Apolyton University Mod (PTW version)

[Dominae]

Playing AU203 I noticed a significant delay in the time when I could begin producing the Forbidden Palace (5 extra cities, to be exact). Although this did not affect my strategy much in this particular game, I can see it being quite relevant on other maps, or when using looser city spacing that I did. Something to keep in mind, I guess.

But yes, 20 cities is too many on a Standard map.


Dominae

[alexman]

Increasing the number of cities needed for the FP by 150% changes the flavor of the game for sure. That's why I removed that change from 1.15, even though it helps the AI in three ways: It increases the likelyhood that the AI might build the FP further away from its capital, it removes a human-only powerful corruption tool from the early part of the game, and it encourages the AI to get enough territory to threaten domination.

[Dominae]

Ok, I just finished AU203, and thought I would share my observations on the new Communism. I know we will not be keeping the current version because of the unfortunate Despotism/Monarchy side effect, but I believe my screenshots are informative and interesting nonetheless (you be the judge!).

The first screen is from the last turn I was in Republic (I've just discovered Motorized Transportation, and am hoping that Flight will not take too long under Communism). The second screen is from my first turn of Communism, only 4 turns of Anarchy later.

Of particular note are the similar Shield outputs and comparable research times (I was expecting research to come to a crawl, to be honest). I had not begun claiming any new land yet (recall the 203 rules), but with my tight spacing I'm quite sure I was over the OCN.

Do we want Communism to be this good? Is this good, or am I just surprised? Note that I was at war (offensively) for exactly 61 turns until the game was over, which is simply not possible with Republic or Democracy.


Dominae

[Dominae]

Republic:

[Dominae]

Communism:

[alexman]

If you ask me, the difference in science output is fine in those screenshots. Republic is almost twice as good for research.

A really good test would be to compare Monarchy to Communism, as those are the two war-time governments. If they are close, it's good. If one or the other is clearly better (especially if that's the ancient Monarchy), then something needs to change. Do you have a save from just before Anarchy was over? You can compare all governments in the same turn that way.

[kettyo]

Originally posted by alexman
If you ask me, the difference in science output is fine in those screenshots. Republic is almost twice as good for research.

Not really.
Republic has -90 gold income while communism has -10 gold.
So it's unfair to compare their science output.
They could only be compared fairly with the nearest gold income values possible.

[kettyo]

Originally posted by alexman
Increasing the number of cities needed for the FP by 150% changes the flavor of the game for sure. That's why I removed that change from 1.15, even though it helps the AI in three ways: It increases the likelyhood that the AI might build the FP further away from its capital, it removes a human-only powerful corruption tool from the early part of the game, and it encourages the AI to get enough territory to threaten domination.

But you'll keep the +50% increase introduced with 1.13 won't you? I feel the +50% cities until FP is no problem.

[ducki]

Originally posted by Dominae
Note that I was at war (offensively) for exactly 61 turns until the game was over, which is simply not possible with Republic or Democracy.

I've actually had some extremely long wars in Republic, though I don't have exact durations, and while protracted wars are normally "impossible", if you have enough "good happiness stuff", it can be done.

Again, I don't have definite numbers, so this is anecdotal(sp? ), but I believe it could be done.

It may, however, be a factor of difficulty level (and "born content" citizens).

[Dominae]

alexman, I'll take a look at Monarchy when I get the chance. I guess the Science issue is where it should be. But what about the Shield output? I've got almost no Corruption!

kettyo, given that the Science slider is maxed out in both examples, the comparison is fair. Just imagine that Communism is earning +80gpt and Republic 0gpt.

ducki, 61 turns is a long time to be the aggressor. Crippling WW would have happened at turn 30 at most (I had all Luxuries save one, Marketplaces and Temples everywhere, Women's Suffrage, Bach's). Being able to declare war and go through with it in one big go is a nice change from my usual games, and is a pretty powerful option. Have you reached Communism in your game? When you do, I think you'll see what I mean (Luxury slider, who needs that!?!).


Dominae

[ducki]

Maybe the difference is that I normally put Cathedrals everywhere and work really hard to get Sistine as well as HG.

I also proactively use the Lux slider and Entertainers as needed.

Perhaps Sistine + Cathedrals(rushed if necessary) in any city that has unhappiness is enough to explain the difference - that and the jump from Regent to Monarch?

[Dominae]

I haven't tested out all the possibilities, but on Emperor, with all the happiness goodies you can ever have, I think you would have a tough time warmongering like I did if the government was Republic. Cathedrals and Sistine should prolong the inevitable by a few turns, but ultimately WW would win out in 60 turns (1400-1700AD).

It would be interesting to figure out exactly how long you can go in the scenario above before WW eats you up. This raises another question: does WW scale with the number of battles, number of units inside enemy territory, etc. or is it just a constant factor for each civ? For example, if I have 1 Infantry inside China's borders (under Republic) will this generate as much WW as if I had a stack of 20 Infantry there?


Dominae

[Stuie]

From what I've seen, WW is reduced by military successes and increased by failures. Granted I play on Monarch, but I've been able to maintain a Democratic government during a protracted war as long as I continue to have successful engagements with my enemies. Note that I also make sure that my luxuries are maxed out, but I never use the luxury slider.

[alexman]

Originally posted by Dominae
But what about the Shield output? I've got almost no Corruption!

Looking at your Iron Works city, it has 102 blue shields, and 4 red ones. Since it has a 200% bonus to each shield, if there were no corruption you would get 12 extra shields. That's about 10% corruption. I think that's OK for an empire slightly larger than the OCN, especially given how low the OCN is compared to human empires in usual games.

[ducki]

Anecdotally, I can also agree with Stuie - when things go well, or when I am winning a purely defensive war, I get little to no WW.

When I am the aggressor, or when I lose a lot, WW seems to eat me up.

Also, razing seems to give me more WW than Take-n-hold offensives.

Again, this is purely anecdotal, supported by no numbers or SAVs, it just seems that way.

[Dominae]

alexman, I guess I was just expecting Communism to be a lot less efficient. The current version is a good model of what we want then, if not for the Despotism problem.


Dominae

[joncnunn]

WW is based on unit loses on both sides outside your Territory and unit loses on your side only inside your territory.

And if the AI declared the war instead of you, you get a negative WW rating initallly.

It also appeared to be calculated seperately for each AI your at war with based on my experence in the previous game where I was technically at war with 2 AIs but almost all my fighting was with 1 of them and when I signed a peace deal with the 1 I was actually fighting, the citizens stoped protesting about the war either though I was still technically at war with another AI.

[Catt]

I rely on the corruption nazi to help me when I am befuddled by corruption issues.

Here is a screenshot from AU 203 -- it shows the city view of a recently captured Chinese city. I have also superimposed a minimap from the same year, showing the German territory. I am playing the AU Mod (v1.14 I believe) on Emperor. It is a standard map. Without dealing with the AU Mod changes just yet, I figure that the difficulty-level modified OCN should be 12 (80% of 16, rounded down). Communism normally increases the OCN by 25% -- so it should be either 15 or 16. I count 33 cities, meaning I have more than 2x the modified OCN. I am Communist, meaning no distance corruption but communal corruption.

Here's my confusion. Note that the city at issue does not have either a courthouse or a police station. It is not in WLTKD (for waste). But it is only experiencing 33% corruption. The adjustment to communal corruption in the AU Mod probably explains some of the better-than-expected productivity, but even so, shouldn't the corruption in this city be far higher?

Catt

[Catt]

Originally posted by joncnunn
WW is based on unit loses on both sides outside your Territory and unit loses on your side only inside your territory.

You seem pretty confident in this view -- have you done testing or seen something that leads you to this conclusion? I have an abiding interest in WW, partly because it is so hard to predict and very hard to test for the formula, so if you have some data I'd love to see it.

The manual, IIRC, states that WW is caused by: (1) units in enemy territory; (2) enemy units in your territory; and (3) engaging in combat. Of course we all know how accurate the manual is . Nonetheless, I have never seen anything conclusive or convincing that shows WW is more directly affected by unit losses than simple combat (i.e., I think WW can grow at a good clip without many unit losses). Furthermore, it seems pretty clear that WW also grows just by having units in enemy territory - without any combat at all. A poster named sumthinelse posted some test results just recently over at CFC -- he was trying to understand whether and to what extent WW is different in MP games compared to SP games. For one test, he created a scenario and put something like 100 of his units in enemy territory. He declared war, but engaged in no fighting whatsoever. In SP games, WW quickly caused a revolt -- again with no combat.

Absent comment from Firaxis or a clear test, I still believe that the manual is mostly right (although I'm open to the view that losing cities or citizens may also have profound WW effects).

Catt

[Dominae]

I thought there was no WW in MP (without bots). My experience is limited, but this seems to be what people are saying (more like complaining about!).


Dominae

[Catt]

Originally posted by Dominae
I thought there was no WW in MP (without bots). My experience is limited, but this seems to be what people are saying (more like complaining about!).

That's what I understand, too. sumthinelse's tests were focused on the "WW in MP??" issue, so he tested both SP and MP games under the same scenario -- in MP no WW, in SP plenty of WW. I just "borrowed" his test results to highlight the related issue that military forces in the wrong territory pretty clearly seem to result in WW, with or without combat.

Catt

[alexman]

Catt, thanks for continuing to try to shed some light on the Communism issue.

If you plug in the numbers, the corruption formula gives about 45% corruption for the city in question. (Rounded to 25% communal plus 20% from cities). In this version of the AU mod, we reduced the flat corruption for communism from 30% to 20%. Since you don't have a courthouse or police station to reduce that value, your 10% less corruption in that city is directly the result of the 1/3 reduction in communal corruption.

No WW in MP?

[Dominae]

Originally posted by alexman
No WW in MP?

Well, that reduces the playable governments to 2 (maybe 3 if you're Religious). Something someone (like me) should bring to Firaxis' attention.


Dominae

[Catt]

Originally posted by alexman
Catt, thanks for continuing to try to shed some light on the Communism issue.

If you plug in the numbers, the corruption formula gives about 45% corruption for the city in question. (Rounded to 25% communal plus 20% from cities). In this version of the AU mod, we reduced the flat corruption for communism from 30% to 20%. Since you don't have a courthouse or police station to reduce that value, your 10% less corruption in that city is directly the result of the 1/3 reduction in communal corruption.

Thanks alexman! Like many things requiring fairly complex formulas to decipher, I frequently duck the actual work of making sure I understand it all and latch on to certain "pieces" of knowledge that distort rather than inform the expected outcome.

In this case, I couldn't get over the fact that my city count seemed far enough beyond the OCN that the city in question should, by virtue of that fact alone, be almost terminally corrupt (the "piece" of knowledge in this case being that cities at 1.5x the modified OCN will be 95% corrupt - a more useful "piece" of info in non-communal governments). Which once again highlights my shockingly, near-total lack of understanding about how the corruption model actually works. I finally downloaded your corruption calculator and was playing with it -- I'm shocked, actually. I could have 80 cities on the AU 203 game and still my corruption, without a courthouse or police station, would be 75% in any given city -- I would have guessed that no more than 50 cities or so would trigger near total corruption.

Catt

[Nor Me]

On the corruption formula, I was trying to test some things with Communism. While the formula is approximately accurate, there were some anomalies.
Things like the Commercial trait not having the same effect as the FP and odd numbers of cities having the same corruption as the next lowest even number (just like with luxuries.)
I did find that the number-of-city-based corruption in Communism can be reduced by another FP. So if any Kremlin sort of idea is floating around making it into an FP will help Communism

Yes, AU203 Communism was good. With the increased OCN, the Chinese certainly made good use of it. At the time of this minimap, the Chinese only have 40% corruption in their unimproved cities! This is a randomly selected city:

[Nor Me]

?

[alexman]

NorMe, the corruption formula is not exact, you're right. But it has evolved to be pretty damn close. I'm convinced that Firaxis is using integers for internal calculations, which suffer from lost accuracy with each division. That would explain those jumps you see (which I also noticed while I was testing), and other weird behavior in the game.

To return on-topic for this thread, that's fantastic news about multiple FPs having cumulative effect for Communism. Unfortunately, I'm not sure government-specific improvements with the small wonder properties work though. I know great wonder properties don't (I gave the Kremlin free maintenance for trade buildings, and still I got the benefits outside Communism). Perhaps the FP will work, since in non-communist governments only one FP works, so the lasting effects outside communism won't matter. We'll have to test that.

[Nor Me]

Originally posted by alexman
Unfortunately, I'm not sure government-specific improvements with the small wonder properties work though.

Your right in this case .
The only way I can see around this is to force both FPs to be in the same city.

[joncnunn]

Manual is definately wrong on #2:

I had a huge stack of Infenty marching thru my territory during the Age of Defense with no WW whatsoever.

(They were trying to reach a city that I had left undefended as bait.)

I was occansionly picking off an isolated Infentry. (And on the first turn, killed all enemyCalvary in my territory.) I suffered no combat losses what-soever and by the end of the war my citizens were still happier than when the war started.

Evenually the AI got tired of fighting (and was suffering WW problems) and settled the war by giving me a city on the island just offshore that they had captured from one of my allies in which I owned the other cities, all their Gold, and some GPT.

On quick offensive wars with low intensity under Republic / Democracy I've suffered very little War Werrious.

Wars in name only I've also had no war werrious whatsoever.

On high intensity offensive wars under Democracy, I quickly get a lot of WW built up until I either sign a peace treay with that nation or else switch governments.

I haven't tested losing a city. (In fact, in all my games in Civ III, I've never lost a city in combat.) This is largely because if I think I can't hold an enemy city it gets razed instead of captured, and on defense, I use very key deployments of troops including which cities NOT to place units in which the AI falls for.

Originally posted by Catt

You seem pretty confident in this view -- have you done testing or seen something that leads you to this conclusion? I have an abiding interest in WW, partly because it is so hard to predict and very hard to test for the formula, so if you have some data I'd love to see it.

The manual, IIRC, states that WW is caused by: (1) units in enemy territory; (2) enemy units in your territory; and (3) engaging in combat. Of course we all know how accurate the manual is . Nonetheless, I have never seen anything conclusive or convincing that shows WW is more directly affected by unit losses than simple combat (i.e., I think WW can grow at a good clip without many unit losses). Furthermore, it seems pretty clear that WW also grows just by having units in enemy territory - without any combat at all. A poster named sumthinelse posted some test results just recently over at CFC -- he was trying to understand whether and to what extent WW is different in MP games compared to SP games. For one test, he created a scenario and put something like 100 of his units in enemy territory. He declared war, but engaged in no fighting whatsoever. In SP games, WW quickly caused a revolt -- again with no combat.

Absent comment from Firaxis or a clear test, I still believe that the manual is mostly right (although I'm open to the view that losing cities or citizens may also have profound WW effects).

Catt

[Catt]

Originally posted by joncnunn
Manual is definately wrong on #2:

I had a huge stack of Infenty marching thru my territory during the Age of Defense with no WW whatsoever.

(They were trying to reach a city that I had left undefended as bait.)

While I acknowledge that perhaps enemy troops in-country may produce less WW than other factors, I am pretty convinced that they do add to WW. I tend to shy away from anecdotal evidence on WW simply because so many factors are at play that it is hard to isolate and diagnose WW causes. I have seen and heard other anecdotal stories that would strongly imply that enemy troops does add to WW. More imp[ortantly, though I'm not certain of the test parameters, I believe that sumthinelse's WW tests included testing for enemy troops in his territory -- i.e., he witnessed growing WW from enemy troops in his territory, even without any combat, unit losses, or offensive forces in enemy territory.

Catt