Realism Invictus

[Sathar]

Explained by the last two posts, but I'll say it out loud:

When attacking a non-adjacent space, it is always good practice to first move adjacent and then attack.

I'd rather have a defensible stack waste their turn standing someplace I don't want to be, than to have that lone valuable unit charge forth only to die because the real odds were 0.2 and not 0.99.

 

[le_harv]

I know this isn't the current SVN but I am getting a repeatable crash on this autosave with version 5079. It will CTD at the end of the turn or at least it does on my machine.

 

[Walter Hawkwood]

Hi! I want to talk about the world map (huge) again. But now I got some statistics! I ran 10 AI autoplay games (1000 turns) on SVN 5076. Here are the details. Unfortunately games take a lot of time to calculate but even with 10 games one can see the regularities.
1) The research speed is too high. At least mid renaissance by 1380 AD feels wrong.
2) These (non-minor) civs get destroyed at least half of the time: Arabia, Germany, Poland, Turkey, China, Mongolia.
3) These civs make it into the top 5 at least half of the time: Austronesia, India, Rome, South China.
4) Some civs fall in both categories with the notable example: India (5 times in top-5, 4 times destroyed).
5) These minor civs get destroyed often: Nubia (8 times), Khmer (7 times).
I hope these findings will help balance the world map scenario though the mod is not about it.

The research speed is balanced for regular games, not for that particular scenario. It is predictably higher there as there are more civs and consequently more tech transfer going on. I could try lowering the research rate for the scenario-only map size, but it is shared with smaller maps too, so I don't want setting it too low.

As for civs that constantly over/underperform: while I try to create a semblance of balance, the Earth map is inherently imbalanced, since starting positions are vastly unequal. Germany and Poland in particular can't be helped much, since they just have too many neighbors on all sides, and just get dogpiled from all sides. I make some tweaks now and then to underperformers, but there is only so much I can do. As for overperforming civs, it is semi-intentional. I like to have some civs vastly important to world history, like Rome, China (at least one of them) and India actually play an important part in most games.

Ever considered adding rebellions to Anarchy, when you are changing civics there should be a chance that rebel forces will spawn just like in slavery or serfdom. Of course anarchy last shorter then you run slavery/serfdom so chance should be higher, and the amount of spawned units should be larger. Also, maybe spawn rebel units after spy action "initiate city revolt", it would also by cool. regarding tech transfer, did you discarded "has trade route" requirement to tech transfer? I remember discussion about that a while back, and i thought it was great idea, did it died?

Some interesting ideas here, thanks! As for trade routes, we haven't yet touched tech transfer in any fundamental way since forever. Probably need to.

Lovely mod, really put a lot of effort into this so well done. A good range of leaders, good use of bonuses and resources (though you may have neutered cottage a bit to much) and a selection of buildings to choose from.

Would benefit from Influence Driven War I think, make fronts more mobile.

Some tweaks I'll be making to my version are to lower the number of world wonders per city, last game it seems like the English managed to build everything in London and basically eclipse anyone but me as a result of this. Also lowered upgrade costs slightly.

I like Influence Driven War as a concept, but I am not sure if it will ever get added due to the amount of effort required to do so. As for wonder hoarding, that's something I'd love to tweak so it wouldn't happen all the time.

When we build many of one type of unit, say archers, the next archer requires more production to build. Couldn't something similar be done with Great wonders?

Yeah, I'm contemplating something like that.

Resumed playing after a few months hiatus, so this might already have been pointed out:

Pastoral Nomadism's 50% extra cost to Wonder building feels a pretty good way to balance it... However, since it doesn't effect hammers used towards building the wonder (ie. -33% penalty towards production) but rather increases the cost of the wonder itself, this can be exploited somewhat. If you're already over the required normal hammer limit, switching civics will cause you to finish the wonder after anarchy is over. Even if you still had several turns to go on nomadism. Switching back and forth like this does carry a penalty, but if you have the Spiritual leader trait you only suffer a sub-optimal resource yield those few turns it takes to switch governments again, but no anarchy.

A very interesting observation, thanks! The effect could probably be tweaked to impact actual production rather than costs.

I know this isn't the current SVN but I am getting a repeatable crash on this autosave with version 5079. It will CTD at the end of the turn or at least it does on my machine.

Well, I didn't manage to find an actual cause of the CTD (most likely having to do with some AI calculations), but I've been able to advance it to the next turn without crashing for at least 10 next turns. So if you want to continue, here is your save. And we'll likely be looking at that CTD in more depth.

 

[OldBoomer]

Also, maybe spawn rebel units after spy action "initiate city revolt..."

I also really like this idea.

@ Walter - As it is now I think the city revolts are too overpowered, so much so that I never use siege units anymore. I don't think you should be able to take a city's defenses down to 0% in 1 turn. I think it would work a lot better if the revolt took place over several turns and each turn the city's defenses were reduced by a random amount. The units inside the city could also have chance of taking a small amount of damage (similarly to how collateral damage works) each turn and how many turns the revolt would last could be partially random and partially depend on how many units are garrisoned in that city. This way in order to take a city's defenses to 0% with a revolt you'd need, time, a bit of luck, and a city without a large garrison.

 

[Dr Roach]

Why isn't there a Khmer or Southeast Asian civilization? It seems like not having one would be a pretty big hole missing. Is it just lack of time or development skill, or is there a design-philosophy objection?

 

[teks]

Why isn't there a Khmer or Southeast Asian civilization? It seems like not having one would be a pretty big hole missing. Is it just lack of time or development skill, or is there a design-philosophy objection?

South China.

If a civ doesn't exist it's normally because their full history can't be fleshed out into a pack of unique units and buildings. I think thats the case for Khmer.

 

[Walter Hawkwood]

I also really like this idea.

@ Walter - As it is now I think the city revolts are too overpowered, so much so that I never use siege units anymore. I don't think you should be able to take a city's defenses down to 0% in 1 turn. I think it would work a lot better if the revolt took place over several turns and each turn the city's defenses were reduced by a random amount. The units inside the city could also have chance of taking a small amount of damage (similarly to how collateral damage works) each turn and how many turns the revolt would last could be partially random and partially depend on how many units are garrisoned in that city. This way in order to take a city's defenses to 0% with a revolt you'd need, time, a bit of luck, and a city without a large garrison.

Thanks for the insight! I don't use espionage much if at all myself, so I welcome that observation and I'll think about what can be done.

Why isn't there a Khmer or Southeast Asian civilization? It seems like not having one would be a pretty big hole missing. Is it just lack of time or development skill, or is there a design-philosophy objection?

Basically what teks pointed out. Since RI, as most other mods, includes a lot of community-made content, there was simply not enough stuff available to flesh out Khmer. They actually were the likeliest civ to be included as playable next, but the that was more or less the exact moment that the guy making excellent unit packs chose to disappear without finishing them. Also, Austronesians are actually THE largest Southeast Asian group of peoples, and they are in, so the region isn't unrepresented at all.

 

[Sathar]

On the City Revolt point: now that Spy units cannot enter enemy-occupied squares, which effectively includes all enemy cities, wartime Espionage acts are far less common. Coupled with the national limit of 3 Spies (which will need to be embedded before the fighting starts), this just won't come up that often anymore.

 

[OldBoomer]

Thanks for the insight! I don't use espionage much if at all myself, so I welcome that observation and I'll think about what can be done.

Thanks!

On the City Revolt point: now that Spy units cannot enter enemy-occupied squares, which effectively includes all enemy cities, wartime Espionage acts are far less common. Coupled with the national limit of 3 Spies (which will need to be embedded before the fighting starts), this just won't come up that often anymore.

I didn't know this. I haven't played the recent SVN. The thing is that it doesn't even need to come up often to be a problem. As spy revolts are now you can take a strategically key, well fortified city in a single turn that you otherwise wouldn't be able to take and this sort of thing often does dictate who will win that particular war.

@ Walter - What was the reasoning for this change? I like the idea of limiting the number of spies (because the AI would often spam spies which was quite annoying), but I personally think wartime spying is an important part of the espionage system and that you should be able to enter enemy territory with spies.

Also since I;m on the topic, one more thing about the espionage AI that always annoyed me was when "friendly," or "pleased" civs would engage in really aggressive and hostile espionage acts against you, for example poisoning your water supply, instigating revolts, sabotage, etc. These are acts of war and diplomatically should be treated much more severely than "-1 your spy was caught causing trouble" - I don't know if this is something that would be possible. I'd also like to suggest making civs who are "friendly," or "pleased" far less likely to engage in exorbitantly hostile espionage missions and more likely to go for things such as stealing techs, influencing civics and religion, etc.

 

[teks]

I agree that wartime spying is a good thing and the resource cost balances it out well enough. People who don't use spies complain how its impossible to wage war in medieval times. I never found enemy spies to be much of a problem either. I just place a spy in every city and its normally enough. With spy limits I couldn't defend my cities from spies so effectively.

 

[plasmacannon]

In my last game I had about 7 spies out spawnbusting to keep the raging barbarians at bay.
The number is going to be reduced to 3?
Oh, that's not good. ;(

 

[Shuikkanen]

It occurs to me that with the unit upgrade cost changes the progressive trait got a considerable buff. Maybe its time to lower its upgrade discount from 50% a bit.

 

[sazhdapec]

It occurs to me that with the unit upgrade cost changes the progressive trait got a considerable buff. Maybe its time to lower its upgrade discount from 50% a bit.

What is the purpose of this? Did you notice a steep increase in performance of progressive civs led by AI? Are progressive civs easier to play than others? Even with such a large but still conditional bonus is progressive better than seafarer or financial? It was a great coincidence (or clever design) that the progressive trait had been buffed by unit cost increase mechanics. Moreover this part of a trait begins to shine the more units you upgrade which means that there can be situations when you won't use this bonus for a long time.

 

[Sathar]

Grenadier requires Black Powder. When Repeating Rifles are researched Powder Mills are disabled causing Macemen to be the build-able Shock unit until War Factories are built/Semiauto Firearms become available.

Suggestion: Grenadiers require Black Powder or Firearms.

Also, Foreign-Trained Infantry doesn't appear right. I am the tech leader and am still at Trench Infantry (Str 18), upon building a War Factory I can now produce Foreign-Trained Infantry even though no one has unlocked Str 20 Infantry yet.

 

[Walter Hawkwood]

@ Walter - What was the reasoning for this change? I like the idea of limiting the number of spies (because the AI would often spam spies which was quite annoying), but I personally think wartime spying is an important part of the espionage system and that you should be able to enter enemy territory with spies.

That's an unintended side effect of the fact that spies are now visible to barbarians, to prevent the exploit of free fogbusting or exploring with a unit nobody can touch. You can still enter enemy territory at war, just not cities - so at war spies are still useful for scouting.

Also since I;m on the topic, one more thing about the espionage AI that always annoyed me was when "friendly," or "pleased" civs would engage in really aggressive and hostile espionage acts against you, for example poisoning your water supply, instigating revolts, sabotage, etc. These are acts of war and diplomatically should be treated much more severely than "-1 your spy was caught causing trouble" - I don't know if this is something that would be possible. I'd also like to suggest making civs who are "friendly," or "pleased" far less likely to engage in exorbitantly hostile espionage missions and more likely to go for things such as stealing techs, influencing civics and religion, etc.

That's AI stuff, and AI stuff is scary.

In my last game I had about 7 spies out spawnbusting to keep the raging barbarians at bay.
The number is going to be reduced to 3?
Oh, that's not good. ;(

More than that, barbarians are now going to eat your spies for breakfast.

Grenadier requires Black Powder. When Repeating Rifles are researched Powder Mills are disabled causing Macemen to be the build-able Shock unit until War Factories are built/Semiauto Firearms become available.

Suggestion: Grenadiers require Black Powder or Firearms.

Also, Foreign-Trained Infantry doesn't appear right. I am the tech leader and am still at Trench Infantry (Str 18), upon building a War Factory I can now produce Foreign-Trained Infantry even though no one has unlocked Str 20 Infantry yet.

Thanks, excellent observations! All fixed.

 

[plasmacannon]

More than that, barbarians are now going to eat your spies for breakfast.

My first response is "Oh Great!" noting how my spawnbusting will become more difficult.
Then, my creative mind starts thinking of naming spies after food.
Green eggs and ham, and 00porkchop. LOL

 

[plasmacannon]

Hey. Any chance that some new leaders will be added in the next update?
I've been interested in trying a Spiritual/Progressive leader and there isn't one.
Also, there are several trait combinations that only one leader has.
It would be nice to try them on another empire other than by using unrestrictive leaders.
Any changes to existing leaders or traits?

 

[Walter Hawkwood]

Actually guys, I realized I want to ask everyone (especially SVN players) their more general opinion. What in general you don't like? What civs feel uninteresting/underpowered? What features seem annoying? Can't promise to fix everything, but you can definitely give me a direction to look.

Hey. Any chance that some new leaders will be added in the next update?
I've been interested in trying a Spiritual/Progressive leader and there isn't one.
Also, there are several trait combinations that only one leader has.
It would be nice to try them on another empire other than by using unrestrictive leaders.
Any changes to existing leaders or traits?

New additions are unlikely at this point, as they have to come with leader-specific city lists, which is a lot of work even with one leader. Reshuffling some traits to cover the missing combo is a more realistic possibility, but I'll see what I can do. Sometimes a good candidate for a particular combo just doesn't come to mind...

 

[Shuikkanen]

I'll probably respond to this with a longer post later, but right now I just want to mention the one thing that makes me want to punch my fist through the monitor most when it happens:

Discover Alphabet.
Hurricane destroys Storyteller Circle.

I 100% never fail to enable cheat mode and just add the circle back into the city if this ever happens. So much do I hate it. I guess the roots of this problems lie from when Schools were moved from requiring Alphabet into a Medieval era building. It does make sense that Alphabet discontinues storytelling from a flavor standpoint, but one now has to wait for a long time for the replacement building. Maybe something in-between, like a Scribe?

 

[Walter Hawkwood]

I'll probably respond to this with a longer post later, but right now I just want to mention the one thing that makes me want to punch my fist through the monitor most when it happens:

Discover Alphabet.
Hurricane destroys Storyteller Circle.

I 100% never fail to enable cheat mode and just add the circle back into the city if this ever happens. So much do I hate it. I guess the roots of this problems lie from when Schools were moved from requiring Alphabet into a Medieval era building. It does make sense that Alphabet discontinues storytelling from a flavor standpoint, but one now has to wait for a long time for the replacement building. Maybe something in-between, like a Scribe?

Well... No, this one is actually by design. Storytellers are from the start intended to go discontinued very early on, so a civ only has them in a few starting cities. Flat-rate buildings are all similar in this respect. What I should probably do to make it less annoying, though, is make it a limited building that doesn't go obsolete until schools.