K-Mod: Far Beyond the Sword

[Lenowill]

(FYI, I wrote this before TMIT and Karadoc posted, and this post doesn't take anything they said into account because at the time I wasn't aware of their posts. I will comment on theirs in a later post, probably.)

Regarding Representation versus Universal Suffrage, I have to admit that it's a sticky issue for me as well. I am leaning in the direction of thinking the nerf to 2 Beakers was not the way to fix the situation (although it was a good and reasonable effort). I'll try to explain why.

First, about me - I used to obstinately run a Pure Specialist economy all the time in unmodded BTS, even on Emperor for a while (usually as either Sitting Bull or Elizabeth) ... so I'll be weighing in on it as someone who has been there and done it ad nauseum, both with and without the Pyramids.

Firstly, I disagree with the statement that a Pure Specialist economy was nonviable as an empire design before the nerf. In regular BTS, the thing about a Pure Specialist economy (on a Phi leader, of course) was that it had a very strong early game (which could be setup with minimal issues on a variety of map starts) coupled with an oddly flexible midgame that emphasized bulbing toward Education and Liberalism and then blazing away on the jet-fueled engine of a massive tech lead while your rivals fumbled around trying to catch up.

The catch though was that a Pure Specialist economy did run into a problem once rival civilizations reached the Industrial and Modern eras. All the Emancipation floating around would start to make your Caste System frown, which would force you to use your Culture Slider more, which would make your cities not give as much research from their Trade Routes ... and all those beefy 7+ Commerce cottages now owned by your rivals around the world (and all that tech trading and tech difficulty reduction stuff that would let them nip at your heels even while you're ahead) would start cutting a great big wedge out of your research-advantage pie.

I guess the biggest trap with the unmodded BTS Pure Specialist economy is in thinking that you should keep growing your cities bigger after you get Biology ... under most circumstances, you shouldn't - getting that much bigger is a nightmare for your healthiness and happiness. What you should do instead is use the extra food as an excuse to start turning some of your farms into Watermills and Workshops in preparation for a switch to State Property and a Hammer/Specialist hybrid economy. Eventually you end up pretty much turning into a Super Hammer Economy with specialists still active in only a few cities (which is also handy for cranking out the inevitable spaceship parts, or Modern Armors and Mobile Artillery, once you get to the point of using your advantage to settle down and win the game). You can stay on Rep through the end of the game in that case (or hop to UniSuff to have the rushbuy option available if you've been building Wealth a lot and would like the added flexibility).

That was what you did to "win" with a Pure Specialist economy. You used it to get a huge tech lead and then evolved your economy into something else. Using your fancy gadgets, you revolutionized your society from a farm-and-philosophers nation into a powerful hammer-heaven that still had a strong core of settled research to back it up.

...And of course, along the way to getting to that "transition" point, you snagged the free tech from Liberalism and probably ate up most of the free great people from Economics/Physics/Communism/Fascism/etc. by beating the other nations to them.

So no, I don't think a Pure Speicalist economy was at all unviable before the nerf - if you used it appropriately and didn't try to make it do something it wasn't good at doing. I do think it is unviable now, after the nerf, because its lategame disadvantages are much harder to make up for in a timely fashion compared to before (because it takes you longer to zoom ahead in the midgame and the rival civs have more time to catch up). K-Mod pretty much forces you to use a Cottage/Specialist Hybrid economy if you would have run a Pure Specialist economy before, which is fine with me in some ways and irritating in others (because if I end up on a map start where cottaging heavily isn't an option, my Phi leaders are inevitably in a worse situation for the midgame than they were pre-nerf).

In my opinion, honestly, it would probably have been better to change the effect of the Pyramids to something different (and by "different" I mean "not highly prone to breaking the game at low difficulty levels and teaching new players the bad habit that wonderspamming is a good idea") than it would have been to nerf Representation's Research benefit. (I might experiment with that in my own private mod-mod of K-Mod and see what happens.)

Hammer/Cottage economies (UniSuff, FreeSpeech, State Property) have always had a ridiculously strong lategame - you seriously can't go wrong with them, if you're lacking in anything you just Wealth or Research process it - even an unexpected defensive war usually gets shrugged off because your production and research are both so high that you can respond to pretty much anything - so it was nice that there used to be a viable alternative that didn't involve Cottage Spam. Now in K-Mod you can choose to Cottage Spam with some specialists on the side, or you can just Cottage Spam hardcore until you get to the inevitable point where Production becomes king.

I dunno, I can't shake feeling like there must've been a better way to encourage hybridization and discourage just sitting on Rep forever. I'm not even going to say that the way it was done was particularly bad - I also thought Rep needed a nerf of some sort for the reasons you gave - but actually playing with it since the nerfs, and noticing how much less I use specialists in general even on Elizabeth, has made me wonder about that.

 

[Lenowill]

@Karadoc

Sorry for the double post, but I wanted to go ahead and reply.

It sounds like you weren't leveraging the bonus hammers from Universal Suffrage right. Either that or you just plain didn't have as many Cottage (Town) tiles as you should. All those little +1 hammers add up, and if you don't build too many buildings in your cities and do use those hammers to Build Wealth or Build Research (the processes), you can use those to help you create a research rate that already wasn't far behind what Representation could accomplish.

It's also possible that you weren't using Foreign/Overseas Trade Routes effectively enough, seeing as those get downright awesome in terms of how effective they are at pumping your research later on.

It could also be something else. I dunno. TMIT might have other insights.

 

[karadoc]

It would be a shame to kill of a strategy like you just described... ... and to be honest, I'm not really too worried about the pyramids; I'm more interested in just taking away the dominance of rep in the late-game. I want it to be powerful in the late game, but not as powerful as it was in the first place.

You might remember that originally I kept the +3 beakers but set it to 'high' upkeep. I don't think that had much of an effect. The thing about upkeep is that it scales linearly; and so if the civic is good for a small civ, then it's still going to be good for a large civ (with the exception of civics with localized benefits, such as bureaucracy; and, I suppose, the happiness from Rep) - so upkeep isn't really a powerful tool for balancing civics. I'm not sure what else to do. I could try something like +3 beakers but with -30% from trade-routes, or something like that.. but I don't want to it be a complicated / clumsy / cumbersome combination of effects. It's best to keep it simple...

(note: -30% from trade routes wouldn't actually mean that you get 30% less than normal, because the trade modifiers are added together before they are applied. -30% would actually be a relatively modest penalty.)

 

[Lenowill]

You could always give it -2 happiness in your [map-size-influenced number] smallest cities, to cutely counterbalance the +2 happiness in your [map-size-influenced number] of largest cities. "We sigh at this under-represented city of ours!"

A much simpler way to code it would be to have it give -1 happiness in all cities but +3 or +4 happiness in the [map-size-influenced number] largest cities. The potential for small places to not have enough say in a pure Representative system is very real even in the real world, and it wouldn't be horrible from a gameplay perspective. It would also help solidy Representation as a nice Civic for close-knit medium-sized empires that have grown a lot "vertically" (in terms of population) rather than REXing to high heaven, while letting UniSuff remain a more attractive option for the giant-world-power cottage spam Civs.

I dunno, something to think about and/or test.

If you wanted to get really silly/realistic, you could go for making it add an additional happiness penalty to overseas colonies ... "No taxation without representation!" Yar!

But yeah. Something with happiness seems to fit the flavor the best, and would encourage more careful thought on the player's part in deciding which cities should become "the big six" or whatever.

 

[Charles555nc]

Maybe Im not nearly as good as I think I am.

But playing on Emperor the last 30 full games, Ive never been able to get the pyramids and never even thought about switching to representation. Just like I would never switch to pacifism (maybe this latest version I will try it). Hereditary rule is almost the only government civic you can run for most of the game, at that difficulty, it seems (and thats kinda lame I guess).

Suggestion 1:
If you think representation is too good if you get the pyramids, how about allowing a beaker bonus once you reach liberalism or something (like once you get liberalism, representation now offers +3 beakers per specialist, up from 2 beakers)

Suggestion 2:
Maybe you should boost the happiness factor of representation or create a new one to make it more balanced between civics.

Suggestion 3: Maybe the upkeep cost of police state is too high. Universal sufferage seems to have a better balance, to me, than Police State. Maybe you could add another "unlocking" feature to make the pyramids not an "insta win" as you call it. Like once you get facism, police state upkeep is reduced.

Side notes:

And just addressing the upgrade cost factor: I think it would add A HUGE amount to replayability to let the player upgrade units easier.

Right now the main strat for me, is to expand like crazy, start rushing a huge economy, and WAITING forever for guilds/civil service/machinery/engineering, while avoiding war with neighbors. The limiting happiness, healthiness, and production values can make the early/mid game terriblely boring. A large part of the early/mid game I dont even want to expand because the corruption/city upkeep costs would kill me.

So you are highly discouraged, as the player, towards almost ANY early game aggression.

(I think its unfun if an ai just barely researches a new military tech, that almost all of their units are easy mode upgraded to the latest tech level over the next few turns)

Suggestions:
I think a 20-25% cost reduction for the player in upgrading units BEFORE rifling would add alot of excitement and potential to the early mid/game. And possiblely a reduced discount for upgrading, for the AI, at all ages.

 

[TheMeInTeam]

Late game sources depend very much on a wide variety of factors.

Suffice it to say, I could easily eclipse 2000 beakers/turn in the current let's play using something lulzy like police state by the end of it with that city volume...probably more.

One of the very best games I've read about, Unconquered_Sun's BOTM 10, relied heavily on specialists and very minimally on representation, ironically (he hammered it all with workshops). In this kind of model, a very unconventional (at the time anyway) setup of workshops with holy rome powered its way to 1400 AD tanks and bombers ready to invade on deity. Rep was an afterthought for the most part; he had tanks available before most players w/o mids can switch to it!

Let's put it this way. Say you have ~10 cities, with following trade route sources:

- Default
- Currency
- Corporation
- Free Market

and average 10 /route. If you have 75% multipliers in those cities, you get 750 only from trade routes if you find a way to deficit research or build wealth + tech...except that in ONE city (bur cap), there's a good possibility you have an academy and oxford, pushing you to 225%...of 60 (thanks to bureaucracy). Thus, if we had absolutely nothing else but trade routes, basic infra, and build wealth, you'd be pulling 870 /turn.

How many default rep specs would you need to punch a dent in that? A lot. You could run 4 pre-nerf rep specs in every one of those cities and the difference between having rep and not having it would be ~200 ONLY if you'd still run those specs while in rep. However, you also have some towns (esp in bur city), lots of passive commerce (any coast, riverside, plantation resources, luxuries etc). In practice, a lot of empire setups would lose 100 beakers or less.

Note that above trade assumption is for large cities (bio farm cities would certainly qualify, pushing into the 20 pop range often and actually getting MORE than that from routes, especially when coastal! High level AI pull such cities too for the foreign trade goodness).

An alternative city setup is shop spam. Throw down 15 state property shops, get your 60 hammers in caste, and then use forge/factory/power to double that to 120/turn. I don't think you need a rocket science degree to see what happens when these cities simply choose to build wealth . 1200 of or per turn, in a city that can VERY quickly set that up. And that's WITHOUT the trade routes or passive commerce, too!

There's no getting away from some marriage to the slider, but you can manipulate it. Trade routes are too strong to ignore it.

Of course, we haven't touched on things like single currency and open markets (another 2 trade routes, meaning even without FM you have 5), or the RIDICULOUS post-beaker multiplier that is tech trades.

When I look at a bio farm feeding a specialist or a SP + caste citizen on a workshop, I'm seeing an utterly miserable case for bothering with a rep specialist. Part of the appeal of the raw is that it's higher! are much more flexible and convert tech into units or projects faster.

Also keep in mind that late game, other civics work at cross purposes with rep. Emancipation pulls the productivity of farms = whips, or at least digs you into more than you can get from rep. If you go rep + emancipation, you're soft on production unless you go SP + shops or work a lot of mines...but mines are comparable to PRE-NERF rep specs in value and late game they have the multipliers too! So specs are a spill-over option late-game unless GPP farming. Unless you specifically dedicate to them their direct research contribution is actually not that staggering...and if you DO commit to them you ARE making concessions elsewhere that balanced them late game.

And a few runs by peachrocks/myself have made me suspect that late-game teching is basically the same. You might have shaved 100-200 beakers/turn, but for 1500 to 3000 beaker/turn empires with MULTIPLICATIVE TECH TRADES that's a drop in the bucket. Games will last maybe 5-10 turns longer for space if and only if there's a tech bottleneck.

The problem is the shattering of the early and mid game value of pyramids and decision of what civics/wonders to pursue is a big price for those 5-10 turns (or a tech or two, or whatever).

 

[Highwayhoss]

BTW Karadoc, I mentioned traits earlier, have you seen Munch's Alternate Traits v4? I've used this modcomp in most of my custom mods and it makes games more competitive; I also suggest reading the thread through as it discusses the evolution of these traits.

 

[karadoc]

I just ran a some test games to try to get a better sense of how much the game is extended by rep nerf. The test games worked like this: I generated a map (Not Too Big or Small, default settings for everything, Monarch difficulty); I then used the world-builder to give archery and 2 archers to the would-be human player; then I saved the starting position and put it on auto-play until the game was over. I then loaded the same starting position with the other version of representation and ran it on automatic again.

I did 7 such pairs like this. Here are the results:
(sorry the games are not side-by-side. It would be better to put them in a 2-row table, but I don't know how to make that on these forums.

+2 Representation:

1. 460 turns, space victory, Suryavarman
2. 352 turns, cultural victory, Louis
3. 409 turns, domination victory, Huayna Capac
4. 323 turns, cultural victory, Huayna Capac
5. 351 turns, domination victory, Suryavarman
6. 364 turns, space victory, Asoka
7. 410 turns, cultural victory, Mansa Musa

+3 Representation:

1. 393 turns, diplomatic victory, Suryavarman
2. 407 turns, diplomatic victory, Suleiman
3. 341 turns, domination victory, Huayna Capac
4. 307 turns, conquest victory, Huayna Capac
5. 326 turns, domination victory, Suryavarman
6. 369 turns, cultural victory, Asoka
7. 378 turns, diplomatic victory, Kublai Khan

(Those 'diplomatic' victories were essentially UN assisted domination. The winners were not what I would call 'popular'.)

Ok.. So it isn't a huge amount of data; and of the times I stopped to look at how things were going, only very few of the AI players were using Representation anyway. Universal suffrage was the more popular civic. But still, I think it shows that there is a non-trivial effect on the length of the game.

The only test case where the +2 game was significantly faster was when Suleiman managed to prevent Louis from getting a cultural victory in the +3 version but failed in the +2 version. It's hard to say how long Suleiman would have taken to win the +3 version if Louis lost his culture somehow. It may look like a similar thing may have happened in the 7th game, but it is not the case. Kublai Khan did not fight Mansa in that game. Mansa Musa was running 100% culture at the end of both version of that game, but he simply was not given enough time in the +3 version.

As I mentioned, most of the time the AI players weren't choosing representation in either game. I personally would expect the turn difference to be larger when human players are involved - because human players actually know how to use representation properly; but I can't really say that with much certainty, and even for AI-only games, 7 pairs isn't really a lot of data.

Anyway, the bottom line is that although I am tempted to return Representation to its former glory, I still think the +3 is too powerful. It's more powerful than the other late-game civics in that branch, and it does apparently have an effect on the length of the game.

And just addressing the upgrade cost factor: I think it would add A HUGE amount to replayability to let the player upgrade units easier.

Right now the main strat for me, is to expand like crazy, start rushing a huge economy, and WAITING forever for guilds/civil service/machinery/engineering, while avoiding war with neighbors. The limiting happiness, healthiness, and production values can make the early/mid game terriblely boring. A large part of the early/mid game I dont even want to expand because the corruption/city upkeep costs would kill me.

So you are highly discouraged, as the player, towards almost ANY early game aggression.

(I think its unfun if an ai just barely researches a new military tech, that almost all of their units are easy mode upgraded to the latest tech level over the next few turns)

Suggestions:
I think a 20-25% cost reduction for the player in upgrading units BEFORE rifling would add alot of excitement and potential to the early mid/game. And possiblely a reduced discount for upgrading, for the AI, at all ages.

I think there is a case to be made for reducing upgrade costs, but I don't think its the case that you are making here. Firstly, I'd assert that the game has no shortage of replay value. It's a very old game already, and we're all still playing it. Secondly, I disagree that the early game is boring and I disagree that the game-mechanics highly discourage early game aggression. In fact, many games are won and lost in early-game wars. Rush strategies are actually quite popular, and sometimes even vital. So... what can I say? I think you're wrong about early aggression being stifled. - As for the AI upgrading all of its units very fast; it's true that their upgrade costs are lower, but the main reason they upgrade so fast is that they consider it a high priority. It isn't uncommon for them to drop their research to 0% for several turns so that they can afford all the upgrades.

BTW Karadoc, I mentioned traits earlier, have you seen Munch's Alternate Traits v4? I've used this modcomp in most of my custom mods and it makes games more competitive; I also suggest reading the thread through as it discusses the evolution of these traits.

Those traits look pretty good, but I don't think I'll use them here. It's just too much deviation from the standard rules.

 

[Lenowill]

@Karadoc

It doesn't surprise me that a full AI game would work like that. A problem is that a human player can make the game much shorter already by just running a Hammer Economy (including quite a few cottages for commerce) from Communism tech onward, instead of bothering with Representation and specialists later on. Mathematically speaking, a properly used Hammer Economy will zoom far and away ahead of every other late-game economic type in overall usefulness under 80%+ of game circumstances. (The only thing that appears to rival it properly is a suuuuper Free Market or K-Mod Environmentalism corporation setup with Cereal/Sushi and Mining Inc, and that takes a billion times more micromanagement to setup than just throwing down a ton of Workshops and ordering cities to build Wealth and Research.)

AI players do not appear to use Hammer Economies very well (can they choose to build Wealth or Research?), so there's no surprise that the nerfed Rep will have a bigger effect there. The AI needed Rep for its auto-assigned specialists to help out later in the game (high-skill human players avoid having tons of auto-assigned specialists in the lategame if they can, because said specialists are inferior to working actual improved tiles except under very particular strategies or trait synergies. Under K-mod I usually tell my cities to emphasize food, production, and commerce to help discourage them from assigning specialists. Heck, most high-skill players won't work non-financial Coast if they can help it at all, and the Governor loves to make the citizens do that. xD)

Again, it's a tricky issue. I still think that +3 beakers with a happy penalty for cities that aren't the "big six" could work out surprisingly well. (A big thing that already kills Rep later on is the rise of Emancipation cutting into the happiness available for your Slavery or Caste shenanigans, forcing you to either sacrifice a significant amount of trade-route commerce to the Culture Slider or deal with your population-bloated cities having an increasingly painful happy cap, so making it already have a happy penalty built in would make it have a shorter shelf-life while still having a chance to shine at what it's meant to do in midgame, especially for Phi leaders.) But there may be a better solution out there. I'd be interested to hear TMIT's further thoughts on it.

 

[keldath]

hi karadoc,

ok ive gotten an oos,

but only when my friend got killed, and right after it oos.

we only played 40... turns - he got crushed by aggressive barbs,

also, after the oos appeared - i could not open the menu to exit the game - so i had to alt tab and close.

u need any log or something?

 

[En_Dotter]

Does one of u have Win XP and other one Win 7 or vista?

 

[Charles555nc]

Id like to make a distinction, that I think early game aggression is essential on a continents type match but much more difficult in a pangea type map- where if your military drops below a certain amount you basically become an open target by many civs,

Edit: Could that supposed "no tech trading" bonus that I theorized about the ai getting be written into the difficulty settings section rather than directly in the settings option? Its really werid playing with "no tech trade brokering", compared to no tech trading, Im almost always number 1 gnp (like for 350 turns out of 400) and even now the ai hates to trade tech to me on an even level...but some ais are 6/7 techs ahead of me...I definitely like the pacing of "no tech trading" more...

 

[keldath]

En_Dotter - yes! how did u know???


besides that, for now we started another, and we did almost 300 turns, so far my friend's alive , so no oos, yet...

 

[En_Dotter]

The one that has vista or win 7 should put compatibility setting for civ to XP and all should work fine. Its not K-Mod problem but seems to be CiV problem. I had the same issue and found out the solution on this thread . That should fix it. It did fix for me and 2 of my mates.

 

[karadoc]

Regarding changing rep from +3 happiness in largest cities, to +4 in largest, -1 everywhere... I was just about set that up for some testing, and I was shocked to discover that civ-wide happiness changes are not available in the civic xml. Shocked I say.. I've actually recommended that exact same give with one hand and take with the other happiness balance thing to other people.. and I always assumed it could actually be done...

The upshot is that the happiness thing can't be done with a simple xml change. It would need to have a bunch of new code for implementing the effect, and having the AI understand how it works. (and the AI will probably be a little bit confused by it anyway, because when the AI evaluates happiness from civics, it evaluates the changes one at a time - and so its estimates for the happiness value might be slightly wrong for civics which have multiple sources of happiness / unhappiness...) The unhappiness from civics would be simply shown as "AAAARRRGHHHH!" unless I add a bit of extra back-end code to support it separately. -- So, the question is, would it be worth it? Is this really the way forward? (To clarify, I'm talking about changing rep to be +3 for specialists, +4 in largest cities, +1 in all cities. The happiness in the largest cities is increased from 3 to 4 to cancel out the -1 that they will also have.)

-

Regarding the Windows 7 + XP OOS thing. It may well be a problem unrelated to K-Mod, I don't really know. If you are able to reproduce the OOS using save from just before it happens, then I'd like to see that save... otherwise, I'll just keep an eye out for things that might cause a problem like that. If compatibility mode on Win7 makes the problem go away, then that's good - but it would still be nicer if it wasn't necessary.

Edit: Could that supposed "no tech trading" bonus that I theorized about the ai getting be written into the difficulty settings section rather than directly in the settings option?

I don't think there is any such bonus. I've looked through the code for everything that mentioned no tech trading, and no tech brokering, and I've checked the code for research cost calculations, and research rate calculations. I have not seen anything that would give the AI an additional boost (or penalty) from those settings.

 

[En_Dotter]

I know this is a bit off topic (maybe not) but will ask it anyway. If there is anyone willing to play K-Mod in multiplayer im pretty much up for it. This could help track down potential OSS and also be fun.
If anyone is interested send me a pm and we can try to organize something. The more the merrier

 

[Lenowill]

I don't know for sure if it's "the way forward" or not. I'm kind of perplexed that there's no direct Civ-wide happiness option already existing in Civics. :O

Hrrm, I dunno, I'd be interested in hearing more input from others, including TMIT, who is about a league and a half better at the game than I am.

 

[keldath]

hey guys,

thank you for the interesting solution with the combability issue. ill tel lmy win7 friend to try it (im on xp).

karadoc, i think i can reproduce the oos, ill tell my friend to kill him self, and ill see what happens.


on another note,
ive noticed the ai - during a short war we had, didnt put too much units on the defenssive point, i think the ai could have put more defence once my friend declared was and sent troops forward.
also maybe something can teach the ai to attack pillaging units around the city.
thats just a few notes ive said ill write, still we are early in the game, so who knows for now.
these are just my first thoughts.
other then that ive noticed a rly small delay when me and my friend are starting our turnm the delay is when our turn starts it take a few milisecs to the orders we give to be applied, maybe its due to hardware slowness (my friend have a much faster cpu then me - im with a wooded pc...).
i guess its not kmod related right?

ty all for the fast answers, i love this mod and work and karadoc!

soon ill release a new version of kmod with a bunch of stuff from my mod.

 

[Lenowill]

That command halting at the start of the turn happens in all Civ4 multiplayer that I've ever been in. Not sure what causes it, but I'm guessing differences in CPU ability could be the culprit. It happens pretty bad when I network my Laptop (midgrade gaming laptop from like 2006) to my desktop computer (much stronger gaming desktop from 2009) for LAN play.

 

[keldath]

En_Dotter,

interesting offer, are u available at evenings?

Lenowill,
sounds reasonable, probably cpu problems, ill report on that when i get a new one

ty.