Balanced Game Playthrough, Strategy & Analysis

[Txurce]

Following up on my earlier post, I won a Science victory in 1884. This is two turns slower than my best, achieved via Darius GAs. But I dominated like in no other Civ5 game so far. Interesting facts:

I was about two eras ahead of the pack, first in everything except land and soldiers, with a literacy edge of 88% to 64%.

I finished with a population of 53.3 in 9 cities, vs a 17.1 avg. My previous high was less than half my total. Does this mean fertilizer - to which I beelined - is overpowered, especially on top of the smokehouse, enhanced mills, and diluted maritimes? My pop doubled in the 20 turns after hitting fertilizer. I would say that these mods have made going for pop with a contained number of cities a very viable startegy.

I went from Wonders in Tradition to full Patronage to half of Rationalism, finishing with Order (?) for a spaceship production boost. In another game I would try converting my 100% farm development on grass and plains to many more TPs, once I have the Rationalism TP policy up next.

Combat was the other area of note. I had a much harder time taking walled cities. Even two mech infantries needed luck to take a size 24 walled city, despite the civ still being pre-industrial. In all I liked the mod forcing me to use siege weapons (next time!).

One possible development rising from the revised city combat system is that there was no runaway civ in the other continent. In fact, 4 of 5 were dead even, with the fifth only a notch below. I wonder how much of this is due to cities bring harder to take - meaning they became next to impossible for the AI?


There's no doubt that the lack of runaway civs makes the game more fun. It also helps the human player not just in having less of a challenge, but in having radically larger trade opportunities. Despite my ever-increasing lead, I stayed friends with the other continents' civs until the nineteenth century - a span of 800 years - when two soured somewhat and a third declared long-distance war. Having said all that, I'd be a little disappointed if the new balance in the other continent were to be a near-ironclad rule. The ideal for me would be the occasional runaway... but I suspect that would require better AI tactics.

 

[Thalassicus]

I've been thinking about doubling the research adjustments in Balance - Research from 0-30% to 0-60%. The modern era in particular seems to still go too fast even with +25% cost, I was popping them out every 4 turns. This would make it every 5 turns. Not a big difference, but meaningful when applied to all techs of the era.

I've found that although there doesn't seem to be a runaway civ conquering the whole continent anymore (primarily due to the ratio change between open:rough terrain) there's often 2 or 3 superpowers that gobble up their neighbors. This happened in the last game I posted a screenshot of. If you look at the minimap, Ramkha and Catherine (in the north) killed their neighbors Napoleon and Montezuma.

 

[Perkus]

Using Harder Free Tech will slow the science down somewhat, but probably at best by 10%. I agree that it's still way too fast. I was getting future tech every 3 turns in my last science win game. I can't remember if I had gradual research already in that one or not, but either way, it was crazy fast. One of the main flaws of the modern era now is that it flies by so fast, and usually you are not even warring any more, so all these neat new units go unused.

 

[learner gamer]

I've been thinking about doubling the research adjustments in Balance - Research from 0-30% to 0-60%. The modern era in particular seems to still go too fast even with +25% cost, I was popping them out every 4 turns. This would make it every 5 turns. Not a big difference, but meaningful when applied to all techs of the era.

@Thalassicus: Funnily enough, this is the very reason why I stopped using your mods and started using the economy mod by Valkrionn...despite all the great changes you’ve made, I still found myself teching much faster than I could build units or infrastructure. Please understand BTW that this is not a criticism of your mods at all; they do an absolutely terrific job of what they’re intended to do – provide more interesting choices whilst remaining true to Civ 5. However, I just happen to believe that Civ 5 is so badly balanced right now that re-balancing it requires making a greater departure from its mechanics than what your mods are designed to do.

Indeed, FWIW, I believe that the first person out there who creates a mod that (i) starts by re-balancing the tech tree and hammer costs on a scale similar to that undertaken by Valkrionn (although I agree with your approach Thalassicus, that later tech research times should be lengthened by more than earlier ones) (ii) makes the granary an early food storage building (allowing for faster regrowth and more working of mines to help get a critical mass of early hammers) and (iii) then adopts a selection of your excellent individual tweaks, will be the person who, IMHO, begins to unlock the real potential of Civ 5. With these kinds of changes implemented, the real issue outstanding IMHO would then be to work out how to prevent these changes from resulting in mass military unit spam in the late game. To that end, the modder could perhaps look to raising later game hammer and unit upgrade costs or use unit maintenance costs. Make no mistake, there is a great game lurking in Civ 5...it’s just that vanilla is so badly balanced right now IMHO that it’s struggling to get out.

Two last things. Firstly, if there is anyone out there considering a mod which combines features from Valkrionn’s and Thalassicus’ mods (giving credit where its due obviously), bear in mind that doing things like cutting hammer costs by 20% (as Valkrionn’s done) may mean that you don’t want to add a hammer to mines etc. at engineering. More broadly, the key issue IMHO will be to avoid making changes that unnecessarily double up / double count. Secondly, genuine kudos to you Thalassicus for coming up with your latest idea re: granting an additional social policy per era (whilst increasing the number of trees required to unlock utopia). Another problem I noticed in my playthroughs was that social policies (like much else in Civ 5) seem to come far too slowly after the initial couple are adopted IMHO; your excellent idea (albeit from the Culture Calibration mod) means they will now continue to arrive during the mid game, potentially allowing gamers to get deeper into the policy tree without necessarily allowing larger empires easy access to a culture victory.

Once again, genuine kudos for all you’re doing to help balance Civ 5.

 

[Txurce]

I've been thinking about doubling the research adjustments in Balance - Research from 0-30% to 0-60%. The modern era in particular seems to still go too fast even with +25% cost, I was popping them out every 4 turns. This would make it every 5 turns. Not a big difference, but meaningful when applied to all techs of the era.

I am all in favor of slowing late research as you propose, because having more time to use later tech is more fun. In terms of conceptual balance, though, how do you factor a late scientific civ having invested more in teching than one that hasn't, or in all likelihood more than any scientific civ from an earlier era? Should this edge still make late research for this civ faster than all earlier research, even with your increased mod? If it doesn't, wouldn't this be an unfair penalty for the effort and expense a good player has made in developing a powerhouse scientific civ?

 

[Thalassicus]

Early techs take 10-30 turns (on Epic), yet late techs come every 3-4 turns even without Rationalism, which is why even a 25-50% increase won't alter the #turns significantly per tech. Since there's five people who have mentioned this now, I'll go with a bigger increase... 1.5x current values for the modern era... seems it needs a power curve too.

Previously

• 100% - Ancient
• 105% - Classical
• 110% - Medieval
• 115% - Renaissance
• 117% - Industrial
• 120% - Modern
• 125% wait time - Research Pacts

New

• 100% - Ancient
• 105% - Classical
• 110% - Medieval
• 120% - Renaissance
• 140% - Industrial
• 180% - Modern
• 200% wait time - Research Pacts

To keep this from being a buff to lightbulbing I'll move one scientist slot from universities to research labs (2->1 and 1->2 respectively).

I'm starting to realize we really need the full sdk for two simple, but critical changes: lightbulbing gives fixed , and defensive buildings increase city hitpoints.

 

[Thalassicus]

(double post)

 

[Perkus]

We heard you the fist time

I'm skeptical about making research pacts that long (2x = 40 turns). It'll take forever to recoup your costs. And since you no longer get the tech if a war breaks out, it makes them than much riskier to invest in. Maybe 1.5x? I'm ok with moving scientists up another building, although I'm surprised you're suddenly so keen on that. As I recall you were hesitant at first to move the one from the Library.

 

[Tomice]

I'm skeptical about making research pacts that long (2x = 40 turns). It'll take forever to recoup your costs. And since you no longer get the tech if a war breaks out, it makes them than much riskier to invest in. Maybe 1.5x? I'm ok with moving scientists up another building, although I'm surprised you're suddenly so keen on that. As I recall you were hesitant at first to move the one from the Library.

I agree, and I'd rather balance them through cost than through time. Even in vanilla epic games they need so long to show results, that they feel somewhat random and not like a deliberate, active player decision.

 

[Thalassicus]

It's mostly just possible to balance through time, the "RESEARCH_AGREEMENT_TIMER" field in GlobalDefines.xml. I'll go with 150%.

I don't remember the exact reasons, but I was probably hesitant to modify libraries because they're so early. In general, I've been avoiding altering the ancient era, since early changes have exponential effects. Universities are halfway through the game so it's not a big deal. The ideal spot for the extra specialist is actually on Public Schools... but that wouldn't make much sense from realism, as universities and research labs have more jobs for scientists.

 

[Wurlf]

Hi Thal,

I'm quite new to the modding aspect of the game though I had played all civs for decades now. I find your work -and some onthers moders' too- really interesting and definitely giving the game the balance and the depth it is needing at the moment. I am afraid that the release of the game was firstly motivated by the stability of the new interface and that additional content will come later to render the all thing more integrated (as it happened for civ IV btw).

Anyway, just a question regarding your mods : I am using a french version (I know, nobody is perfect) and your added buildings or wonder (aquaduc, agra fort...) are not showing any text, or rather erraor kind of text, in the tech tree or the civilipedia. Is this to be expected or is there any language thing that would prevent me to see english text there?

Now, i'll bring my little stone to the science pace debate above. I do agree that towards the end of the game you do not have enough time, as turn-time, to enjoy playing with most of the unit or buildings. Slowing down everything seems to be a good approach to balance the game to fix this 'issue'

In the meantime, for reality stimulation purpose, one must to agree that sciences in modern time are evolving faster than in antiquity. kinda exponentially faster. So you would expect discovering a new thing more often, turn wise. Where I feel the game is unbalanced is that not every discovery should lead immediately to a technological advantage such as a new type of unit much stronger than the previous one. In fact you shoud need 5 or 10 of these techs to get to use the new unit or to use it a its full potential.

The way I see it, all the issue is down to a poor tech tree actually. It is quite linear with somehow two double branches that never ramify much. They actually fuse in future time. Modern time should have four to height more techs 'vertically' to allow you to move on the next level -not sure I am making myself clear. This means that you would need lot more prerequisite to actually progress, and also that some sciences will have less and less connection, such as medicine, physics, media etc... the bridges should remain but be more and more tiny, so researching a tech or a bunch of tech would become a more strategic decision and disadvantage you on other fronts... Of course you could not have a new unit each tech, but one can imagine that a couple of tech would provide some free promotion to such or such unit for the next round of production. Realistically new advanced weapon are sometimes less effective than old proved ones which have been tweaked every possible way. Thus, expanding science could also be a way to balance war (and food, hammering or even hapiness) the same way.

To sum up, science in modern era is progressing with to wide steps to be enjoyable and slowing down the steps might not be good enough to make them feel less 'gappy'.

Now, i know that I am well away from your mod-balancing current aims, and to go back to how slowing science in late games, well, as a sicentist myself, I do know that nowadays it takes a lot more people to produce the same scientific production (as in scientific breakthrough of the same magnitude for the society) than some times ago. Pasteur or Curie discovered a lot on their own, while sequencing the genome required dozens of equally knowledgeable scientists. My suggestion: allow more scientist slot in modern era as you propose, yes, but halve the science input of all scientists (from all buildings). This will certainly nerve things down.

Once again, thanks for your work, really appreciated.

 

[Thalassicus]

I understand you, it is a good argument: science does evolve faster. This is why I am changing modern technologies to approximately ~5 turns. It is slower than the normal game, but still faster than ancient technologies (~10-20 turns). I also decreased scientists from 3 to 2.

The next official patch will somewhat solve the language problem you describe.

Added support for fallback languages (if mod is not translated, fall-back to English so text keys are not showing).

I can not access the language files for other languages, so I can not easily translate to Français, Deutsch, Español, etc.

However, all text is in 1 place in each mod. Juances19 has translated a few to Español, you can see an example of his work in the Balance - Civilizations mod folder. (BL - Language_ES_ES.xml)


To show it in Français, basically:

1. Copy the English file
2. In the copied file, translate English text to Français

Here are details:

1. Create a copy of the file "BCD - Language_en_US.xml" in the mod folder.
2. Change the name of the new copy to "BCD - Language_FR_FR.xml".
3. Open "BCD - Language_FR_FR.xml" in notepad or notepad++.
4. There are many things like this below. New buildings are in the "National Wonders" section.
   

   <Update>
       <Where Tag="TXT_KEY_BUILDING_GRANARY" />
       <Set Text="Smokehouse" />
   </Update>
   <Row Tag="TXT_KEY_BUILDING_GRANARY_HELP">
       <Text>Each source of [ICON_RES_COW][ICON_RES_DEER][ICON_RES_SHEEP] and [ICON_RES_FISH] worked by this City produces +1 [ICON_FOOD] Food.</Text>
   </Row>

5. Replace each Text part with a translation in Français. Do not change things in between [ ]. For example, it might look something like below. I do not know Français (used an online translator) so this attempt is not correct, but hopefully demonstrates the idea.
   

   <Update>
       <Where Tag="TXT_KEY_BUILDING_GRANARY" />
       <Set Text="Maison Fumée" />
   </Update>
   <Row Tag="TXT_KEY_BUILDING_GRANARY_HELP">
       <Text>Chaque [ICON_RES_COW][ICON_RES_DEER][ICON_RES_SHEEP] et [ICON_RES_FISH] travaillé par cette Ville produit +1 [ICON_FOOD] Nourriture.
   </Row>

6. Send me the translated "BCD - Language_FR_FR.xml" file and I will add it to the mod.
   
   (You can also add it manually. I will make this easier in the next version of the mods.
   1. In Balance - City Development (v 10).modinfo create a line in the <Files> section for the new "BCD - Language_FR_FR.xml" file.
   2. Use this tool to create a md5code (copy-paste the file name to the top box and click Generate).
   3. Add a line for the file in the <OnModActivated> section.
   4. In the file "BCD - Language_FR_FR.xml", replace all "Language_en_US" with "Language_FR_FR".

 

[Perkus]

It's only possible to balance through time, the "RESEARCH_AGREEMENT_TIMER" field in GlobalDefines.xml.

Anybody know what RESEARCH_AGREEMENT_MOD does? It's 0. I wondered if maybe that affects the RA cost formula. How about BASE_RESEARCH_RATE (= 1) ?

 

[Slowpoke]

Wait, research agreement costs can be based on era, I remember doing it for my mod. Hold on.

Edit:
Yup, here you go.

<?xml version="1.0" encoding="utf-8"?>
<!-- Created by ModBuddy on 10/21/2010 2:09:04 AM -->
<GameData>
<Eras>
<Update>
<Set ResearchAgreementCost="250"/>
<Where Type="ERA_CLASSICAL"/>
</Update>
<Update>
<Set ResearchAgreementCost="450"/>
<Where Type="ERA_MEDIEVAL"/>
</Update>
<Update>
<Set ResearchAgreementCost="600"/>
<Where Type="ERA_RENAISSANCE"/>
</Update>
<Update>
<Set ResearchAgreementCost="800"/>
<Where Type="ERA_INDUSTRIAL"/>
</Update>
<Update>
<Set ResearchAgreementCost="1000"/>
<Where Type="ERA_MODERN"/>
</Update>
<Update>
<Set ResearchAgreementCost="2000"/>
<Where Type="ERA_FUTURE"/>
</Update>
</Eras>
</GameData>

 

[Perkus]

Nice, I thought I saw that being done in a mod somewhere before...

 

[Thalassicus]

Yeah, I'd forgotten about that.

The problem with by-era is it isolated vs crowded start locations. A civ with a crowded start could sign 20 research agreements in early eras, getting a big boost, and it provides more benefit since early gains give exponential returns. In contrast, if a less-crowded start doesn't have as many neighbors and can't sign most agreements until post-Astronomy, they've got to pay more for the same number of free techs, and it's later in the game where the effect won't be as amplified.

Part of it is random map generation luck, part of it is who gets killed off early. Even if you're a peaceful empire, if one of your warlike neighbors kills off a few others you'll get penalized more than someone on a politically stable continent. On top of this, isolated locations already get penalized for resource trades. I'm worried solving one problem (overpowered pacts) will just cause another (escalating isolation penalties).

The best way to balance research pacts is increasing cost by the number of agreements done. This is the method social policies, GPs, golden ages, border expansion etc all use. I don't know why they used a different method for research pacts. I've got it on my todo list for when we have full sdk access down the line.

 

[Slowpoke]

Still, I prefer increase by era actually because it relates to more of a correct gold per beaker ratio. With your ideal method, you could essentially save agreements until medeival, then dump your gold for massive gains.

 

[Thalassicus]

That's true, yet it goes back to the exponential gains point.

Saving 20 turns from a free tech in the ancient era might result in you winning the game 40 turns earlier at the end. It's a domino effect because of the amplification of early gains. In contrast, saving 20 turns on research in a later era won't have as long to amplify the effect. If you wait to sign agreements you're getting more beakers per pact, but also losing out on the exponential gains.

Basically, it needs to be balanced where isolated starts don't get penalized, and you get about the same cost/benefit ratio of a research pact in the ancient era as the modern era. In short, a combination of increasing costs per era (like the game has) and costs per pact. Since this isn't possible yet, and I want to avoid moving the problem from one area of the game to another, just waiting for the full sdk.

 

[Slowpoke]

Heh that's true, I guess I'm thinking too linearly One bad thing about the mod I'm working on is your power goes up a lot less rapid, so I forget how it works in other mods. Perhaps your way is more befitting your mods' style.

 

[Txurce]

I tested v1.0.2 with Harder Free Tech, Emigration and Liberation – essentially the package Thal put up. I used Persia to see how its GA bonus would be affected. Again, I set a science victory as my pre-game goal.

I found myself in a large peninsula in a northern continent with Paris just on the other side of the connecting bottleneck. I immediately made a critical decision to build my second city far from the capital and past the two local CS, right on the Parisian border. This effectively sealed off the French to build one city to the south (where unknown to me they were hemmed in by the Iroquois) and in a far-north connecting loop that was almost all tundra.

This kept Napoleon at bay until I was ready to whittle him down and take what I wanted, once my four cities were thriving. Like last game, I maxed out population growth, with the intent to fill as many specialist slots as possible.

The downside of my territorial strategy was that Napoleon controlled the coastal waters, so I was unable to trade or sign one RA until after astronomy. This slowed down my tech start and overall score. Still, by the time of my launch in 1940 or so, my population was at 46.9, almost twice the average. I love fertilizer! Literacy was first, at 93. When I ended the game (in a GA) I had 1563 beakers per turn.

Foremost of note from a playtesting perspective, my first GA was 22 turns, my second one 33, and I extended that one for the rest of the game with GG’s, GA’s and a GE at the end– an additional 91 turns. GS’s still went to bulbing, which gets us into Harder Free Tech. I felt the mod didn’t limit me at all. One time I waited a turn for a tech to tick off Hard. Otherwise I could always find something useful with which to bulb. On the other hand, I didn’t have a single opportunity to knock off a long tech while researching a short one. So it seems as if it’s very well balanced in terms of intended impact and playability. (I’ll follow up in my next game.)

Emigration gave me a nasty jolt when, some time in the Classical period, I dipped under zero by one or two and started losing a citizen for three turns in a row. I leveled off after that, and had no other problems, going under once late game when the loss meant nothing. In the moment, though, I felt that the loss was a bit harsh. My people were barely unhappy, and the impact on me was small but still meaningful. I can’t advocate liberalizing the rules yet, because frankly it caught me off-guard – next game I would be aware (hyper-aware!) and try to avoid it happening. I don’t think lowering the threshold to –5 would help, by the way, because then –5 would become the new zero.

I thought I’d never have a chance to try liberation, then had an opportunity to take Lyons from the Iroquois. That’s when I remembered that it has to be the capital, not a city. Maybe next game.

In terms of the Balance mods’ effect on game play, once again the AI sizes remained balanced. I still like this, but look forward to seeing the exception to the rule with a runaway AI. I would guess this would have to occur early on, before the building of walls. I can’t imagine the AI breaching them until the arrival of artillery. All in all, the AI seem to fight more intelligently.

The AI’s flow continues to be similar, with a burst of aggression in the nineteenth century aimed at other AI (my prior game) or CS (this game). Humorously, the Aztecs – alone on their own island – had a game-long war against Geneva which they won two turns before I launched. That was an example of very poor AI tactics. On the other hand, the Aztecs did launch a seaborne invasion against me (!) that involved over half a dozen units. Since I operated the second half of the game with a minimal strike force, this was scary until I dipped into the piggy bank for some mech infantry.

Curiously, Geneva kept losing land to the Aztecs. I didn’t know that was possible. By the end it was down to a spit of land and its water hexes.

And on a final side note, having a permanent GA made the Persian cannon junior artillery. I never bothered to research dynamite.

Now – how about adding the Capitol Railroad Boost to the ensemble? It seems to make sense.