially appearing anywhere are a bit too much to handle for me, if I like to use my main army for anything else including defense against another AI. Still Poland is one of my favourites.

Yeah, I don't think it's good to have that strong rebel forces that risk randomly spawning in your empire. Although in the middle ages rebels have some upsides too, they give some easy great general points. The 6 Str. units caught me by surprise the first time around with Poland though, as they were stronger or similary as strong as all my normal units.

 

[AllTheLand]

This mod has removed that upper limit to those costs, so they keep rising. On a larger map, you can build so many cities that even your most prosperous city in the modern age would go bankrupt.

To change this back, you need to do a dll change. I discussed it with Walter a few dozen pages back. The change is not so complicated, but I struggled a bit with compiling the dll. A search for city maintenance in this thread should find the discussion and may help you if you want to change it yourself.

I think that removing that upper limit was a wrong design decision. But I'm not just concerned about what happens when you get 50 cities in the Renaissance or 100 cities in the industrial era, I'm also concerned about having empty land ripe for the taking or having a strong enough military to conquer from a neighbor, and instead having to just stay idle and wait to avoid collapsing the economy. Even with 6 or 7 cities, adding a new city too early can be very damaging.

I also find it unfortunate that road infrastructure has no effect on city distance maintenance, but I can understand that it would be difficult to have it work properly.

Inflation has nothing to do with the real world notion of inflation. It is just a mechanical cost increase over time.

Yeah, I know, which is part of why I dislike this mechanic. It's completely artificial.

You could introduced building costs, but I don't know if the AI considers those when deciding to construct a building.

Higher maintenance costs already exist for more modern units i this mod. I have increased unit costs and done away with inflation. This gives me costs that are more related to gameplay and less to 'time passing'. You can do a replace all in some xml files, but if you have some very detailed cost progression in mind for units, then it will take more work.

Oh, higher maintenance cost for more modern units are already possible? That's great to hear. I think it doesn't show in the civilopedia however, which is disappointing.

"Costs that are more related to gameplay and less to 'time passing'" are what I'm after.

If you already have something like this working for you, could you perhaps share it somewhere? It seems we have some shared taste in this regard so I'd rather not reinvent the wheel. Has Walter expressed any opinion about the general concept? I imagine there is concern for any large-scale change that could affect balance, but would there be some potential for some variation of this concept making its way into RI?

As for the AI and building costs, I suppose that would have to be checked in the code or tested... Cathedrals already have a building cost, but it's a single building and it's small enough that the AI ignoring completely this cost wouldn't be noticeable.

But cities still do indeed get a lot of damage by being captured. Many buildings have a chance of being destroyed, which you could change, I think... Or was that just wonders... I don't know. I never had such a problem with cities getting damaged by war, so I haven't looked into it.

I don't usually suffer from this problem either, but that's mostly because I will go to great lengths to not have to deal with it. Having some back and forth war where you lose a couple cities and then regain them could be quite interesting if it didn't leave them so devastated.

This is even more true of separatism. I don't like how little warning there is after the initial message giving a choice of action to limit the rebellion for a city to breakaway. But it would generally be a lot more tolerable to have parts of your empire breaking away and having to reintegrate them back in war if the setback wasn't as long-lasting.

I have noticed that this mod uses temporary buildings to create special effects, such as the "lenient governor" that reduces separatism, hammer output and tax revenue. I wonder if "damaged [building]" temporary buildings that would give a boost to build back the corresponding building and would disappear when it is rebuilt would be feasible. That could be a great way to have cities suffer war damage while not making it too crippling

Yeah, I noticed a while back that the chance for epidemics was incorrectly shown in cities and looked up the chance calculation in the dll.That's the discussion with Walter about it that you probably saw. Walter changed the display in a recent svn update and now the displayed chances are correct (and a lot lower) per level of epidemic.

That's great to hear. I'll definitely pick up the SVN version whenever I start a new game then.

 

[Watermelon]

I can't speak too much about the very early game, yes conqueror very early on might make it easier to take out or cripple a neighbouring civ. But when there is a lot of empty land around, going into a war with another civ instead of expanding into empty land is just setting yourself behind in my opinion. In free-for-all scenarios with more than two competitors, it's rarely a good play to go for war if that's not required by immediate expansion plans unless you are trying to take advantage of a glaring weakness (such as a big tech advantage over the other civ).

Yeah, not long before I also didn't really war early unless there was pressing reason to (maybe boxed in with 3 cities or single AI competitor on the continent). But I think good settle spots in the early game are really rare. Killing an AI neighbour early on can really give you breathing room and spare you a lot of trouble later down the line. Take my current Germany game for example with Bismarck on huge RI Planet generator:

I saw an opportunity in taking Rome and razing another city, with a couple of shortswordmen and militia. That netted me a great capital location with 2 currently Pastoral Nomanism boosted cows, as well as gold and spices and lots of river tiles. A good capital location, certainly, but not uncommon and very worth it, because it basically eliminated my nothern rival Rome, and still kept a single city north to zone the Spanish, while I can take his last decent city at my leisure.:

You might notice the Spanish who settled west of me, but it's no deal, because this capital location projects power, enabling me in a short while to train levies rather quickly. Toast the spanish town when I decide to. You see all this empty land on the map? Easy expansion and boom right? No. Because right now I'll queue up soldiers to gather another capital:

Spoiler Next item on the shopping list :

A piece of cake compared to Rome, a hill city with walls. Also has nice pastures, nice river tiles and couple of hill tiles.
Why not just tech up, build up some proper infrastructure? Well because there is currently no tech available cheaply that I really want. Who cares about Priestdom, when I have Animism due to Jungles (even though I founded Judaism)? Mining's no good without slave farms, which I don't need due to my low happy cap and many pastures. Drama's execellent (most of the time my breaker tech into classical, because it offsets the malus in happy faces), but crazy expensive atm, with ahead of time and no AI having it.

Why not settle all the emtpy space instead (ignoring the "dead rival on my border is a no city spamming rival on my border")? Well..:

Spoiler All the possible city locations, some even decent :

The Austronesians annoyed me here, because they blocked the southern locations with horses and jungle pigs, freshwater and coast, which I would certainly settle. Top left is bad, because little food and only one flat river tile for farm. Right is bad atm because no farms, lack of fresh water and no more food that the one banana.

Dye location is "okay", but riverless jungle sucks esp. when switching out of animism, so low early food. Below I lack fresh waters and rives, its too little coin for me to settle now.

Red location is also "fine", but single jungle river tile is not enough with other features lacking and barb town next door. Doesnt really gain me anything, and I can "zone" areas just fine for later by usage of fire and sword. Green is fine as well. But no freshwater hill city on border to Egypt, who can actually threaten me later on (because I took out Rome, who was their rival as well) and Egypt UU is strong enough. Culture could also be problematic, barb town still there. Location not immediately imba enough to be worth settling and garrisoning.

Damn the Spanish for settling there, right? Nah, triple food and copper looks really good, but corn and rice are dry, nerfed by Pastoralism (which I'm gonna keep atleast for now) and no hill, no fresh water, no rivers. From my experience such cities look promising but then starve on pop 4 or so, with only a couple of hammers to show for it. West to it might be a decent location, depending what hides in the fog. Most likely though, no big rivers and a sea food or sea commerce which wouldn't make it a priority still. (Though probably the next settle, if there is a small river or lake somewhere)

Tbh I struggled quite a bit to get Rome, because they build a wall like 3 turns before my planned attack. Luckily they sent many soldiers on an attack against Egypt and away with a settler, about 4 units manned the walls when I came knocking without rams. With conqueror it would be quite a bit easier to pull of. I think many more players might be surprise of their gains and the accessibilty of early conquest with shortswordmen. (Take a look at the treasure, you get quite a bit of cash for your troubles, so you don't lag behind in tech too much as you can see)

But hearing you had 150 gpt saved through Legislator in 650, which could be spent into science was quite amazing, and is a very strong boost! How many cities did you have at that moment?

I think it would be fun, if we had kind of a standard map in this thread. So players could try out different strategies, compare them with each other and learn new tricks watching other peoples saves. I think the regular Civ 4 BTS have this as hall of fame games?

 

[Walter Hawkwood]

Happy to share my edits if anyone's curious about how to go about modifying the map generation script towards their own preferences.

Please do, and not just edits themselves, but also insights into what particular ones do. Quite often when I tried messing with map scripts, the effect of a certain change was far less than I would imagine.

Keep it simple and you get better gameplay.

Ultimately yes, I wouldn't want to overcomplicate stuff. But I don't think players would get too confused by a wonder or a leader trait that boosted tech transfer, or a wonder that eliminated one from your civ. What I want is simple but impactful changes with more player agency.

-Turkish Janissaries certainly come too late with Absolutism, which is on the same column as flintlock musket, while Janissaries only have 10 Str. Historically these guys appeared much earlier, and adopted firearms once they become widespread. Maybe just make them available at Arquebus like the Tercio (Spain and Ottomans being quite the rivals at the time as well)? Or at the latest Administration, which was kind of an Ottoman thing anyways.

Agreed. It's one of those where I look at something and wonder why it is like that in the first place. Due to the age of the mod, it might well be from a time where the tech tree layout was quite different. Just Arquebus it is now!

-Spanish Mineria just feels sad. It gives one commerce more on an improvement you might get 2 in your entire empire. If we would reeeallly stretch the definition of serfdom, it might be fitting to give the spanish Minerias about 2 Commerce extra under Slavery (Classical Spanish Silver mines) and Serfdom (New World silver mines), because unlike Protectionism you are going to be in these civics for a while.

I kind of disagree here. While it is not among the most awe-inspiring improvements, I'd say one can reasonably spend the latter half of the game in Protectionism. I may buff it even more though.

I played around with RI recently. I had an old 3.5x version installed previously but I had graphical issues, so I only started with 3.60. There has been some significant changes since that release, but since I had already started games and the saves are not compatible, I didn't jump to 3.61 or the SVN version yet. But I think most of my experience on 3.60 remains relevant.

Thanks for the detailed feedback! It is always welcome.

the Civ3 music in the RI Classical era feels great

IIRC it was from 0AD, and its lack will probably be one of the negatives for you switching to a later version.

There are some elements I liked in AdvCiv I tend to miss in RI, notably the AI being smarter about war declarations and the way city revolts work.

I have great admiration for AdvCiv, and ported quite a few smaller things from it. Unfortunately, many of the more involved features are extremely complex code-wise and hard to port.

World Wonders are somewhat too important. This is perhaps subjective, but some of them give extremely strong bonuses and they are the dominant source of GP points for the early eras, with GP points being quite valuable. One of the clearest example is the Kreml. The -10% maintenance bonus across all cities is by itself overpowered (It's almost effectively -20% once you have both a courtroom and a town hall in a city), and it also gives GP points and if you get Servage it's +1 hammer / + 1 gold per city with a building that's very easy to get.

I mean, isn't it one of the leitmotifs of Civilization game series as a whole? But yeah, I agree with you with some caveats.

1) You wrote yourself that you hand-picked opponents with wonder construction penalties and optimized your start location for the needed resources. I'd say in an average game where one player doesn't get to hoard wonders, their effects are more balanced.
2) Kremlin is indeed rather powerful, if situationally so, but Poland you picked is an absolute no-brainer for it; this is probably one of the strongest civ/wonder synergies in game.

The Heroic Epic that gives +100% GP points gives a massive incentive to just put every single GP-giving wonder in the same city. In my first game, I tried building a few wonders in a side city, but its GP generation was so behind my capital that it managed to produce a GP only once (I ended the game into the Middle Ages). The tradeoff between losing some GP generation and increased control of the GP type that gets created is simply not there. The no limit on national wonders per city, which I don't mind by itself, makes that choice even more obvious.

Well, it's a vanilla thing. If anything, the presence of additional modifiers in RI should dampen its effect (+150% in the capital vs +50% in all cities is less of a difference than +100% in the capital vs +0% elsewhere). I don't actually have a definite feeling on that one. To me it feels like from certain point onwards (around late Renaissance), other cities are able to catch up in most cases (again, though, I have to point out that a specific "hoard all the wonders in a perfect capital" playstyle can severely impact this). I think (but haven't checked) the GP point curve plateaus at some point; that would make having several cities generating a lot of GP point more effective in the long run.

Despite the high number of buildings available to each city, it's quite possible to reach a situation where most of the cities (except some more recent ones) has all the helpful buildings, and nothing really useful to do as the Civ has already enough military units. I had almost never used the "convert hammers to gold/science" options in previous games, I often had half my cities or more using it here. I think there are benefits in not needing half the game to get a city up to standards (and the "old building upgrades into new building" trick, such as Well into Aqueduct is really clever), but reaching a point where most cities don't have much useful to do with their hammers seems strange. It certainly taught me the value of building cottages, because hammer-prioritization doesn't seem that useful when the limiting factor is gold instead. In AdvCiv I could mostly focus on farms, mines and sawmills, not anymore.

I know what you're talking about, even though I've been seeing less of it in the later versions. Maybe raising the costs of everything across the board would be a good way to go. Oh and farms are still subjectively a better choice later in game IMO, as you can support several specialists from one farm.

Number-of-cities maintenance costs are outrageous. I played both games with number-of-cities tech cost scaling, and I think it's a fine option by itself. Beyond the realism justifications that can be given (difficulties of coordinating research and spreading research across a vast empire), it helps to limit the snowball benefits of expansion. However, number-of-cities maintenance create situations where you can have completely unclaimed space fit for settling a new city, and still let it aside for dozens or even hundreds of turns because all your old cities are suddenly going to be losing much more money. So you just wait for the next tech that can improve tax revenue enough to justify the expense of settling open terrain or conquering some territory from a neighbouring civ. This mechanic from the base Civ4 game is conceptually flawed, and in RI it feels even worse. Unfortunately it's not just a matter of tweaking a few numbers. Reducing drastically the number-of-cities maintenance would give too much of an advantage to expansionist strategies and make them too easy once the ball starts rolling. But introducing new mechanics is much more involved than simply tweaking numbers.

I mean, for me it's the other way round. "Expansion is a no-brainer" approach where a bigger civ is almost always better was always a turn-off. To be fair, a tall-vs-wide balance in the 4X genre is a very tricky thing, and the Civilization series itself also seems to flip-flop on that historically.

AIs are really stupid when evaluating Open Borders. With Open Border tech transfer, there is limited interest in giving Open Borders to someone who is really behind in techs, but there is a strong interest in keeping Open Borders with a civilization that's ahead, as it propells a civ's own tech research. Very early in the game, civilizations are giving Open Borders relatively easily, but then they seem to be completely set against it. In my first game, I had cancelled most of my open border agreements when I got so ahead in techs that I couldn't get benefit from it in this regard, but later on as I expanded I wanted to make some deals to increase my trade route revenue. For dozens of turns, multiple civilizations that had diplomatic relations with me in the +6 to +10 range (thanks to shared religion) repeatedly refused signing Open Borders, despite the massive tech boon this would have been for them (compared to the mild benefit I would have received). The only way I managed to sign some deals was when I joined a war alongside some of these civs after they demanded me too.

I don't recall if it was before or after the version you played, but there was a major bug I fixed where most of the code for evaluating Open Borders simply didn't get called. So unfortunately I can't say whether you're criticizing the brain-dead version or the version after I already fixed it.

The "You are overtaking us" mechanic feeds into this. I had a rival civilization cut Open Borders and fall behind in tech as a consequence.

It was both toned down and the formula for it rewritten to be more gradual. Should be much better now.

AIs have a nasty tendency to just make stupid diplomatic demands, such as joining a war (but without paying anything to do so or having existing deep ties with the demanding civilization), converting to a religion that's present in maybe 2 cities out of 20, changing civics, and so on. Then they get upset when the obviously stupid demand is rejected. On the other hand, they almost never seem to come to propose a mutually beneficial trade deal.

Well, that's Civ 4 AI in a nutshell. Not much (if anything) changed there.

AIs have a nasty tendency to declare war to civilizations that are dozens of tiles away. This was especially obvious in the Europe scenario. Getting embroiled in useless wars sets the AI behind a lot. A "We are far from each other" diplomatic relation bonus ought to be considered. Or some way to reduce typical negative modifiers such as wrong religion when the countries are distant.

That one is in my sights, but I haven't done anything particular with it yet. A relation bonus seems unnecessary, but an additional factor in war preparation calculations might be in order.

The way AI considers civilization currently at war with its target to evaluate the military balance is deeply problematic, as it will often be possible for the target to achieve a separate peace and then the opportunistic declaration of war appears really foolish. In my second game, I had Aztecs in the South and Poles in the North, and after Aztecs attacked me with a proper army, Poland jumped in. But they never sent a proper attacking army my way (only 3 skirmishers), and after I repelled the Aztec assault and made peace with them, I could easily turn back on Poland and took two cities with very little resistance.

Again, that's a vanilla thing we didn't touch. And vanilla is actually quite stupid in this regard, as the civ that accepts an invitation to war doesn't have a war plan in place. Would make more sense if it first spent several turns getting its troops in place. But I'm not coding that...

Inflation is really annoying, and there is nothing realistic or engaing about it. It's just there because the gold revenue from normal tech progression and city expansion increases too quickly compared to the costs. If we take the example of a hypothetical stagnant civilization (no new tech, city expansion, workers improving tiles...), revenue and expenses should be flat over time, but with inflation the expenses explode.

I don't really get the hate. This is barely a mechanic at all, as players can't interact with it - just a simple modifier that ticks linearly over time. But you'll be glad to know that recently I cut away a chunk of K-mod inflation-related code which should result in lower inflation in most games.

The loss and capture of cities causing destruction and setbacks is expected, but the damage tend to be so high that losing and then reconquering a city feels really bad. This problem is perhaps even worse in vanilla Civ4, but at least for me it fosters a mentality of not letting even one city fall, unless it was founded quite recently.

Sort of. You yourself mentioned that cities can be built up relatively quickly, and later in game ministries help out a lot. But I agree there is, especially later on, a clear distinction between your core cities, losing which hurts a lot, and the periphery that is more extractive and less productive, which can be lost and regained with relative ease.

I noticed earlier in this thread that the difference between the displayed epidemic rate and the real epidemic rate was discussed. I found it very puzzling when playing, because in my first game I went to great effort to keep epidemic rate low but despite having it often at 2-3% in a few cities, I only got my first epidemic after a few hundreds of turn. I was even asking myself if the feature was not broken.

Yeah, that was not quite broken, but very misleading, using "%" when what displayed wasn't actual probability per turn. The formula is non-linear, meaning 1-2 epidemic strength is a very low outbreak probability. Now there is a better tooltip.

I haven't analyzed it in detail, but some leader traits seem so much better than others. Despite all the many sub-bonuses given by a trait like Conqueror, it still doesn't hold a candle to a trait like Legislator. Clear-cut economic bonuses are helpful for every single strategy and in every single situation, whereas some other traits, and especially military traits, have a very narrow niche. Since economic bonuses also make it possible to afford more units in the first place, I think some significant changes are warranted.

I feel there is a lot of difference in utility between individual playstyles, and I've seen people with vastly different ones. From your observations, I'd place you firmly into the economic player camp, for whom military-related traits are relatively worthless; but for some other players they can make or break a game.

The Folwark is OP. Most custom tile improvements are very niche, only being available on a small number of tiles, but here it's a strong +1F/+2G on half the tiles until mechanized farms become good enough to beat it.

I agree it's one of the stronger ones. Though I can't say I see Poland overperforming in AI-only test games. There are some civs that do, sometimes mysteriously so (I don't really know what Armenia's deal is, but I almost never see a weak Armenia).

I don't like the penalty for resarching techs of the next Age before a certain date, especially as going into the next age earlier already forces you to deal with a +1/+1 unhealthyness/unhappyness penalty. Thankfully I could simply switch it off for my second game with the tick of a box, so I don't have anything to complain about.

Eh, I am not too partial to it myself, but there used to be a very vocal player subset that demanded research speed to match the displayed date. Also, due to the non-linear tech cost increases across the tech tree, the actual impact on the progress speed is actually smaller than one feels it should be at the first glance.

Random map generation for RI is a whole can of worms by itself. Some resources are so important that a map that doesn't place enough of them is going to be very unbalanced (which is made even worse by how reluctant AIs tend to be with resource trading and how little excess resources available for trade there is). In my second game, it didn't feel good to have a "thermal baths" unique tile improvement that I couldn't build even once within a 20 tiles radius around my capital...

I can't really remember when, but I overhauled the resource placement at some point. Again, it may or may not be the version you've played.

I also wanted to ask: are there some ways to have AIs play by themselves and watch what they are doing? I would find it very interesting to make some AI tweaks and look at the results, or simply to observe how different civilization and traits fare in different games. Although I also need to learn how to compile the .dll and I don't know the codebase, I shouldn't have much trouble tweaking the C++ code.

As others already pointed out, RI has AI autoplay. It gets disabled for the release versions though, so you'll have to use SVN for that.

- find CivilizationIV.ini in My Documents > My Games > beyond the sword, find "CheatCode" and set it to "chipotle". This will allow you to use Ctrl+Z shortcut which reveals the whole map so you can see how everyone is playing
- press Ctrl+Shift+X (I hope I remember correctly) and enter how many turns you want to autoplay

It can't be stopped as far as I know, so it's better to run it in some reasonable intervals.

It actually can be stopped with the same key combo, but as the game is fairly unresponsive during autoplay, you often need to bash it repeatedly several times for it to stop.

To change this back, you need to do a dll change. I discussed it with Walter a few dozen pages back. The change is not so complicated, but I struggled a bit with compiling the dll. A search for city maintenance in this thread should find the discussion and may help you if you want to change it yourself.

I am actually curious - from the time you made the change, did you ever reach 40+ cities? Because from my experience, the economy is balanced to work and works as intended at least until that amount. I often see AI civs winning the game have 30+ cities.

Conquered cities have a separatism modifier to avoid them revolting away immediately. I think newly settled cities should also get one, as it's not rare for them to suffer from unhealthiness or other troubles early on before you can build a few basic things to allow them to benefit from your resources. Since I often raze AI cities to build "better placed" cities instead, the difference here was clear.

Actually makes sense, given that culture doesn't go away totally on razing a city. I'll check.

The barbarian civs settling down are way too advanced technologically. They would be expected to be slightly behind civilizations that have been around for much longer, but instead a civilization that settled from barbarians in my second game ended up being the worldwide tech leader for a very long time (until it got destroyed in war by its neighbours). I caught a glimpse of this being discussed in this thread recently, and I would definitely like to see some change here.

Weeeell... A settling barbarian civ is already at a disadvantage when it comes to most other things, most crucially the quality of its territory. Starting spots for initial civs are not only picked from the best on the map, but also usually additionally sweetened, and then they organically expand to the best spots around them, whereas a settling barbarian civ is basically a collection of cities in random spots. If these emerging civs aren't given an edge somewhere, they'll just be fodder. Generally speaking, my headcanon for "settling barbarians" is not something like Mongols riding in from the steppes, but rather an organized society that was looked down upon by their more ancient neighbours but may actually turn out to be more dynamic and progressive than the already long-established ones, such as Greeks as seen by Egyptians.

I have not played any game that far out so it's hard for me to tell. Still, having most of the Middle Ages and Renaissance with a massive food and gold boost is quite significant. Even without Serfdom, the Folwark is still +1/+1 over a standard farm is still very appreciable.

Poland is one of the civs that has a very clearly defined "Golden Age" when it comes to gameplay. A large bonus during a limited time window might not necessarily be better than +1 of something for the majority of the game... I am not totally closed off to nerfing it, but Poland overall doesn't feel OP to me.

Yeah I absolutely get that there needs to be increased sources of spending to counterbalance increased revenue, and that there needs to be measures to limit the effectiveness of territorial expansion. I just dislike the incentive pattern that's created by the way the mechanics we currently have work. Sitting doing nothing to avoid upkeep costs exploding is not very fun, and I find it to also go against immersion. That's why I'm thinking of mechanics such as decreased revenue from distant cities (possibly tied in with separatism mechanics), increased upkeep of more advanced military units (if that's possible), building upkeep costs... to avoid players using a 100% science rate and still having a gold benefit. Obviously the design, balancing and testing would not be easy.

I feel "wider" civs still generally dominate the rest. Not so completely as in vanilla, but still.

But when there is a lot of empty land around, going into a war with another civ instead of expanding into empty land is just setting yourself behind in my opinion. In free-for-all scenarios with more than two competitors, it's rarely a good play to go for war if that's not required by immediate expansion plans unless you are trying to take advantage of a glaring weakness (such as a big tech advantage over the other civ).

I will disagree with you here. Starting spots for capitals are usually the best city spots there are in the world; and early on when your overall number of cities should stay low, the quality of individual city spots is very important. A foreign capital is often worth the investment in taking it.

Yeah, I don't think it's good to have that strong rebel forces that risk randomly spawning in your empire. Although in the middle ages rebels have some upsides too, they give some easy great general points. The 6 Str. units caught me by surprise the first time around with Poland though, as they were stronger or similary as strong as all my normal units.

Not stronger; they are as strong as the weakest unit (irregular) you can currently build. While the transition period immediately after flintlocks can be a bit painful, you're able to build/upgrade to very cheap units of your own at the same time if defenses anywhere seem to be lacking.

I think that removing that upper limit was a wrong design decision. But I'm not just concerned about what happens when you get 50 cities in the Renaissance or 100 cities in the industrial era, I'm also concerned about having empty land ripe for the taking or having a strong enough military to conquer from a neighbor, and instead having to just stay idle and wait to avoid collapsing the economy. Even with 6 or 7 cities, adding a new city too early can be very damaging.

Do you feel that settling / taking a city should always be a no-brainer net positive then? As it currently stands, there is a clear economic curve over time where expansion usually happens after several clearly defined jumps in economic efficiency, and yes, overextending beyond what you can support is painful. I feel it is a design decision that fosters more deliberation over where one wants to settle/expand to. To get a colonial empire going, one first has to have banks at home!

And by late industrial era, you can actually absorb a lot of new cities, which sets the stage neatly for World Wars.

 

[JohnCarterOfMars]

Some thoughts in keeping tech transfer "simple" but with "player agency:"

1. Must have open borders.
2. Bonuses for civ desiring tech
   1. Has trade routes
   2. Has same state religion
   3. Have some common civics
   4. In a vassal relationship
   5. Performing counter-espionage mission on civ with desired tech
   6. Civ with the desired tech is very far ahead of other civ
3. Penalties for civ desiring tech
   1. Had previously been at war with other civ
   2. Had been at war with friends of other civ
   3. Had denied requests from other civ
   4. Is enemy of other civ
   5. Distance between palaces

I'm leaving out personal traits as I think these will ultimately influence the items on my list.

 

[Watermelon]

I kind of disagree here. While it is not among the most awe-inspiring improvements, I'd say one can reasonably spend the latter half of the game in Protectionism. I may buff it even more though.

Huh, I probably find it weak, because I seldom use protectionism. Pretty much always in Free Market. Still I think it should be stronger in the early game as well, because many games are decided before protectionism can ever be adopted.

Again some other observations, that might or might not be shared by other players:

1. I find scouts too be a bit too squishy to scout reliably. I much prefer using a militia for scouting, who has much more staying power. I'd get a scout after Farming -> Animal Husbandry -> Tool working, so may it's intentional they get absolutely eaten by archers and barbs. But I think they do tend to die too often to a pair of cheetahs hidden in the fog, compared to using a militia. Even when staying on forests only. Maybe make them a bit more survivable against animals, but until them I find myself skipping scouts entirely.
2. On a similar note I often only get 2 horsemen (the 5-Str. guys), mostly because the first two are so cheap. They are mobile, but lack punching power against pretty much anything except a lone shortsword man or recon unit in the open, or to raid improvments while advancing. They are fine, but I usually use skirms for these tasks, which they often perform better. Maybe give the horsemen some +25, 50% against archers? You don't have odds on a lone hill archer, which feels odd for these fast horsemen. Horse archers on the other hand are an instant recruit, when they become available.
3. Fortifications don't provide protection against gunpowder units. This is surely not intended, because star forts caused some of the longest sieges in history. I think there might be two types of defense: the one ineffective against gunpowder units, like walls, and the one which provides defense in general like (IIRC) Chichen Itza, and the wrong one is registered on post gun powder defense buildings.

 

[Roland Johansen]

Oh, higher maintenance cost for more modern units are already possible? That's great to hear. I think it doesn't show in the civilopedia however, which is disappointing.

"Costs that are more related to gameplay and less to 'time passing'" are what I'm after.

If you already have something like this working for you, could you perhaps share it somewhere? It seems we have some shared taste in this regard so I'd rather not reinvent the wheel. Has Walter expressed any opinion about the general concept? I imagine there is concern for any large-scale change that could affect balance, but would there be some potential for some variation of this concept making its way into RI?

As for the AI and building costs, I suppose that would have to be checked in the code or tested... Cathedrals already have a building cost, but it's a single building and it's small enough that the AI ignoring completely this cost wouldn't be noticeable.

I am not sure what you are looking for. Removing inflation is a simple xml change in the file CIV4GameSpeedInfo.xml in the mod folder where you set (for all the game speeds) the value iInflationPercent to 0. It will affect the game that you are playing (backup the old file).

What I did on unit costs is double the value of iUnitCostPercent for every difficulty level (the AI plays at noble) in the file CIV4HandicapInfo.xml. But that is a dirty way to do it as I just double the costs of all units. But it was a quick and easy 'proof of concept'. But what you'd really want to do is increase the costs of units dynamically over time. Maybe no extra costs in ancient time. Maybe 1 extra costs in medieval times and then increasing it more rapidly going into modern times. And given how many units that there are.... it is quite a bit of changing. You can see in the game that for instance, modern armor has a 2 extra cost. This originates in the file TR_Armor_CIV4UnitInfos.xml for the unit modern_armor in the value <iExtraCost>2</iExtraCost>. You'd have to change that for every unit in the game, also the local versions of every unit.

The change that I did for city maintenance is in the dll, but you don't seem to be asking about that one. For each svn version, I will need to create a new dll-file.

I don't usually suffer from this problem either, but that's mostly because I will go to great lengths to not have to deal with it. Having some back and forth war where you lose a couple cities and then regain them could be quite interesting if it didn't leave them so devastated.

This is even more true of separatism. I don't like how little warning there is after the initial message giving a choice of action to limit the rebellion for a city to breakaway. But it would generally be a lot more tolerable to have parts of your empire breaking away and having to reintegrate them back in war if the setback wasn't as long-lasting.

I have noticed that this mod uses temporary buildings to create special effects, such as the "lenient governor" that reduces separatism, hammer output and tax revenue. I wonder if "damaged [building]" temporary buildings that would give a boost to build back the corresponding building and would disappear when it is rebuilt would be feasible. That could be a great way to have cities suffer war damage while not making it too crippling

I think it is purely related to the value iConquestProb in the file CIV4BuildingInfos.xml. It is quite low for a lot of buildings. At a quick glance it is 0% for about every military building and 66% for many other buildings. I agree that this could be a lot higher. I guess it makes conquest less profitable. Ideally, it would be related to the number of units killed to take the city or something like that. A very old game like Master of Magic even made it related to the number of turns of fighting for the city (it has a separate turn-based combat screen where units fought each other).

Ultimately yes, I wouldn't want to overcomplicate stuff. But I don't think players would get too confused by a wonder or a leader trait that boosted tech transfer, or a wonder that eliminated one from your civ. What I want is simple but impactful changes with more player agency.

It mostly needs to be transparent and gameplay would indeed benefit from player agency. I liked the idea that the turns of peace/turns or open borders had an effect on tech transfer percentage. I could see people arguing to give a Great Library an effect. But when you look at it, many technologies were not transferred on paper, but just by craftsmen traveling. Historically, the library was good for maintaining written knowledge. Maybe the creation of modern nation states and Espionage Agencies could lower the percentage, when it became more 'us' versus 'them'. And computers or the internet could increase it again when it is harder to apply borders to knowledge.

But turns of open borders would for instance give more player agency as it means that there is a big incentive to keep long-term good relations with advanced civilizations.

I am actually curious - from the time you made the change, did you ever reach 40+ cities? Because from my experience, the economy is balanced to work and works as intended at least until that amount. I often see AI civs winning the game have 30+ cities.

I actually haven't played so much lately. I was involved in play testing a mod for a different game. But I am still very interested in this excellent mod for the brilliant civilization IV (and VII doesn't look so hot to me).
I looked it up. I have now 23 cities at the early medieval age. I haven't had a war yet. Just peaceful expansion. I was playing on a large map. I couldn't expand any further a few cities earlier with nice open spots left. I forgot whether I was playing emperor or immortal level.
I don't know what you mean with saying that the game should work till 40 cities. The number of cities maintenance costs go up with the number of cities and the size of the cities. So, with no upper bound of the costs, at some point, you won't be able to expand anymore, even with late game prosperous cities. I wouldn't be able to cover the continent that I started on with 40 cities. And that was like 1/3th of the world.

For me it is a different design philosophy.

• I don't mind that empires can get somewhat big already at the end of the classical age. It happened in the real world and some were fairly stable. At least, they didn't immediately crumble due to city maintenance costs. The revolution aspect is something separate from city maintenance and you can choose to play using that.
• The most artificial part of civilization is that there is one guiding light from 4000BC till 2000AD. That makes an empire a lot more stable. To mimic real transfer of power through all those generations requires a different game.
• I want the game to be able to be won by conquest victory. It is a victory condition. Given that I play on a large map, that could be like 120 cities or so? Not that I really expect to win by that victory condition, but I don't want to be limited. And I definitely don't want the AI to be limited.
• I want the fear of being conquered. Not just losing a border city. Not just by an early game rush. No, the fear of finding another continent completely conquered by another civilization that is three times my size and might be coming for me.
• I don't think that the AI actually really knows about the number of cities maintenance. It knows maybe that it shouldn't expand when it already has a high city maintenance. But is it really integral to all its decision making? Does a militaristic, aggressive leader understand that it is really useless to create a huge offensive army and declare wars when it can't really maintain another extra city?
• I love the extra research costs per added city. I never liked the technological advantages that used to be the case for big empires. Conquering should definitely not be the best way to win any type of victory. It adds poor underdeveloped cities to your empire that barely do any research and have a maintenance that they can only pay for if they get well-developed. Meanwhile they drag down the research rate. I like that balancing element against expansion.
• I like to play larger maps because it is harder to control what is going one with many empires and different continents. Given, the size, it is actually hard to conquer it all during a game. Maybe if you are a very aggressive war monger and really going for conquest. I usually do a bit of a mix of developing peacefully and war. I think that up till now, I have been the most peaceful civilization of all in the game which actually helped me focus on expanding peacefully.

Thanks for this excellent mod. I may have modded some small parts, but we all have slightly different preferences. I love the way that the economy starts working differently during the ages. I think that organic change is the biggest accomplishment of this mod. It doesn't feel forced, it feels like it gradually happens. I also love how resources work very differently and realistically in this mod.

 

[[Y]]

Below are all the changes I've made to the Totestra map generator, reflecting my preferences (you might want to alter differently to reflect yours)

Normalize Tectonic Plate Shapes
Changed lines #514 and #515 from

self.plateGrowthChanceX = 0.3
self.plateGrowthChanceY = 0.3

to

self.plateGrowthChanceX = 0.38
self.plateGrowthChanceY = 0.4

Spoiler Explanation :

The way Totestra map generation works (maybe other maps too, but this is the only one I've looked at) (from my current understanding, I haven't read through the entire script and all the relevant parts) is by dividing the map into tectonic plates and using the shapes of those plates to derive land. To create the plates, it randomly places a number of plate seeds, then runs a loop of (possibly) growing plates. Whether a a plate grows during each iteration of the loop is determined by a growth direction (either north/south or east/west), a growth chance, and whether a randomly generated number is less than that growth chance. `plateGrowthChanceX` is the growth chance for north/south and `plateGrowthChanceY` is the growth chance for east/west. It runs this loop until every tile on the map belongs to a tectonic plate.

So with the original code, it basically runs a loop where it takes a tile belonging to a tectonic plate, and for each available direction (north, south, east, and west), checks whether a randomly generated number is less than 0.3, and if it is, it adds the tile in that direction to the current tile's tectonic plate, as long as that new tile doesn't already belong to a plate. Since 0.3 is a small chance to grow a plate, the plates can grow very inconsistently and unevenly. Some might be relatively even, some might grow in thin branches, leading to the snaky continents.

By upping the numbers from `0.3` to `0.38` for north/south and `0.3` to `0.4` for east/west, we make it much more likely for growth to occur in each loop iteration, bringing the plate structure just a tad closer to an even all-round growth. And by making it bigger for east/west, we add a bias for plates to be wider than they are taller, which I think yields better maps since maps have more width than height in the first place. Much more space for continents to grow without crashing into map edges. I think it does some smoothing/forcing of specific things afterwards, but that's the gist of it.

Use Fewer Tectonic Plates
Changed lines #988, #990, #992, #994, and #996 from

self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0016) #988
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0032) #990
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0024) #992
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0008) #994
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0004) #996

to

self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0014) #988
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0028) #990
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0020) #992
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.006)   #994
self.hmNumberOfPlates = int(float(self.hmWidth * self.hmHeight) * 0.0003) #996

Spoiler Explanation :

Use fewer plates, meaning each plate can get bigger before running into another plate. I think this helps yields more stable continent shapes, but I'm not sure. I'm also not sure I modified the numbers here enough to actually make a difference. Each line yields fewer plates than the one above it, based on how patient the user is willing to be waiting for the map to generate. Though oddly, despite offering 6 levels of patience here, I'm pretty sure the map options only give 2 levels of patience.

Increase Landmass
Changed lines #663, #665, and #667 from:

self.landPercent = 0.29 #663
self.landPercent = 0.43 #665
self.landPercent = 0.19 #667

to

self.landPercent = 0.41 #663
self.landPercent = 0.54 #665
self.landPercent = 0.27 #667

Spoiler Explanation :

I like having more land on the map and more space to play in. More chaos that way. #663 changes the percentage for standard water level, #665 for lower water level, and #667 for high water level.

Adjust Land Type Rates
Changed line #698 from

self.DesertPercent = 0.15

to

self.DesertPercent = 0.06

and line #703 from:

self.PlainsPercent = 0.40

to

self.PlainsPercent = 0.30

Spoiler Explanation :

Totestra creates a rainfall map, giving each tile an amount of rainfall that varies based on distance from coast, wind directions, and other stuff. It uses the percents above to determine the likelyhood of a tile being desert, plains, or grassland (tundra has other mechanics, I guess). If rainfall is between 0 and 0.15, it's most likely to be a desert, with those closer to 0.15 having a chance to be plains instead. Those between 0.15 and 0.4 are most likely to be plains, with tiles closer to 0.15 possibly being deserts, and those closer to 0.4 possibly being grassland. And tiles with 0.4 or more rainfall are most likely to be grassland, with those nearer to 0.4 possibly turning up as plains.

Since the original numbers reflected a relatively smaller/interesting distribution for the original plate shapes and continent sizes, normalizing land masses and/or increasing land percentage can lead to continents mostly being desert, since rainfall doesn't penetrate too deep into the continent. These number changes help ease the distribution more so that you still get big deserts in the middle of continents, but also have big stretches of grassland and plains. Typing this out I realize that the "proper" approach would probably be to leave these numbers alone and instead modify the rainfall algorithm, but my understanding of the map script (and meteorology) doesn't extend that far, so this will have to do (and it does fine).

Remove Swamps
Changed line #375 from

self.HotMarshPercent  = 0.04

to

self.HotMarshPercent  = 0.00

Spoiler Explanation :

Swamps contribute nothing to the game other than randomly providing some players with a handicap. Removing them is just a better game for everyone.

Map examples (cropped minimaps, sorry for small sizes):

 

[[Y]]

Another thought I had recently, which I kept forgetting to mention here and is somewhat frustrating, is how building a road on a forest tile (26 turns at onset of game, no pastoral nomadism) takes more turns than clearing that tile of forest entirely (14 turns) plus building a road (11 turns). This is on 3.6, so if it's been changed since, great! But I skimmed the SVN release notes looking for "worker" and "road" and didn't find anything relevant. I like that roads in general take longer to build than they used to, but maybe the malus for building on a forest should be reduced?

 

[AllTheLand]

Some other thoughts (still based on playing in 3.60):
- I got Poland to capitulate in my Germany game but I had also taken some of their city outright. There was some separatism (mostly driven by culture and by the massive penalties for being ahead), and I decided to test out what happens if I don't try to contain it by piling up military units. It turns out that the cities would turn back to Poland, while Poland was still my vassal civ. While trying to return revolting cities with a strong foreign culture to the appropriate foreign civs makes a lot of sense, this situation seems very odd. And I can't just ask my vassal the city back, I would have to break the treaty and take it by force.
- Cities directly bordering freshwater lakes should enable (some of) the buildings that boost the exploitation of coastal tiles. My RI Planet Generator map has multiples freshwater lakes of 5-6 tiles. These are interesting features, but there should be ways to make them somewhat useful.
- Cities bordering freshwater lakes should not be able to build the Great Arsenal. In my game, the Aztecs built the Great Arsenal in a city that was bordering such a lake but with no direct or indirect access to the sea and no ability to produce ships.
- I forgot to mention it earlier, but only the english and german versions are playable. The other languages have some translated strings (including a couple of incorrect translations), but they don't just default to the english string when there is no proper translation. When the english string is something like "unit type (western)", the other languages have "western" as a string. You can imagine the massive confusion it creates when you are trying to understand the composition of an enemy unit stack or trying to pick what a city is going to build next without going in the city screen.

Yeah, not long before I also didn't really war early unless there was pressing reason to (maybe boxed in with 3 cities or single AI competitor on the continent). But I think good settle spots in the early game are really rare. Killing an AI neighbour early on can really give you breathing room and spare you a lot of trouble later down the line.

I saw an opportunity in taking Rome and razing another city, with a couple of shortswordmen and militia. That netted me a great capital location with 2 currently Pastoral Nomanism boosted cows, as well as gold and spices and lots of river tiles.

I think a major factor leading us to different conclusions is map size.

From your screenshot, there seems to be about 10 tiles between your capital and Rome, which is very little. That's one of the reasons I don't like playing small or standard maps, you get boxed in almost instantly.

For my Germany game, I went with a large map, and the Aztecs capital was 19 tiles to the south and the Polish capital was 17 tiles to the North. This means that I had to deal with strong barbarian attacks for a while, and that barbarian cities that settled as civilizations were quite strong, and it also means that it's much less enticing to go steal a capital spot. It creates a lot of issue, stretching your military resources, requiring a lot more work from workers to build connecting roads, incurring meaningful city-distance maintenance costs.

Otherwise, thank you for the write up, it was interesting.

But hearing you had 150 gpt saved through Legislator in 650, which could be spent into science was quite amazing, and is a very strong boost! How many cities did you have at that moment?

I didn't have these savings in my Germany-Immortal game, I just said that's the savings I would have had if I had that trait. I had around 20 cities. I will say that the free city raider promotion and the logistics boost were quite appreciable when besieging Aztec cities, but I still would have taken a big economic boost instead.

In my Poland-Monarch game, I had 147 gold per turn of savings with 27 cities when I called it a win in 480 AD.

I think it would be fun, if we had kind of a standard map in this thread. So players could try out different strategies, compare them with each other and learn new tricks watching other peoples saves. I think the regular Civ 4 BTS have this as hall of fame games?

It would be interesting, but because of different preferences, I'm not sure we could agree on a standard. I would personally lean towards a Large map with someone having gone over it to manually add some critical resources (in my opinion, while it's fine to only have one or two major food source types nearby, it's not fine to not have any copper or iron that could be reasonably contested) - not just for the player civ but for the territory handled by the AIs as well. Actually RI Tortestra with the option to put in a lot of resources seemed fine from a cursory look, but the massive deserts it generate on large maps (and even worse with low sea level - which is otherwise a nice option to allow realms to have some depth), and its general predictability

In my Germany game, there was only 3 silk resources on the entire map (for 11 civs + settling and breakaway civs) and fur was extremely sparse too. Since these don't give just happiness but are key to unlock a +10% gold building, I see that as a massive flaw. Perhaps needing these resources to get the building in the first place at all should be avoided.

I'm sure that other players would favor different generations. Of course it would be possible to do one type of map, then another... But with how involved completing a RI game can be, I don't know if there is the interest for that.

Thanks for the detailed feedback! It is always welcome.

And thanks for the detailed answers.

IIRC it was from 0AD, and its lack will probably be one of the negatives for you switching to a later version.

0AD music should have licensing allowing to use it in RI without issue, though? But the musics I was talking about are specifically Civ3 musics. They are not all the musics of the classical era (at least in 3.60), but are definitely some of them.

Anyway, it should be quite easy to customize the music anyway even if it's not present anymore in more recent versions.

I have great admiration for AdvCiv, and ported quite a few smaller things from it. Unfortunately, many of the more involved features are extremely complex code-wise and hard to port.

Yeah, I can understand why this creates a barrier to getting some of these features to RI.

AdvCiv is also a good tool to learn to play Civ4-variants, RI is reputed harder than vanilla but after AdvCiv the equivalent difficulty level feels relaxing.

I mean, isn't it one of the leitmotifs of Civilization game series as a whole? But yeah, I agree with you with some caveats.

1) You wrote yourself that you hand-picked opponents with wonder construction penalties and optimized your start location for the needed resources. I'd say in an average game where one player doesn't get to hoard wonders, their effects are more balanced.
2) Kremlin is indeed rather powerful, if situationally so, but Poland you picked is an absolute no-brainer for it; this is probably one of the strongest civ/wonder synergies in game.

Yeah, the Civilization game series tend to put a big emphasis on the World Wonders. But that's something I disliked even when I was playing Civ3 many years ago too. I see a couple issues with it:
- Once you have a good knowledge of the game, if you crank up the difficulty to be really challenging, you start to have only AIs getting all of the early wonders (usually by the late game you are in a winning position and leading techs again). If the difficulty is too low, the situation is reversed, you scoop everything.
- I don't like the dynamics of failing a world wonder because another civilization got it first. "We built the Pyramids at 99% your highness, we are just missing some stones at the top, but we heard that some civilization in a continent we don't even know about on the other side of the planet finished their Pyramids first, so we are going to just scrap the entire project." I understand why it works that way for the gameplay, but I find it problematic. I like the localization some wonders got but it's even more ridiculous in a way: "Your highness, we are stopping the work on our Statue of Odin because we heard that some other civilization completed a big Statue of Zeus." National Wonders avoid this pitfall.
- While some of them are weak enough to be taken opportunistically (or not) depending on the specifics of a game (civilization, starting location, etc.), some of them really feel "make or break" because of how strong their bonus can be. I admit there is a lot of personal preference here, but I would lean towards wonders being moderate boosts that bring some flavour and uniqueness, rather than so centralizing as they can be in vanilla Civ.
- Great Wonders are the only meaningful source of GP points in the ancient eras, and they keep being a massive boost for GP points for a very long time. In RI, passing up on wonders means mostly passing up on the great works of arts and the great works of science (I suppose having a specialist work for a while can net you one or two) during the early ages.

Yes, my choice in my Germany game to pick opponents with wonder construction penalties (basically because I wanted to play at Immortal difficulty without having my old Civ3 experience of not being able to get any wonder) allowed some wonder-hoarding, although it wasn't really any more than I had gotten as Poland in Monarch difficulty.

Regarding the Kremlin, maybe that's different on small or even standard maps, but on large maps I would call it a no-brainer for every single civilization on all possible types of map generation. Even if not running Serfdom, the 10% maintenance boost alone is very strong. This is even clearer in maps like the Europe scenario (a lot of land, little sea) or on low-sea generations. I got it not only in my Poland-Monarch game but also in my Germany-Immortal one.

Also related, I got the Solar Cult in both of my games because while it has drawbacks (late and limited missionaries) and some other religions have interesting bonuses, ultimately the city upkeep savings is too good to pass up. I'm convinced that, at least on large and huge maps, Solar Cult is objectively the single best religion in RI if you can do better than just maintaining your "natural borders".

And yes, I was calling number-of-city maintenance costs outrageous with both Solar Temples and Kremlin on providing -20% maintenance (so with a courtroom, nearly -30% to the post-courtroom cost). That's part of what I dislike with number-of-city maintenance as it's currently designed. It's very centralizing in its own way.

Well, it's a vanilla thing. If anything, the presence of additional modifiers in RI should dampen its effect (+150% in the capital vs +50% in all cities is less of a difference than +100% in the capital vs +0% elsewhere). I don't actually have a definite feeling on that one. To me it feels like from certain point onwards (around late Renaissance), other cities are able to catch up in most cases (again, though, I have to point out that a specific "hoard all the wonders in a perfect capital" playstyle can severely impact this). I think (but haven't checked) the GP point curve plateaus at some point; that would make having several cities generating a lot of GP point more effective in the long run.

Assuming the formula for GPs is the same as in vanilla, the additional number of GP points needed for the next GP becomes constant so percentage-wise the increase from each GP to the next gets less and less, which is probably the plateau you are referring to.

Yes, I suppose in the late game when you can feed a lot of specialists you can hope to get GPs from other cities, but it's still best play to put all the GP-generating wonders in the Heroic Epic-city. And I freely admit my playstyle warps things, but I picked this hoarding method precisely because of how effective it is. Because time goes by slower than in normal Civ4, you very rarely have situations where you need two different cities to concurrently work on two different wonders if you want to get them both, it's mostly a matter of getting the key tech and the key boosting resources in time.

My capital city in my Germany game generated 19 great people I can account for by 800 AD (7 settled, 1 golden age, 2 religious buildings, 1 academy of science, 1 glassworks, 3 great works of science, 4 great works of art - I never get techs with GPs because although immediate benefits are worth more than long-term ones, I prefer the long-term ones), needless to say it's going to take a very long time for a wonderless city to produce a GP.

Considering the Heroic-Epics are actually poems going by the localized names for the civs I had a look at, couldn't we have something like +35% in the city it's built in and +15% in all cities (so effectively +50%/+15%)? After all, the Odyssey didn't inspire Greeks only in the home city of Homer, it inspired Greeks in all Greece.

Notice that while I did this wonder-hoarding in the capital, I'm precisely arguing for balance changes that would disincentivize it.

I know what you're talking about, even though I've been seeing less of it in the later versions. Maybe raising the costs of everything across the board would be a good way to go. Oh and farms are still subjectively a better choice later in game IMO, as you can support several specialists from one farm.

I still have not played enough RI to know what should be done. An idea I had back when playing AdvCiv before touching RI was something like having the possibility to build a second library, a second courthouse, and so on. With the second building costing the same as the first one, but with a smaller benefit, making it more situational. If we had some building maintenance cost or some way to tie things with city size (imagine a single courthouse losing efficiency if the city population growth above 12, etc.) the concept could be even more powerful. On the other hand, I have to say that this would risk inflating the list of buildings in the sidebar of the city screen quite a lot... The UI is not really designed for such a concept.

I wish there was some way to have both farm tiles and town tiles be meaningful choices, but since we can't really have food transfers between "farm cities" towards "megalopolis" cities and each city tile is going to take away food from a potential specialist, I guess that doesn't work.

Isn't it strange that farm tiles can be more interesting in the modern era when most people live in cities (although this is admittedly represented by specialists) than in ancient eras where making town-tiles is such a good gold-source.

I mean, for me it's the other way round. "Expansion is a no-brainer" approach where a bigger civ is almost always better was always a turn-off. To be fair, a tall-vs-wide balance in the 4X genre is a very tricky thing, and the Civilization series itself also seems to flip-flop on that historically.

The reality of the world is that, if you are unopposed, bigger is just better. One of the main reasons the USA are currently the number one world power is that they could expand on a massive amount of land against very limited opposition and form a large country with enough in common to not split apart (although there was a close call).

Map-painting strategy games are always about investing now into later success. It may be better now to work on improving your core city by building things there, by getting workers to improve tiles, or by investing into creating a new city, or by investing in attacking a neighbour. But in the later stages a world power is always going to need some territorial expansion if it wants to be the number one power and to have access to more resources.

Now I'm not saying strategies that aim to maximize the amount of territory you control in the early and mid-game should be dominant.

I just think that the completely artificial number-of-cities upkeep mechanic is not just guilty of irrealism, it's guilty of creating unfun gameplay incentives.

Is it really a debate about "tall vs wide" when all my core cities are basically as tall as can possibly be (all buildings except a couple of military buildings for units I don't intend to build there anyway, all or nearly all tiles fully improved), and I'm just having to wait for the next tech that'll allow me to have something to build or more money to make expansion viable?

In my Germany game I ended up attacking Poland (the map generation really set up the joke) and got a few more new cities, but that was in part because i was bored and they were easy prey, money-wise it's going to take a long time to pay back.

The kind of mechanisms I envision as limiting rabid expansion in more interesting ways:
- Having to fight barbarians and other civs to actually gain the territory. It's of course already the case to a good degree, but I want to emphasize that usually should be a major factor.
- Diplomatic desire for a "balance of power". I don't necessarily mean having all other civs hating on you and refusing any trade even if beneficial to them (as it can too easily happen) just because you have a big territory, but there should be some attempts by other civs to curtail civs that expand too much.
- Reducing the economic benefits from distant cities. They would still build their improvements at somewhat normal rates, but the production of military units and the tax rate would be cut down. Tying it with separatism could be very interesting if the player had some ways to adjust the rate within distance and tech based upper and lower limits, risking revolts if trying to get too much or if revoking formerly granted tax-breaks... Doing it right would likely be a massive amount of work, so I'm certainly not demanding it. But I'm persuaded that such a system has a higher potential. In some ways the notion of far-flung cities being inefficient has Civ3 vibes, except Civ3 completely killed all kind of production including buildings, leading to what I used to call "zero-cities", cities that had no usefulness whatsoever apart as military stepping stones. And I know the Civ4 mechanic was introduced as a way to pace expansion compared to Civ3, but in RI it really hurts.
- Having to deal with internal troubles. Random disasters are not too fun in general if they can't be sufficiently anticipated. Unfortunately, this may be something that the Civ4 engine can't deal with properly. If you have ever played the Crusader Kings series, you'll know what I'm thinking about. The separatism mechanic goes towards this to some degree, but while it's one thing to have to conquer back cities when the land have a foreign culture, it's another when they are culturally similar : if the military might of the rebels is broken and a couple cities are retaken (including the rebellion's capital), the others should surrender and join back your civilization. Making this sort of revolt slightly easier to deal with would also open room to making it happen more often. Ideally, the optimal gameplay should not be to always manage for such a revolt to never happen, but instead to have it happen and have to deal with it ; at least if playing a large realm in the early eras.

I will also note that while I didn't say a word about it so far, the cost-increase on units based on how many units you already have of the same type is strongly a boost to smaller civs. When you want to put archers in all your cities and you see a +250% penalty to build another archer, you start to ask yourself if you shouldn't just put recruits there...

I don't recall if it was before or after the version you played, but there was a major bug I fixed where most of the code for evaluating Open Borders simply didn't get called. So unfortunately I can't say whether you're criticizing the brain-dead version or the version after I already fixed it.

Let's assume it was the broken version, I'll see if it happens still when I'll start a game on the SVN version.

That one is in my sights, but I haven't done anything particular with it yet. A relation bonus seems unnecessary, but an additional factor in war preparation calculations might be in order.

Good to hear!

Again, that's a vanilla thing we didn't touch. And vanilla is actually quite stupid in this regard, as the civ that accepts an invitation to war doesn't have a war plan in place. Would make more sense if it first spent several turns getting its troops in place. But I'm not coding that...

It would be a lovely feature but you are the best judge of what your time is best spent on and what's within your coding abilities.

I don't really get the hate. This is barely a mechanic at all, as players can't interact with it - just a simple modifier that ticks linearly over time. But you'll be glad to know that recently I cut away a chunk of K-mod inflation-related code which should result in lower inflation in most games.

I can't interact with it, but I sure feel the 600 additional gold I'm having to pay per-turn in 800 AD in my Germany game. Now don't get me wrong - I should have to pay most or all of that anyway for things to be balanced, but it coming from sources I can control and where I make a deliberate investment, instead of just a blanket increase on everything. Inflation is one of the reason why the number-of-cities upkeep feels more oppressive as when playing in AdvCiv (it still felt oppressive there).

I'm certainly glad to know you cut that code. If you could explain briefly what kind of triggers led that code to increase inflation, that would satisfy my curiosity.

Sort of. You yourself mentioned that cities can be built up relatively quickly, and later in game ministries help out a lot. But I agree there is, especially later on, a clear distinction between your core cities, losing which hurts a lot, and the periphery that is more extractive and less productive, which can be lost and regained with relative ease.

Yeah, I suppose that the possibility to build up cities quicker in RI make it less of a concern here.

Although in my current game I better not let cities fall considering the "culture victory off" bug that exists in 3.60 leading to AIs razing all the cities they capture.

I feel there is a lot of difference in utility between individual playstyles, and I've seen people with vastly different ones. From your observations, I'd place you firmly into the economic player camp, for whom military-related traits are relatively worthless; but for some other players they can make or break a game.

Playstyles certainly matters a lot, but I think there is some sound data behind valuing economic bonuses highly.

I saw in the changelog you grabbed some voicelines from AoE2. Well, speaking about AoE2, every single pro player agrees that typical economic bonuses (say +10% gathering of some common resource) are more valuable than typical bonuses just making units stronger (say +10% strength to some unit type), because economic bonuses are flexible. In Civ4 terms, a stronger economy can be used to research techs faster. To build more wonders. To build more building-improvements in cities. To get more workers to improve tiles around cities. To build a bigger army. To invest in settling new cities.

While a military bonus requires you to start by building specific military units (which you have less resources to afford since you don't have the economic bonus), and then you need a war (often with some specific traits). When I spent dozens of turns defending against barbarian assaults (a pro Civ4 player might have managed to spawn-bust barbarians, but I didn't really try it), the city raider 1 promotion on my military units sure felt useless. But the hammer-boost in my cities was most welcome. In my first wars against the Aztecs, I was on the defensive so again the City Raider promotion felt useless.

Even if the end effect comparing a specific use of the economic bonus and of the military bonus happen to be equally strong, the flexibility afforded by greater economic power make it strategically more valuable.

In the end, I mostly think that the traits that ought to be boosted are the one I'm not likely to pick in the first place, so I'm not sure if the fans of military-traits would find it so objectionable.

I agree it's one of the stronger ones. Though I can't say I see Poland overperforming in AI-only test games. There are some civs that do, sometimes mysteriously so (I don't really know what Armenia's deal is, but I almost never see a weak Armenia).

I suppose you shouldn't nerf if you don't have much data suggesting it to be too strong.

Eh, I am not too partial to it myself, but there used to be a very vocal player subset that demanded research speed to match the displayed date. Also, due to the non-linear tech cost increases across the tech tree, the actual impact on the progress speed is actually smaller than one feels it should be at the first glance.

I see. Since it's just an option, everyone can be happy.

I can't really remember when, but I overhauled the resource placement at some point. Again, it may or may not be the version you've played.

You are also handling the PlanetGenerator generating code?

As others already pointed out, RI has AI autoplay. It gets disabled for the release versions though, so you'll have to use SVN for that.

Ah, thanks for the precision.

I am actually curious - from the time you made the change, did you ever reach 40+ cities? Because from my experience, the economy is balanced to work and works as intended at least until that amount. I often see AI civs winning the game have 30+ cities.

You didn't ask this to me, but I'll still answer. I got to 27 cities in 480 AD in my Poland-Monarch game on the Europe scenario. I'm at 27 cities in 800 AD in my Germany-Immortal game on a random map. I have a couple spots I could settle if I didn't want to limit my city number. If playing on large/huge maps (especially with a high proportion of land vs sea), 40+ cities is really not that crazy.

Weeeell... A settling barbarian civ is already at a disadvantage when it comes to most other things, most crucially the quality of its territory. Starting spots for initial civs are not only picked from the best on the map, but also usually additionally sweetened, and then they organically expand to the best spots around them, whereas a settling barbarian civ is basically a collection of cities in random spots. If these emerging civs aren't given an edge somewhere, they'll just be fodder. Generally speaking, my headcanon for "settling barbarians" is not something like Mongols riding in from the steppes, but rather an organized society that was looked down upon by their more ancient neighbours but may actually turn out to be more dynamic and progressive than the already long-established ones, such as Greeks as seen by Egyptians.

I see. Playing with 11 civs on a large map with the 40+% ocean option in PlanetGenerator, my barbarian settling civs had for the most part a fair amount of cities and some interesting spots, although indeed not the best.

I advocate for more resource-rich random maps, that would also help settling barbarians.

Since you've already mentioned you want to change the way the techs they get is determined, I don't have much to add, just that I hope it's going to be done.

Did you consider using the temporary building trick to give a more passive boost to the capital of settling barbarians for the first 100 turns or so? Instead of settling down and behind on par or ahead of most other civs, they could have a "surge" where they go from being a little backwards to being much more competitive.

Is it also possible in theory to trigger some event that would sweeten a bit the quality of their capital's lands?

Poland is one of the civs that has a very clearly defined "Golden Age" when it comes to gameplay. A large bonus during a limited time window might not necessarily be better than +1 of something for the majority of the game... I am not totally closed off to nerfing it, but Poland overall doesn't feel OP to me.

Better wait on more feedback then.

I feel "wider" civs still generally dominate the rest. Not so completely as in vanilla, but still.

I will disagree with you here. Starting spots for capitals are usually the best city spots there are in the world; and early on when your overall number of cities should stay low, the quality of individual city spots is very important. A foreign capital is often worth the investment in taking it.

I'll refer you back to what I said to Watermelon above in this post. 10 tiles vs 17/19 tiles distance to reach a foreign capital changes calculation quite a bit (although I can believe it would be worth in some circumstances to go for the grab).

Not stronger; they are as strong as the weakest unit (irregular) you can currently build. While the transition period immediately after flintlocks can be a bit painful, you're able to build/upgrade to very cheap units of your own at the same time if defenses anywhere seem to be lacking.

I didn't have any of the new irregulars yet (I changed civics just after getting the tech), and I didn't have Stirrups. So I had 6 Str. swordsmen, 4 Str. archers, 4 Str. pikes, 5 Str. horsemen and 4 Str. skirmishers. None of which are actually stronger than the 6 Str. irregulars unless you get promotions.

I'm not requesting any change in the balance, I'm just saying I got completely blindsided by it when it happened and I had to really scramble to get new units ready because I had to put down two revolts that started almost immediately after I enacted the new civics. I had readied some troops for revolts but they were too weak.

Do you feel that settling / taking a city should always be a no-brainer net positive then? As it currently stands, there is a clear economic curve over time where expansion usually happens after several clearly defined jumps in economic efficiency, and yes, overextending beyond what you can support is painful. I feel it is a design decision that fosters more deliberation over where one wants to settle/expand to. To get a colonial empire going, one first has to have banks at home!

No, I don't think it should always be a no-brainer net positive. I already said a lot of things on related topics earlier in this message, but I'll add some thoughts here.

It represents an opportunity cost (creating a settler, creating military units). It represents ongoing costs (supplying workers to improve its tiles, supplying military to defend it, civics upkeep cost...).

But to take the easiest case of settling a city in empty land, I think in most situations it should be a benefit (it should not be if you can't defend it militarily, supply it with a worker, keep it from quickly revolting, etc.) - but it might not be the biggest benefit. The benefit might be (a lot) bigger to wait and focus on something else first, because the investment is costly and the opportunity cost too high.

What I take issue with is the economy completely collapsing because you just got a new outpost of your civilization to trade with... Settling a new city should not suddenly rise your realm-wide expense by 40 or 50 gold per turn (which is what I'm currently seeing with 25+ cities despite a lot of anti-maintenance buildings)...

And I'll give you two easy examples, the Roman Empire and Ancient China. Of course, both were plagued by instability, civil wars, and so on. There definitely were issues. The "power per population" and "power per territory" of the central ruling authority was definitely a lot smaller than for smaller realms. But they didn't instantly collapse because "you first need to have banks at home" to get a large empire going. Thanks in part to good internal trade (and communication) routes, unity could persist for fairly long periods of time. But have a "Roman Empire" scenario on RI's Europe map, and the Roman Empire would instantly collapse because its "number-of-cities" maintenance would bankrupt it instantly.

I wonder how the game would play out if newly settled cities tended to be even slower to grow, but you could quicken it up by sending more settlers from your core cities...

I find scouts too be a bit too squishy to scout reliably.

I second this. I tried making a scout or two in my first game and I lost them very easily to barbarians.

On a similar note I often only get 2 horsemen (the 5-Str. guys), mostly because the first two are so cheap. They are mobile, but lack punching power against pretty much anything except a lone shortsword man or recon unit in the open, or to raid improvments while advancing.

I second this. 5-str horesemen are maybe good enough to quickly help against slave rebellions during the classical age, but their combat strength is so pitiful that their niche is very small, since even if they get to the place of the fight quickly, they don't offer much beyond the mobility support. These units are pretty cheap and are not expected to pack the same punch as later Cataphract-class 7 Str. horsemen, but their useful timeframe is quite restricted. Perhaps discovering Stirrup could give them some boost, or does that go against design practices?

 

[AspiringScholar]

What a sudden flurry of activity in this thread! The list of interesting things to jump into the conversation about is getting too long.

 

[AllTheLand]

Some quick notes and questions:
- I got the SVN version running.
- The AI doesn't seem to take negative yields for buildings into account or at least not properly. I did a test adding a -5 "coins" (the thing that can be changed into both science and tax revenue, I was editing the XML so I was guessing the field) yield to the warehouse, and I confirmed it applies the penalty. However, one of the AI cities still ended up building one. While tying a cost to some of the most important buildings shouldn't lead to the AI making gross miscalculations, it still could lead to misevaluations on what's the best next build. Tying it to more optional buildings would cripple the AI.
- How do I disable the custom RI textures? The installer for the release version offered an option but on SVN I suppose I need to remove some files.
- Regarding the music, I suppose I was just talking about the default Civ4 selection, as it seems the Civ3 music I mentioned came packaged with Warlords. Somehow I don't recall hearing it when playing AdvCiv.

I am not sure what you are looking for. Removing inflation is a simple xml change in the file CIV4GameSpeedInfo.xml in the mod folder where you set (for all the game speeds) the value iInflationPercent to 0. It will affect the game that you are playing (backup the old file).

What I did on unit costs is double the value of iUnitCostPercent for every difficulty level (the AI plays at noble) in the file CIV4HandicapInfo.xml. But that is a dirty way to do it as I just double the costs of all units. But it was a quick and easy 'proof of concept'.

Thanks for the pointers.

Editing all units wouldn't be that much of a pain if each unit didn't have 15 regional variations...

While I'd like to completely remove inflation, just doubling the upkeep of units would still leave expenses too low I suspect. Of course the AI would benefit too.

The change that I did for city maintenance is in the dll, but you don't seem to be asking about that one. For each svn version, I will need to create a new dll-file.

I'm not in a hurry to recompile the .dll.

I'm going to use the easy way of tweaking the forms of governments to reduce number-of-cities maintenance for the time being and see how it goes.

I think it is purely related to the value iConquestProb in the file CIV4BuildingInfos.xml. It is quite low for a lot of buildings. At a quick glance it is 0% for about every military building and 66% for many other buildings. I agree that this could be a lot higher. I guess it makes conquest less profitable. Ideally, it would be related to the number of units killed to take the city or something like that. A very old game like Master of Magic even made it related to the number of turns of fighting for the city (it has a separate turn-based combat screen where units fought each other).

Something based on how much fighting there was for the city would make some sense but it's probably too involved to be done.

I'll try this for my next game:

Spoiler :

- Walls get a 70% likelihood to be kept.
- 66% buildings become 80% buildings instead. This means that if a city is lost and reconquered, approximately 64% of buildings will remain, instead of approximately 45%.
- 33% buildings becomes 60% buildings instead. 33% buildings include notably many kinds of religious buildings and resource-dependent buildings such as the Tailor, but also the airports.
- The zoroastrian Temples (25%), the jewish temple and the muslim temple (40%) all get to 60%.
- Taverns and zoroastrian cathedrals get 80% instead of 50%.

 

[Walter Hawkwood]

- I got Poland to capitulate in my Germany game but I had also taken some of their city outright. There was some separatism (mostly driven by culture and by the massive penalties for being ahead), and I decided to test out what happens if I don't try to contain it by piling up military units. It turns out that the cities would turn back to Poland, while Poland was still my vassal civ. While trying to return revolting cities with a strong foreign culture to the appropriate foreign civs makes a lot of sense, this situation seems very odd. And I can't just ask my vassal the city back, I would have to break the treaty and take it by force.

A good observation; I should add a check for the rebels wanting to rejoin their parent civ for that civ not to be a vassal. In this situation, you should end up with two Polands instead!

- Cities directly bordering freshwater lakes should enable (some of) the buildings that boost the exploitation of coastal tiles. My RI Planet Generator map has multiples freshwater lakes of 5-6 tiles. These are interesting features, but there should be ways to make them somewhat useful.

Yes, maybe. Shouldn't be too hard to code.

- Cities bordering freshwater lakes should not be able to build the Great Arsenal. In my game, the Aztecs built the Great Arsenal in a city that was bordering such a lake but with no direct or indirect access to the sea and no ability to produce ships.

That's interesting to hear. I didn't think they could. I'll look into it.

- I forgot to mention it earlier, but only the english and german versions are playable. The other languages have some translated strings (including a couple of incorrect translations), but they don't just default to the english string when there is no proper translation. When the english string is something like "unit type (western)", the other languages have "western" as a string. You can imagine the massive confusion it creates when you are trying to understand the composition of an enemy unit stack or trying to pick what a city is going to build next without going in the city screen.

Yep, I am one person without an army of translators at my disposal. I only actively maintain the English version. A while ago a kind person did a full German translation as well, so I try keeping it up to date since then too (I will freely admit I mostly use Google Translate for that; I do speak some German, but having been a professional translator earlier in my life, doing it in my free time feels too much like work, so I "delegate" it to Google and only sense-check so that it doesn't output active nonsense
).

In my Germany game, there was only 3 silk resources on the entire map (for 11 civs + settling and breakaway civs) and fur was extremely sparse too. Since these don't give just happiness but are key to unlock a +10% gold building, I see that as a massive flaw. Perhaps needing these resources to get the building in the first place at all should be avoided.

That's by design; certain such buildings (Trading Posts, Jewellers) are additional boosts you only get if you have resources that are supposed to be scarce.

0AD music should have licensing allowing to use it in RI without issue, though? But the musics I was talking about are specifically Civ3 musics. They are not all the musics of the classical era (at least in 3.60), but are definitely some of them.

Anyway, it should be quite easy to customize the music anyway even if it's not present anymore in more recent versions.

0AD is free, so feel free to grab yourself some of it! It's all in mp3 format anyway, IIRC, and you can listen to it just as easily. Ultimately, I decided (after actually switching music on for the first time in many years) that, while undoubtedly good, it was too clashing with the rest of the Civ 4 soundtrack and that either the rest of the eras needed to be redone in the same way (from god knows what sources) or it had to go. I would actually love an AI tool that could take a basic tune and rewrite it in several different styles - that would be ideal for things like leader music - but so far I haven't seen anything like that, even though procedural AI music generation has actually come a similarly long way as the image generation did.

AdvCiv is also a good tool to learn to play Civ4-variants, RI is reputed harder than vanilla but after AdvCiv the equivalent difficulty level feels relaxing.

That's a first! But I guess if something would have been harder on an equivalent difficulty than RI, it would be AdvCiv...

- Once you have a good knowledge of the game, if you crank up the difficulty to be really challenging, you start to have only AIs getting all of the early wonders (usually by the late game you are in a winning position and leading techs again). If the difficulty is too low, the situation is reversed, you scoop everything.

Then bear in mind too that you are probably a better player than most people in this thread, myself included. My own comfort level in RI is Monarch, and I know most players also don't do anything higher - so bear that in mind when commenting as well, the majority of the players play far less optimally than you do.

- I don't like the dynamics of failing a world wonder because another civilization got it first. "We built the Pyramids at 99% your highness, we are just missing some stones at the top, but we heard that some civilization in a continent we don't even know about on the other side of the planet finished their Pyramids first, so we are going to just scrap the entire project." I understand why it works that way for the gameplay, but I find it problematic. I like the localization some wonders got but it's even more ridiculous in a way: "Your highness, we are stopping the work on our Statue of Odin because we heard that some other civilization completed a big Statue of Zeus." National Wonders avoid this pitfall.

I actually like the dynamics. It's an investment that either pays off or fails. It brings an element of risk assessment to what would otherwise be "click a button to see numbers increase".

And from a purely flavour perspective, in a given timeline, consider the "failed" wonders to, say, not having been impressive in the first place ("Oh so the great pyramids are supposed to be more than 5m tall?") or perhaps not having been completed in the first place ("But I got reports that my glorious pyramids were almost ready! What's the meaning of this empty field?!") or maybe having been a psyop all along ("Ah yes, now that the Egyptians have unveiled their actual pyramids, our excellent scheme to funnel money into treasury by telling people that they're paying additional tax for pyramid construction will no longer work...").

- While some of them are weak enough to be taken opportunistically (or not) depending on the specifics of a game (civilization, starting location, etc.), some of them really feel "make or break" because of how strong their bonus can be. I admit there is a lot of personal preference here, but I would lean towards wonders being moderate boosts that bring some flavour and uniqueness, rather than so centralizing as they can be in vanilla Civ.

I do try to eliminate the ones that are an absolute must have regardless of one's strategy and circumstances, and admittedly from my own observations the Kremlin comes closest to that currently.

- Great Wonders are the only meaningful source of GP points in the ancient eras, and they keep being a massive boost for GP points for a very long time. In RI, passing up on wonders means mostly passing up on the great works of arts and the great works of science (I suppose having a specialist work for a while can net you one or two) during the early ages.

That is true, though for the most important types of GP in the ancient eras (subjectively prophets and scientists), dedicated buildings and specialists become available quite early. Still, I see a good case for halving the wonder GP point gain.

Regarding the Kremlin, maybe that's different on small or even standard maps, but on large maps I would call it a no-brainer for every single civilization on all possible types of map generation. Even if not running Serfdom, the 10% maintenance boost alone is very strong. This is even clearer in maps like the Europe scenario (a lot of land, little sea) or on low-sea generations. I got it not only in my Poland-Monarch game but also in my Germany-Immortal one.

And you're beginning to see my dilemma of having to balance for a wide variety of playstyles, map sizes etc.

Also related, I got the Solar Cult in both of my games because while it has drawbacks (late and limited missionaries) and some other religions have interesting bonuses, ultimately the city upkeep savings is too good to pass up. I'm convinced that, at least on large and huge maps, Solar Cult is objectively the single best religion in RI if you can do better than just maintaining your "natural borders".

What can I say, you're a smart person. An "administrative" religion that's best suited for an economic buildup.

Considering the Heroic-Epics are actually poems going by the localized names for the civs I had a look at, couldn't we have something like +35% in the city it's built in and +15% in all cities (so effectively +50%/+15%)? After all, the Odyssey didn't inspire Greeks only in the home city of Homer, it inspired Greeks in all Greece.

Notice that while I did this wonder-hoarding in the capital, I'm precisely arguing for balance changes that would disincentivize it.

That is indeed a good suggestion, well worth considering, and easy to implement.

Isn't it strange that farm tiles can be more interesting in the modern era when most people live in cities (although this is admittedly represented by specialists) than in ancient eras where making town-tiles is such a good gold-source.

I actually feel it is a good simulation of the exodus of people from the countryside during the Industrial Revolution. I see towns as still predominantly rural communities, and their bonuses as representing traditional rural crafts, whereas the true urban centers are the cities.

I still have not played enough RI to know what should be done. An idea I had back when playing AdvCiv before touching RI was something like having the possibility to build a second library, a second courthouse, and so on. With the second building costing the same as the first one, but with a smaller benefit, making it more situational. If we had some building maintenance cost or some way to tie things with city size (imagine a single courthouse losing efficiency if the city population growth above 12, etc.) the concept could be even more powerful. On the other hand, I have to say that this would risk inflating the list of buildings in the sidebar of the city screen quite a lot... The UI is not really designed for such a concept.

Well, since we have building "tiers" (Toll House / Tax Collector / Bank etc), if the cost difference between the more basic and the more advanced buildings is greater, it can provide meaningful diminishing returns.

The reality of the world is that, if you are unopposed, bigger is just better. One of the main reasons the USA are currently the number one world power is that they could expand on a massive amount of land against very limited opposition and form a large country with enough in common to not split apart (although there was a close call).

Off the top of my head, 2 of the top-5 economies in the world are small islands with relatively low population (at least compared to US/China, not talking Luxembourg-sized obviously). And most definitely, some of the most technologically advanced countries are rather small, a feat that would be impossible in vanilla Civ 4. The internal logic of "bigger is better" provides excellent simulation of the leading superpowers, but really fails to capture the second-tier world powers.

Is it really a debate about "tall vs wide" when all my core cities are basically as tall as can possibly be (all buildings except a couple of military buildings for units I don't intend to build there anyway, all or nearly all tiles fully improved), and I'm just having to wait for the next tech that'll allow me to have something to build or more money to make expansion viable?

Kind of; it is about striking a balance where you aren't hitting the "tall" ceiling in process.

The kind of mechanisms I envision as limiting rabid expansion in more interesting ways:
- Having to fight barbarians and other civs to actually gain the territory. It's of course already the case to a good degree, but I want to emphasize that usually should be a major factor.
- Diplomatic desire for a "balance of power". I don't necessarily mean having all other civs hating on you and refusing any trade even if beneficial to them (as it can too easily happen) just because you have a big territory, but there should be some attempts by other civs to curtail civs that expand too much.
- Reducing the economic benefits from distant cities. They would still build their improvements at somewhat normal rates, but the production of military units and the tax rate would be cut down. Tying it with separatism could be very interesting if the player had some ways to adjust the rate within distance and tech based upper and lower limits, risking revolts if trying to get too much or if revoking formerly granted tax-breaks... Doing it right would likely be a massive amount of work, so I'm certainly not demanding it. But I'm persuaded that such a system has a higher potential. In some ways the notion of far-flung cities being inefficient has Civ3 vibes, except Civ3 completely killed all kind of production including buildings, leading to what I used to call "zero-cities", cities that had no usefulness whatsoever apart as military stepping stones. And I know the Civ4 mechanic was introduced as a way to pace expansion compared to Civ3, but in RI it really hurts.
- Having to deal with internal troubles. Random disasters are not too fun in general if they can't be sufficiently anticipated. Unfortunately, this may be something that the Civ4 engine can't deal with properly. If you have ever played the Crusader Kings series, you'll know what I'm thinking about. The separatism mechanic goes towards this to some degree, but while it's one thing to have to conquer back cities when the land have a foreign culture, it's another when they are culturally similar : if the military might of the rebels is broken and a couple cities are retaken (including the rebellion's capital), the others should surrender and join back your civilization. Making this sort of revolt slightly easier to deal with would also open room to making it happen more often. Ideally, the optimal gameplay should not be to always manage for such a revolt to never happen, but instead to have it happen and have to deal with it ; at least if playing a large realm in the early eras.

Well, most of it is already present in one form or another. It's just again a question of finding the right balance then.

I will also note that while I didn't say a word about it so far, the cost-increase on units based on how many units you already have of the same type is strongly a boost to smaller civs. When you want to put archers in all your cities and you see a +250% penalty to build another archer, you start to ask yourself if you shouldn't just put recruits there...

Indeed, and was intended to be a net positive for smaller civs. And to be honest, you should; cities away from fortified borders should be garrisoned by irregulars at least partially.

It would be a lovely feature but you are the best judge of what your time is best spent on and what's within your coding abilities.

Diplomacy is something I try touching as little as possible. And a lot of stuff there seems to be hardcoded anyway.

I can't interact with it, but I sure feel the 600 additional gold I'm having to pay per-turn in 800 AD in my Germany game. Now don't get me wrong - I should have to pay most or all of that anyway for things to be balanced, but it coming from sources I can control and where I make a deliberate investment, instead of just a blanket increase on everything. Inflation is one of the reason why the number-of-cities upkeep feels more oppressive as when playing in AdvCiv (it still felt oppressive there).

I'm certainly glad to know you cut that code. If you could explain briefly what kind of triggers led that code to increase inflation, that would satisfy my curiosity.

K-Mod had a large and unnecessarily complicated blob of code that was supposed to increase global inflation if the global tech rate was progressing too quickly. I am also 80% sure it caused an integer overflow somewhere and led to strange inflation values later in game.

Although in my current game I better not let cities fall considering the "culture victory off" bug that exists in 3.60 leading to AIs razing all the cities they capture.

Oh yeah, that was a fun one to catch. I recall that...

Playstyles certainly matters a lot, but I think there is some sound data behind valuing economic bonuses highly.

I saw in the changelog you grabbed some voicelines from AoE2. Well, speaking about AoE2, every single pro player agrees that typical economic bonuses (say +10% gathering of some common resource) are more valuable than typical bonuses just making units stronger (say +10% strength to some unit type), because economic bonuses are flexible. In Civ4 terms, a stronger economy can be used to research techs faster. To build more wonders. To build more building-improvements in cities. To get more workers to improve tiles around cities. To build a bigger army. To invest in settling new cities.

While a military bonus requires you to start by building specific military units (which you have less resources to afford since you don't have the economic bonus), and then you need a war (often with some specific traits). When I spent dozens of turns defending against barbarian assaults (a pro Civ4 player might have managed to spawn-bust barbarians, but I didn't really try it), the city raider 1 promotion on my military units sure felt useless. But the hammer-boost in my cities was most welcome. In my first wars against the Aztecs, I was on the defensive so again the City Raider promotion felt useless.

Even if the end effect comparing a specific use of the economic bonus and of the military bonus happen to be equally strong, the flexibility afforded by greater economic power make it strategically more valuable.

In the end, I mostly think that the traits that ought to be boosted are the one I'm not likely to pick in the first place, so I'm not sure if the fans of military-traits would find it so objectionable.

Well, you're invoking "pro players" here, and from what I gather so far, you're also a fan of hyper-optimal playstyles. I will freely admit that such competitive players are not what I'd consider the core demographic, though, of course, I try to accommodate a wider audience.

You are also handling the PlanetGenerator generating code?

The bits of code I added then were extrinsic to map scripts, they got called from them, so yes, it should have covered PG as well. But you've already been playing the results of that, it was quite a while ago. Before that, certain resources would almost never spawn at all.

You didn't ask this to me, but I'll still answer. I got to 27 cities in 480 AD in my Poland-Monarch game on the Europe scenario. I'm at 27 cities in 800 AD in my Germany-Immortal game on a random map. I have a couple spots I could settle if I didn't want to limit my city number. If playing on large/huge maps (especially with a high proportion of land vs sea), 40+ cities is really not that crazy.

If there's one thing I'd retroactively change about RI now is I'd remove the largest map sizes (and retire the Huge World Map and other huge map scenarios, but I know it has a lot of fans). After many years I feel that Civ 4 in general was designed with smaller maps in mind, probably no larger than Large.

I see. Playing with 11 civs on a large map with the 40+% ocean option in PlanetGenerator, my barbarian settling civs had for the most part a fair amount of cities and some interesting spots, although indeed not the best.

I advocate for more resource-rich random maps, that would also help settling barbarians.

Since you've already mentioned you want to change the way the techs they get is determined, I don't have much to add, just that I hope it's going to be done.

Did you consider using the temporary building trick to give a more passive boost to the capital of settling barbarians for the first 100 turns or so? Instead of settling down and behind on par or ahead of most other civs, they could have a "surge" where they go from being a little backwards to being much more competitive.

It is already so; barbarians do get a large temporary boost as well. It has been surprisingly difficult to ensure emergent civs being at least moderately competitive.

Is it also possible in theory to trigger some event that would sweeten a bit the quality of their capital's lands?

In theory yes; it does feel somewhat gamey to me (at least more so than the measure we already have in place).

I didn't have any of the new irregulars yet (I changed civics just after getting the tech), and I didn't have Stirrups. So I had 6 Str. swordsmen, 4 Str. archers, 4 Str. pikes, 5 Str. horsemen and 4 Str. skirmishers. None of which are actually stronger than the 6 Str. irregulars unless you get promotions.

I'm not requesting any change in the balance, I'm just saying I got completely blindsided by it when it happened and I had to really scramble to get new units ready because I had to put down two revolts that started almost immediately after I enacted the new civics. I had readied some troops for revolts but they were too weak.

What I am actually hearing here is that there should be more signposting for new players regarding that.

And I'll give you two easy examples, the Roman Empire and Ancient China. Of course, both were plagued by instability, civil wars, and so on. There definitely were issues. The "power per population" and "power per territory" of the central ruling authority was definitely a lot smaller than for smaller realms. But they didn't instantly collapse because "you first need to have banks at home" to get a large empire going. Thanks in part to good internal trade (and communication) routes, unity could persist for fairly long periods of time. But have a "Roman Empire" scenario on RI's Europe map, and the Roman Empire would instantly collapse because its "number-of-cities" maintenance would bankrupt it instantly.

To be fair, the Roman empire didn't last terribly long, and were it not for a spectacular successive intervention by Aurelian (a military genius) and Diocletian (an administrative genius), would have existed shorter still. Also, Rome very clearly ran into its expansion limits. At its most powerful (during the reign of the Five Good Emperors), it even willingly abandoned certain territories with an understanding that they couldn't be held. China was generally not expansionist beyond its core territories, being more content with having tributaries beyond its immediate borders rather than direct control.

I second this. I tried making a scout or two in my first game and I lost them very easily to barbarians.

Well, to be fair their intended purpose was to be useful before barbarians arrive. I might hand them more bonuses specifically against barbarians, but generally speaking, they're not supposed to hold their own against archers even under favourable circumstances. Archery units are supposed to be hard counters for recon units.

I second this. 5-str horesemen are maybe good enough to quickly help against slave rebellions during the classical age, but their combat strength is so pitiful that their niche is very small, since even if they get to the place of the fight quickly, they don't offer much beyond the mobility support. These units are pretty cheap and are not expected to pack the same punch as later Cataphract-class 7 Str. horsemen, but their useful timeframe is quite restricted. Perhaps discovering Stirrup could give them some boost, or does that go against design practices?

Again, to be clear, the design decision here was that horsemen are supposed to be "bad cavalry" for the civs that can't build horse archers. They have relatively limited usefulness by design - having them rather than horse archers is supposed to be a drawback, and their main intended use is not direct combat engagements but scouting, pillaging and combat aid (basically how pre-Hellenic era Greeks or early Romans used their very limited cavalry).

What a sudden flurry of activity in this thread! The list of interesting things to jump into the conversation about is getting too long.

Yep, I've been quite blindsided by this as well. Interesting to still see such animated discussions after all these years.

The AI doesn't seem to take negative yields for buildings into account or at least not properly. I did a test adding a -5 "coins" (the thing that can be changed into both science and tax revenue, I was editing the XML so I was guessing the field) yield to the warehouse, and I confirmed it applies the penalty. However, one of the AI cities still ended up building one. While tying a cost to some of the most important buildings shouldn't lead to the AI making gross miscalculations, it still could lead to misevaluations on what's the best next build. Tying it to more optional buildings would cripple the AI.

Might well be there is no code in place to evaluate negative yields at all. I don't recall there being any such buildings (at least none where it isn't more or less symbolic and dwarfed by the actual bonus), so nobody might have bothered to account for that in the first place.

How do I disable the custom RI textures? The installer for the release version offered an option but on SVN I suppose I need to remove some files.

Delete RealisticTerrain.FPK (it will spring back to life each time you update SVN though, so be prepared for that).

Regarding the music, I suppose I was just talking about the default Civ4 selection, as it seems the Civ3 music I mentioned came packaged with Warlords. Somehow I don't recall hearing it when playing AdvCiv.

A lot of older mods failed to include some stuff that was added to Civ 4 after initial release. For instance, I only recently discovered the limits on air unit capacity that were introduced in BtS and that we completely failed to port over to our mod back then.

 

[AllTheLand]

I read your entire post with interest but I'll only answer a few points for now.

Yep, I am one person without an army of translators at my disposal. I only actively maintain the English version. A while ago a kind person did a full German translation as well, so I try keeping it up to date since then too (I will freely admit I mostly use Google Translate for that; I do speak some German, but having been a professional translator earlier in my life, doing it in my free time feels too much like work, so I "delegate" it to Google and only sense-check so that it doesn't output active nonsense
).

I completely understand this. I only wanted to point out that the strings for the other languages manage to be worse than just falling back to the english string.

If you have say a pike, an archer, a horseman and an axeman on a tile, and they all get called "western" or "polish", it's extremely confusing as you have no clue what units are actually in the stack.

Is there any process for translation? Editing the XML directly is exceedingly unpleasant, doubly so if there is a need to worry about special characters encoding.

That's by design; certain such buildings (Trading Posts, Jewellers) are additional boosts you only get if you have resources that are supposed to be scarce.

I have perceived those quite differently. If you don't have any gemstones or precious metals, you can't make use of the jeweller. If you have some, you can boost your economic output slightly in the places that produce them and get happiness elsewhere. Having more luxury resources is always appreciable, but missing out on the jeweller doesn't feel too bad.

The same could be said of other resource-tied buildings, except for the Tailor, because +10% gold in every city is quite significant. Things were different when the buildings each had their tiny +2% gold bonus.

Considering that Tailors largely existed independently of the local availability of silk (and to a good degree fur, but fur is relatively more abundant unless you restrict it to specific highly-sought after luxury furs ; the Europe scenario has quite a lot of furs placed around and rightly so), my suggestion would be to remove the resource requirement to get the building done - you already need to get a tech that doesn't have really any other purpose except to unlock future techs. So a Tailor with no resources would give the gold but no happiness, and you'd need it to get the happiness from fur or silk.

I still think fur ought to be more frequent. The 0.5 per player and the two 25 RandApp don't create much (I've read up on the RandApp system - a number is generated between 0 and the max value defined there, but how is the numerical value interpreted in terms of how many resources are placed on the map?)

I will also point out a further issue (I just discovered one hour ago about the "CIV4BonusInfos.xml" system): a lot of resources are distributed "per player" rather than per tiles. Which makes sense to some degree, but once you get outside of the narrow "default" parameters it's quite fragile. Personally, I find it interesting to have empty space that gets filled by Barbarian cities that settle in new civilizations. It happens quite naturally on a large-sized map, but the new settling civilizations, or the breakaway civilization from separatism, appearing after the map had been generated, don't trigger the spawning of the per-player resources. This feeds the excessive resource-scarcity for some key resources I have been concerned about. I was really lucky in my Germany game to have two limestones available to make paved roads and get the construction bonus for some buildings, some civs didn't even have one (before I went into the world editor and added a bunch of them).

Delete RealisticTerrain.FPK (it will spring back to life each time you update SVN though, so be prepared for that).

Thank you.

Again, to be clear, the design decision here was that horsemen are supposed to be "bad cavalry" for the civs that can't build horse archers. They have relatively limited usefulness by design - having them rather than horse archers is supposed to be a drawback, and their main intended use is not direct combat engagements but scouting, pillaging and combat aid (basically how pre-Hellenic era Greeks or early Romans used their very limited cavalry).

This is quite helpful in clearing up the intended design. I didn't even realize they were there instead of horse archers as I didn't play a civ that has them yet.