CE vs SE Head-to-Head Experiment!

[DaveMcW]

I am bashing SE.

The numbers that WERE posted in this thread (and there weren't many of them) showed CE blowing SE away in every category.

You can talk all you like, but the data supports CE as the clear winner.

 

[mutax2003]

SE favours early domination or conquest victory, while CE favours space race and late domination victory, what do you guys think of this statement?

 

[frob2900]

I'd pretty much agree with that. Also I'd say that since SE favors generally higher production, if you can keep up in tech somehow a relatively space race victory is very doable in a proper SE economy. I've managed to launch in 1702 AD with an SE using Peter. Pumped out all space parts easily, the heavy parts (dock/engine) took less than 10 turns in my capital (massive settled engineers + ironworks).

Has anyone tried the late game SE using Frederick/Germany with massive engineers? Put the national epic + ironworks in a heavy food site with rivers + a few plains hills. Then pump all engineers into this town and finally round out with a few watermills so the city is static at 20 + (amount of engineers runnable with ironworks/assembly plant/forge). Crazy production.

I've never really run a cottage heavy game WITHOUT being terrified of the enemy civs. With SE I can out produce them easily on Monarch.

BTW off topic: Whats the earliest space launch you guys have managed on prince in 2.08 patch? I've never managed pre 1700 but it should be doable..

 

[Gunner Jones]

Would it be possible for the SE proponents to take a crack at Wodan's save? It's quite possible Wodan's SE effort could be improved. I'm having a go myself, but I'm still grapping with the basics.

Selfishly I'd like to keep the analysis going - watching the head to head experiments and the demo games is the only way I get to accelerate my undertanding of the internals of the game.

Gunner.

 

[futurehermit]

I am bashing SE.

The numbers that WERE posted in this thread (and there weren't many of them) showed CE blowing SE away in every category.

You can talk all you like, but the data supports CE as the clear winner.

It's clear you don't know what you're talking about in the sense that you don't know how to run a SE. I did EXTENSIVE testing of my own and have played umpteen games with it and prior to liberalism SE >>> CE.

LATE GAME YES the CE is superior, but in my games, I've already won by that point

ps., your data are bogus if the player running the SE doesn't know what he's doing...

 

[futurehermit]

For those of you who want some real proof of the power of the SE check out this save game...

I'm playing as Rameses II Monarch, Continents, Normal, Improved AI mod.

I had stone, but didn't have horses (had copper and iron).

I built the pyramids and ran a SE using police state at times to crank military.

I achieved Military Tradition in 1100AD and am now cranking out cavalry with police state, theocracy (state religion confucianism), and vassalage.

I also completely wiped out the USA in an early war.

I have a clear-cut points lead already and will soon have a sizeable empire-size and tech lead, thus being able to dictate my victory condition (Hannibal is friendly and not backwards, leaving even diplo open to me).

For those of you who "can't finish a SE game", why not see if you can finish this game, I think I've left it in a "winnable" position...

EDIT: P.S., Note two things about this game: 1) Despite running many scientists in my capital, I lucked out and got another GE. I am saving him for the SoL. 2) I believe I could've got the time under 1000AD. I printed off a copy of the GS tech preferences and they must've changed it or the copy was wrong because a number of techs came up prior to liberalism that I had to research manually (the other civs couldn't/wouldn't trade them) and that set me back. The techs included: drama (???), theology (???), compass, etc. I purposely ignored sailing forever in this game thinking to avoid having to research compass, calendar, etc., but for some reason the GS prefers sailing before liberalism??? Has this always been the case or has the tech preference been changed recently?

 

[idiot_savant]

I am bashing SE.

The numbers that WERE posted in this thread (and there weren't many of them) showed CE blowing SE away in every category.

You can talk all you like, but the data supports CE as the clear winner.

While I'm not going to say the results were bogus, I've done some stats and graphs (to be shared once I get them analyzed) and there was no such thing as a clear win for the CE. In terms of total productivity, the CE and SE were neck-and-neck until about 1650s (on Mr. Cynical's saves anyway). It took about 25 turns for Liberalism to kick in and the cottages fully mature was my view after reading the logs. Moreover, I think Cynical played the games very closely and missed doing some of the things that the SE strategy NEEDS to work properly:

1) Faster -- not better -- GPP production: GPPs were about the same until late. As people have shown, particularly Acidsatyr, you should be maximizing these earlier for an SE to fully work. My impression is that Cynical is right and wrong: a good GP farm can come close to overall GP output as the curve flattens, but wrong in terms of how fast you can get them going. There was no clear distinction between when GPs were produced in either case.
2) Lightbulbing: Cynical did not use GPs in this manner. That's fine. But an SE demands you grab an early advantage and then exploit it quickly. You need to lightbulb. The SE strategy proposed ramps up research with the express intent of turning it off before the CE really gets going. I would call this something like "burst beakers" (a la burst damage in DPS, etc.).
3) Whipping: Cynical whipped more in the CE than the SE. Hence, there really is no data against the point being raised. While the statistical analysis was clear that whipping impacted pop growth, there was no indication of negative impact.

That is not criticize Cynical. He was very painstaking about documenting all of his activities. We should all be thanking him for actually doing it (I am hopefully here). But I don't think the SE was played the way it is being recommended here.

I think the conclusion is ultimately inconclusive based on the way the two games were played. However, they do indicate that the SE could conceivably do better in terms of productivity if these elements were added more into the mix in the early game.

From an analytical POV, these are my takeaways:
a) until someone plays and documentsan SE the way Acid or Hermit are recommending, the data will never show proper results;
b) the real winner here for a full game is the SE early/CE late variant that no one has played yet. I'm going to get there eventually, but would be great if someone actually went that road;
c) other games have shown that there are other ways to manage your economy (trade routes for example) -- I think we should be expanding our horizons past these two alternatives a bit.

My $0.02.

 

[idiot_savant]

I'm not disagreeing with you, future, but having Pyramids skews the results just like Financial does so for the CE. Based on this thread, "everyone" agrees that:

Financial must be CE for best results. Pyramids (and Philosophical) must be SE for best results.

Comparison of Cynicals results show that an SE (even not one played as you would advocate) can keep up with CE in the early game as you've stated. The only real question here is what do you do when none of these preconditions exists (Fin, Philo or Pyr).

On Lightbulbing: Once you secure Fishing, a GS will prioritize water techs to a certain (historically inaccurate if you ask me) degree. If you ignore fishing and have Philosophy, GSs will start to pop techs along the bottom track as well. (See the beeline for GP or Kind word and a gun challenge.)

 

[futurehermit]

Two comments on that:

1) SE > CE prior to liberalism/democracy. Therefore, if you are going for early conquest/domination win, SE = better (see my posted savegame).

2) CE > SE post liberalism/democracy. Therefore, if you are going lategame science victory--and started with a SE--it is good to transition to a hybrid economy running cottage civics lategame (by hybrid I mean primarily working cottages, but getting some free specialists from SoL, mercantilism and running specialists off of surplus food sources). The exception may be the Russians, especially Peter whose UB makes a spacerace SE more attractive (mostly skip industrial going straight for computers then backfill where necessary).

Whether transitioning from SE to CE is better than running CE from the beginning is very hard to tell imo. SE is clearly stronger early game, meaning you'll unlock the CE civics much sooner thus you can reap the early-game benefits then start the transition using CE civics earlier thus lessening the catch-up (to full towns) period to a full-blown CE. Plus you have to weigh the benefits of the early game SE vs. a late game CE which is like apples and oranges imo.

 

[futurehermit]

I'm not disagreeing with you, future, but having Pyramids skews the results just like Financial does so for the CE. Based on this thread, "everyone" agrees that:

Financial must be CE for best results. Pyramids (and Philosophical) must be SE for best results.

Comparison of Cynicals results show that an SE (even not one played as you would advocate) can keep up with CE in the early game as you've stated. The only real question here is what do you do when none of these preconditions exists (Fin, Philo or Pyr).

On Lightbulbing: Once you secure Fishing, a GS will prioritize water techs to a certain (historically inaccurate if you ask me) degree. If you ignore fishing and have Philosophy, GSs will start to pop techs along the bottom track as well. (See the beeline for GP or Kind word and a gun challenge.)

When Fin/Philo/Pyr are not in question, I would assume that CE will be superior for reasons I've already stated. However, with spiritual (as I'm finding out with Rameses) a SE may still have power in the sense that you can have periods of running caste system + pacificism (going back to slavery/OR over time without anarchy).

Philosophical and financial seem to be on the endangered species list of traits (few leaders have them it seems) and pyramids is not guaranteed without stone and/or industrious, so the question is valid. I just feel that based on extensive testing I did earlier that CE will be better than SE without philo and/or pyramids.

Why? (I'll repeat my argument here): 1 scientist requires 2 pop (1 surplus food and 1 specialist) and results in 3 beakers/turn. 2 cottages at 100% sci will produce 4 beakers/turn as soon as they grow to size 2 (fast). Without philosophical, the surplus GP generation (beyond a CE's GPfarm) is less significant meaning the "burst beakers" provided by lightbulbing is significantly less. Therefore, I side with CE in this situation.

 

[frob2900]

IMHO a point that's being overlooked is how much war there will be in a game. For periods of constant warfare (e.g. more powerful Montezuma/Alexander as neighbor) where it is impossible to immediately cripple the AI and also unthinkable to accept their demands (e.g. demanding a city), the CE will tank, because all those nice cottages are producing culture..

Thus another game that should be tested is CE - Heavy Warfare vs. SE - Heavy Warfare. Just add a bunch of psychos as the AI with Aggressive AI settings...

In of my games where I played on Monarch with Elizabeth engaged in a series of wars against an unrelenting Catherine my cities were constantly in revolt, forcing me to run 100% culture. Even Hereditary Rule didnt help since most troops were needed at the front.

In a pure peace game I'd pretty much have to agree that the SE is about 50% as good as the CE at science.

 

[idiot_savant]

However, with spiritual (as I'm finding out with Rameses) a SE may still have power in the sense that you can have periods of running caste system + pacificism (going back to slavery/OR over time without anarchy).

Good point here that hasn't really been explored too much yet.

Without philosophical, the surplus GP generation (beyond a CE's GPfarm) is less significant meaning the "burst beakers" provided by lightbulbing is significantly less.

Not sure I agree here. By 600, in a beeline to Gunpowder test as Louis (same 3 wonders in all tests), I had produced 8 GPs (1 GP, 7 GS). I recreated this with Ottomans as well. With a GP farm in capital and cottages elsewhere, I max'd out at 4-5.

In these tests, Cynical generated 4 over same period while Wodan had 2 if I read their spreadsheets correctly. When the reqs are low, you can speed it up and have more control by running multiple small GP farms.

Whether transitioning from SE to CE is better than running CE from the beginning is very hard to tell imo. SE is clearly stronger early game, meaning you'll unlock the CE civics much sooner thus you can reap the early-game benefits then start the transition using CE civics earlier thus lessening the catch-up (to full towns) period to a full-blown CE. Plus you have to weigh the benefits of the early game SE vs. a late game CE which is like apples and oranges imo.

To me, this is all a matter of turns and timing. I'm not sure it's as apples/oranges as it sounds. If you're SE has a limited amount of CE-based towns which was part of the initial setup here, it makes the transition slightly easier.

The key needs would be: major number of workers, unlocking techs at appropriate time (probably 1000-1100), and impact of buildings on the CE and how fast you can get these up. Timing a GM trade mission or two at this time would be huge as well.

Sorry, boring slow day at work.

 

[futurehermit]

^^^Ok, nice to know that SE still generates nice #s of early GP over a dedicated GP farm. I am surprised, but since I haven't tested it myself I will take what you say for cash unless someone wants to contradict. I still feel the CE will be > SE without pyramids or philosophical, but if someone can prove this wrong then I will never try a CE ever again

IMHO a point that's being overlooked is how much war there will be in a game. For periods of constant warfare (e.g. more powerful Montezuma/Alexander as neighbor) where it is impossible to immediately cripple the AI and also unthinkable to accept their demands (e.g. demanding a city), the CE will tank, because all those nice cottages are producing culture..

Thus another game that should be tested is CE - Heavy Warfare vs. SE - Heavy Warfare. Just add a bunch of psychos as the AI with Aggressive AI settings...

In of my games where I played on Monarch with Elizabeth engaged in a series of wars against an unrelenting Catherine my cities were constantly in revolt, forcing me to run 100% culture. Even Hereditary Rule didnt help since most troops were needed at the front.

In a pure peace game I'd pretty much have to agree that the SE is about 50% as good as the CE at science.

Agree: SE > CE for heavy warfare game (which most of mine are, and I'm the initiator usually )

Disagree: CE > SE at science in pure peace game. SE is better prior to liberalism/democracy (proven many times over on these boards and in my own games, and those of acidsatyr, etc.).



EDIT: PS, I would like to see some head-to-head games between equally skilled opponents of the CE vs. the SE (CE played by CE advocates and SE played by SE advocates). Using Liz (Phi/Fin). SE only has 1 cottage city (capital). CE only has one specialist city (GPfarm). Goal 1: First to liberalism, Goal 2: Must eliminate 1 neighbour, Goal 3: Fastest victory year, any victory condition. Both starting from same save. Condition: Must have stone in capital or 2nd city to allow for pyramids to be built (If 2nd city, capital must have decent production and/or enough forests for chopping--yes, pyramids are situational and I am not claiming otherwise). We could also have a 3rd person play SE without pyramids to see how that does.

 

[frob2900]

Disagree: CE > SE at science in pure peace game. SE is better prior to liberalism/democracy (proven many times over on these boards and in my own games, and those of acidsatyr, etc.).

For the first years 4000BC - 1000BC you wont run many scientists, hence here cottages are the only non-gold/gem/silver research boost method.

1000BC-0AD you can run both, fair enough + its probably a good idea to pop a great scientist here.

BUT ca 500BC to 500 AD Civil Service hits. How can you say that non-representation scientists are better than a well developed cottage in the capital together with the 50% bonus?

It's crazy not to cottage the capital unless your a mega-warmonger who really wants those hammers from bureaucracy hence needing a really big capital.

For general robustness the SE is better IMHO but for raw science i'll generally go for cottages. HOWEVER sometimes the map wont support this. Some maps/situations do prefer research through specialists.

I think its impossible to say that either system is better. It depends on the situation.

 

[futurehermit]

Note: I am NOT saying that CE is better than non-representation scientists. Please read more carefully. My version of the SE assumes pyramids and representation. Plus, I WILL cottage my capital and run CS when it hits. Prior to 1000BC I don't care about research (I run 70% without prioritizing cottages, even in capital, although I'll build them in capital). Why? Because I care about maximizing HAMMERS in this early critical period to get my early empire up and running, including pyramids, libraries, barracks with a total of 3 early cities. I only care about hitting alphabet and then I'll drop to 0% science (without worrying even about scientists for a period) and crank military to the point where I'm almost going on strike. Then I'll go wipe out a neighbour with my huge army while switching back to representation from police state and start to run scientists getting ready to settle into a peaceful run to liberalism (and, preferably an early military tradition for my 2nd and final major period of war).

 

[frob2900]

Yeah, I'd agree that for a warmonger the hammer economy (the third type is the most important, plus for me this always leads to a mixed shield/cottage economy since i'll end up with a few nice AI cities surrounded by villages or towns. But for a game where i'll never fight an offensive war I'd never run a specialist economy except, perhaps with peter/russians and pyramids and a heavy wonder building strategy (pyramids+oracle+angkor wat) using my cheap workers to chop.

The few times I've pulled this off Mansa Musa or someone similar has been on the map allowing me to trade heavily and only research select techs so that even though my pure specialist economy is research weak by the end game i'm still drowning in hammers so I can pull of a space race victory even though the AI is as advanced as me. In the best case there's always a backward monty/alex around to bribe with all my techs if the advanced AI's are completing the spaceship ahead of me.

In such inevitably heavy tech trading games I've managed to get apollo program ca 1500-1600 AD on prince, on monarch its either ca 1750 AD or noticing that the AI is really getting ahead of me leading to a massive rifleman/cannon or infantry/artillery war which can be touch and go).

A really early space race might be doable with cottages, but I've not really tried.. Does anyone have a 1600 AD launch on prince/monarch with 2.08?

 

[Wodan]

Would it be possible for the SE proponents to take a crack at Wodan's save? It's quite possible Wodan's SE effort could be improved.

I'm positive it could. I'm too much a dabbler and theoretician to min/max properly.

Wodan

BTW good thoughts from everyone. I hope some people take up some of the challenges / hypotheticals thrown out here.

 

[futurehermit]

I would take a crack at the save, but I'd rather we did the head-to-head elizabeth game. That being said, however, I still think the original point of the thread (to compare non-phil, non-financial situation) is a good one. It's just that I feel, a priori, that the CE is better in a neutral situation. Maybe I'll try a non-pyramids, non-philo game with SE to see how well I can do. I just worry my base research rate will be too poor!

 

[DaveMcW]

I'd be willing to play a head-to-head game with stone and Frederick or Peter. I still think CE would dominate post-liberalism, but maybe it would be somewhat close.

 

[futurehermit]

I'd take you on with Peter. I'll even play for space race instead of domination (easier imo). I've wanted to try the Peter "skip the industrial era" strat for awhile now, so that would be fun to try

Do you agree with these three goals:

1) First to liberalism

2) Must eliminate at least one neighbour.

3) Fastest victory (space race)


???

P.S., The gauntlet has been laid down and now picked up

EDIT: P.P.S., What skill level/settings do you normally play under. If we're not in the same skill range then our challenge would show nothing imo. I normally play comfortably (80+% wins): Monarch/Normal/Continents/Improved AI mod, rest default. Much less comfortably, but with some wins (~25%? wins--I play this less often, so hard to judge for sure my win rate), Emperor/same stuff.