Specialist Economy - are you guys really buying this?

[uberfish]

Acid, you could play at emperor or immortal if you wanted and see how it compares to the people doing prince

 

[MrCynical]

Well the draft looks fine to me. All the points raised in this thread seem to have been covered by it.

 

[VoiceOfUnreason]

why prince again?

"Because!", he explained.

 

[VoiceOfUnreason]

Hmm... What I would really like to have included in the report issued every 20 turns is total number of each type of specialist, and total number of each type of cottage being worked, as a touchpoint on how well the target strategy is being worked.

Of course, what I would really like is to be able to drop the save file into a tool like the log utility, and retrieve all of the information automatically. I'm going to have to write the code for that at some point.

 

[Mutineer]

Looks good. I can comit to playing this.
As I said before, I think the CE should be allowed to assign a priest in 1 city (probably the oricale city) and a couple of scientists in another for the first 2 great people. The low GPP cost of these makes them so much more worth while than later ones, and I think it would be a significant handicap for the CE to not be allowed to do this

Perhaps the rule could be changed to one of;

The CE is only allowed to use specalists where the GPP are the aim rather than the beakers.
After the 1st 2 great people the CE should only use specalists in their GP city other than “free” specialists (e.g., Mercantilism), or when population surpasses city size limits.
The CE is only allowed 1 type of specalist in 1 city, ie. scientists in 1, preists in another, an enginner in another.

Of course if you think this is making it a hybid ecconomy we shall stick to your rule.

You missing the Main Point of SE.

In SE main amount of beakers come from GP up to 6th GP, not from specialists themself.

Stick to one GP farm.

 

[Wodan]

Looks good. I can comit to playing this.

Kewl. I'll add you to the list.

As I said before, I think the CE should be allowed to assign a priest in 1 city (probably the oricale city) and a couple of scientists in another for the first 2 great people.

Oh, yes. I don't have a problem with that. Have to figure out how to word it. Your suggestions help a great deal.

This is long winded, but gets the idea across.
-- CE may not manually assign specialists in the core empire, with the following exceptions: “free” specialists (e.g., Mercantilism), when population surpasses size 21 (because specialists are the only option), if using citizen automation (see below), in a “GP Farm” (any number allowed the whole game), or before the 2nd GP of the game is generated (any number of specialists may be assigned in any cities, with the goal to spawn an early GP of a specific type).
-- If CE is using citizen automation, specialists are permitted but no “forced” specialists are allowed; however, note that the citizen automation feature often assigns a specialist that may not be desirable to the CE player.

Hmm... What I would really like to have included in the report issued every 20 turns is total number of each type of specialist, and total number of each type of cottage being worked, as a touchpoint on how well the target strategy is being worked.

Of course, what I would really like is to be able to drop the save file into a tool like the log utility, and retrieve all of the information automatically. I'm going to have to write the code for that at some point.

Your screen name is aptly chosen.

I can see some value there but it's a TON of work. Not worth the return.

If someone really wants to see that data they can always open the save games.

why prince again?

I've been putting off answering this.

Prince instead of Noble or lower, because
-- it's not going to change the difficulty that much, not to players of the caliber who will be doing the test in the first place
-- it is closer to the higher levels in gameplay, thus it will be easier for people looking at those levels to accommodate for that delta

Prince instead of Monarch or higher, because
-- less risk of the AI throwing a wrench into the works
-- closer to the lower levels in gameplay, thus it will be easier for people looking at those levels to accommodate for that difference

Thanks for the note, MrCynical. My suggestion for timing is to open the new thread and allow games to begin on Friday, unless someone else notices a huge problem.

Wodan

 

[Wodan]

You missing the Main Point of SE.

In SE main amount of beakers come from GP up to 6th GP, not from specialists themself.

Stick to one GP farm.

The main amount of beakers CAN come from GP... depending on the strategy of the SE player, he might well want to settle great scientists into his super-science GP farm.

Some rough math. 3 GPP / turn / scientist. Say 8 cities at 2 scientists each = 6 GPP / turn / city. That means turn 800/6 before the 8th city produces a great scientist.

Meanwhile, 3 beakers per turn (non-Pyramids) * Library etc. Say average of 7 * 2 scientists per city = 14. In 100 turns that's 1400 beakers per city.

Seems to me that the scientists are giving the lightbulbs a run for their money.

Meanwhile, the settled great scientists are producing about 30 beakers each * 100 turns = 3000 beakers each.

That's really rough math and may well be wrong.

On the other hand, lightbulbing definitely gives speed. It's the speed, not the actual value of the beakers, that makes lightbulbing all worth while.

Wodan

 

[Mutineer]

Yes, that what I am saying.

Ligth bulbing give beakers, but you assume I mean only beakers that included direclty into ligth bulb.
I cound a beakers that we are getting from trades from an expecive technology we got very early, or access to bonus technologies, like Music or CS or Liberalism. It is dificult to put actial value, it depends on situation, but in general it tend to double/ triple amount of beakers obtained.

 

[Phrederick]

-- CE may not manually assign specialists in the core empire, with the following exceptions: “free” specialists (e.g., Mercantilism), when population surpasses size 21 (because specialists are the only option), if using citizen automation (see below), in a “GP Farm” (any number allowed the whole game), or before the 2nd GP of the game is generated (any number of specialists may be assigned in any cities, with the goal to spawn an early GP of a specific type).
-- If CE is using citizen automation, specialists are permitted but no “forced” specialists are allowed; however, note that the citizen automation feature often assigns a specialist that may not be desirable to the CE player.

I would say: “free” specialists (e.g., Mercantilism), when population surpasses size 21 (because specialists are the only option), if using citizen automation (see below), in a "GP Farm", or to get a desired type of GP for the first and second GPs.

Really, I think that any case of using specialists for GP points should be acceptable. I often run a priest in my stonehenge/oracle city for an earlier shrine, or throw in an engineer for a chance of a GE, even after the first two GPs. If I never got a Great Prophet, I might throw a priest into my GL city, just on the off chance of getting a prophet. In that case: “free” specialists (e.g., Mercantilism), when population surpasses size 21 (because specialists are the only option), if using citizen automation (see below), or when the purpose is to increase the rate of GP emergence or to influence the flavor of GP.

That assumes people will be reasonable about that sort of thing, but you both seem like fairly reasonable people.

 

[Wodan]

Ligth bulbing give beakers, but you assume I mean only beakers that included direclty into ligth bulb.
I cound a beakers that we are getting from trades from an expecive technology we got very early, or access to bonus technologies, like Music or CS or Liberalism. It is dificult to put actial value, it depends on situation, but in general it tend to double/ triple amount of beakers obtained.

I'm not sure that's entirely fair.

For one thing, a civ that gets accelerated research will have techs available to trade to the AIs, too. Yes, maybe not as much or as early, but it'll be there. You're right, it's difficult to put an actual value. But, if we say it's a benefit for the one we have to say it's at least some benefit to the other too.

Secondly, I think it's important to stress the relative gain of your tech vs the AI's tech. If you get a tech by lightbulbing, and then trade it to a couple of AIs, then yes you gain some techs, but the AIs gain a tech too. You're absolutey correct that with skillful trading you can trade to multiple AIs for multiple techs. So, you're getting a lot of bang for the buck. Nevertheless, we need to point out that as part of the process the AIs get a tech for free. In fact, it costs them less than it cost you... all they did was give you a tech that all the other AIs already had. You had to give them a tech that NOBODY else had.

Anyway you have good points.

I would say....

Good suggestions. It was a little too wide open... I'll back off a bit.

Wodan

 

[Wodan]

Hey boys and girls, here's the new thread:

http://forums.civfanatics.com/showthread.php?t=186961

Enjoy! I for one will be very interested in everyone's results!

Wodan

 

[Wodan]

Copied from http://forums.civfanatics.com/showthread.php?t=186961

Hence, for the purpose of this testing I would definitely change that as most people will end up with hybrid economies and there won’t be great different for whatever economy you go. This is my view of the things, but I am open to suggestions.

Personally I feel that there is a HUGE difference in the game between a SE with a cottage city or two and a CE with a specialist or two. Whereas there is hardly any difference at all between a SE with no cottages and a SE with a cottage city or two, and there is hardly any difference between a CE with no specialists and a CE with a specialist or two (perhaps in a GP farm).

When it comes down to it, slapping a label on something doesn't make it any different than it is. And, bottom line, running your civ with a large number of specialists requires significantly different gameplay than running a CE. In addition, the gameplay benefits are different.

Adding a cottage city to a SE without otherwise changing the mechanics of running the SE does not color that leopard black.

Wodan

 

[Jonny211]

Hi all,

A fantastic debate and I've just played my first Monarch game running a SE... normally I'd run a CE and fall way behind in the tech race but with only a couple of cottage tiles I was ahead until the 1700's with only 3 cities, and only the capital had more than a couple of grassland tiles.

Somehow I had the Pyraminds, Parthenon and Great Library though. Not sure if I can pull that off every game though

Off home soon to experiment some more...

 

[CivScientist]

So I ran a little experiment to compare a specialist economy to a cottage economy. To begin with, I assumed I had just built another city in my empire on land that has 10 grassland tiles (NOT including the city), all next to rivers. The remaining tiles were mountain tiles and unusable. I also assumed that I was unable to acquire the Pyramids and that I have a GP factory city elsewhere in my empire that is so effective, I will be unable to ever make any GP in this city. Thus, any benefits a specialist may provide in producing great people will not be realized in this city. Futhermore, the timeframe of this experiment ends around the middle ages. Therefore, many of the civics that give a boost to either specialists or cottages/towns/etc. are still unavailable.

The question is whether I should build this individual city following the SE method or the CE method. If I am building using the SE method, I would allow my city to grow as fast as possible by building a farm on each of the 10 grassland tiles. Once my city reached it's full potential in size, I would start using specialists. Also, I switch to caste system just in time to be able to go beyond the set limits of scientist specialists normally set on the city. I found that doing it this way was optimal in terms of the commerce and beakers produced.

If I am building using the CE method, I would build a cottage immediately after a new tile was being worked. Thus, tiles were never fallow and cottages/towns/etc. were always worked. I did not build farms before cottages to accelerate my city's growth early on. Thus, I should note I may not be following the optimal building strategy here.

What I measued was economic output at various time intervals where both and equal one unit of economic output. Thus, if a city produces 10 in and 20 in , the economic output was 30. I then graphed the progression of the city's economic output following each economy. The graph I made is below:



Here's a graph plotting the difference between the two versions of the city for the first 200 years:



For the first 150 years or so the version of the city following the specialist economy was fairly close to the cottage economy version of the city. Sometimes the SE version of the city faired better and sometimes it faired worse. After that point, however, the CE version of the city started breaking away as tiles that once had cottages grew to towns.

So, why should I use the SE method on this city given my situation? Also, is there something wrong with my model?

FYI: I can post the excel spreadsheet I used to generate this experiment. I will need to label everything I did first so it makes sense. First, though, I'd like it if people can help me refine my model by telling me how I can go about building the SE version of my city better given my situation (ie, no pyramids, no GP produced by this city, etc.).

 

[CivScientist]

So, why should I use the SE method on this city given my situation? Also, is there something wrong with my model?

Some more questions in regards to my last post:

Is there a civic or building I am forgetting that gives my specialists a bonus (other than GPP)?
Can a SE city build buildings like libraries faster than a CE city?

 

[UncleJJ]

Sorry but you have introduced a great deal of experimental bias in favour of the CE and so your results are unbeleivable.

It seems that you have left out the effects of important buildings such as granary, forge, courthouse, library, market and so on. A SE should run slavery and install those buildings before running specialists. Large food surplusses give an extraordinary amount of hammers when run over a long period and buildings provide substantial bonuses (to hammers, beakers and gold) which you've ignored. With those buildings this city would not need Caste System but could run Merchants and Scientists. Factor in those bonusses and the early loss of beakers (due to delay in running scientists) is rapidly recovered in the long term. Your CE with only 1 hammer / turn will take 60 turns to build the granary alone and will soon be commerce rich but weak on bonuses. On the otherhand the SE can easily build the infrastructure needed to make up the diffrerence, it is commerce weak but hammer strong and hence strong on bonuses.

It is not clear from your description but I assume that you have a city square giving 2F, 1H and 1C plus 10 grasslands which are either cottages (maintaining +2 food) or as farms (maintaining 2 food + 1 per farm upto +12 food). That is a very strange assumption as most cities are based on at least one good food tile which gives +4 or +5 food surplus with the city square, then the grassland tiles are worked as cottages or farms.

Even without the Pyramids (a fair assumption) you cannot seriously assume that Representation is not available for the whole 200 turn period. It is easy to research Constitution by 1000 AD. Also why not assume Pacificm is being run ... that gives 100% GPP boost?

Using your figures, I guess you assumed that the SE was running 6 scientists (under Caste System but not Pacifism). That means they will produce 18 GPP per turn and therefore 1800 per 100 turns Making reasonable assumptions about other cities in a SE (which does not have a single GP Farm) at that rate I assure you it would produce at least 1 G Scientist in the 200 turns you consider. Late in the game that GS would be worth 2000 beakers if lightbulbing a tech and that would account for most of the difference you show between the CE and SE . If produced after 100 turns (say) that GS could be settled in the Science City (which I would build in my SE) and after Representation is researched he would be giving 9 beakers with a Research multiplier of +225% (Academy, Library, Uni, Observatory and Oxford) to give 29.25 beakers / turn. I argue that those 29 beakers should be credited to the city that produced the GS which would then gain 2900 beakers over the next 100 turns. Again this would account for the differnence between the SE and CE.

So there are many things wrong with your assumptions and most of them favour the CE over the SE and that is why your results are not useful. Don't be discouraged by my comments, I hope you revise your models and get closer to a more realistic set of conditions. You need to assume that key civics are researched and introduced into a game that lasted 200 turns or more and you can't ignore GPPs when running for that sort of time.

 

[Wodan]

So I ran a little experiment to compare a specialist economy to a cottage economy.

In addition to what UncleJJ said, I'll add a couple other two cents.

If I am building using the SE method, I would allow my city to grow as fast as possible by building a farm on each of the 10 grassland tiles. Once my city reached it's full potential in size, I would start using specialists.

Honestly, I never do what you just said. I use Slavery to whip critical buildings.

Also, whether running Slavery or not, I usually assign a scientist before reaching max growth. My personal preference is to have some science coming in throughout. Furthermore, your method would suggest that, once max growth is reached and you assign 2 scientists, you thereafter remove those scientists anytime you whip (rather than reduce working population), which again is something that I personally don't do. I keep the scientists and allow the whipping to reduce the working population. Occasionally there is not enough food income to quickly regrow so I change from 2 scientists to 1, but hardly ever all the way down to zero.

What I measued was economic output at various time intervals where both and equal one unit of economic output.

I don't think this is a safe assumption. Barring extensive shrine income or other means of economic income, neither CE nor SE will be running at 100% on the science slider for that length of time. Given the huge difference between effect of that slider on CE vs SE, one unit of commerce != one unit of research.

And, if you have shrine or other means of economic income, then the whole idea of comparing "economic output" in this fashion is out the window.

So, why should I use the SE method on this city given my situation? Also, is there something wrong with my model? Can a SE city build buildings like libraries faster than a CE city?

Several things, and yes absolutely there are some fallacies in your model. Ignoring Slavery alone is a huge difference. As UncleJJ said, the CE city will take, what, 90 turns to build the Library alone, plus it will require a Market and other buildings. The SE city, meanwhile, can whip those buildings quite quickly.

Frankly, even if I was running a CE, if I had a city site with all grassland like that, I would plant farms first, whip a bunch of buildings, and then cottage it later. But then, cottage maturation is delayed.

We can't simply compare numbers like you did... it's too complex an issue.

It's for this reason that we decided to run the test. To have real life data, and to attempt to capture all the various issues. By recording actual research etc, we capture the bonuses provided by whipped Libraries and so on. Also, we record whipped hammers etc.

I posted the link to that thread up above. As of today, MrCynical has just finished his game. I am most of the way through mine, though I'm thinking of doing a couple of them. Another player says he can't start until next week, so there's time to join in! The more data points we have from more people, the better our comparison at the end will be.

Wodan

 

[MrCynical]

As UncleJJ said, the CE city will take, what, 90 turns to build the Library alone, plus it will require a Market and other buildings. The SE city, meanwhile, can whip those buildings quite quickly.

While I agree with much of what you're saying, you're greatly exaggerating the advantage of Slavery to an SE. As you'll see when you look through my game logs (I've posted them now, I'll see about finding somewhere to upload the saves later), there is virtually no difference in the effectiveness of slavery under SE and CE. Since slavery (especially if we're talking about whipping things like libraries) works best for small city size, the vast majority of the food surplus is from a couple of tiles which will be present in both economies.

There's also the point an SE can run all its specialists, or have a greater food surplus than the CE, but not really both at the same time. To increase the growth rate, you have to remove specialists, and you're very unlikely to regrow more than a turn or two faster than an SE during the game phase when slavery is very effective.

 

[UncleJJ]

Originally Posted by CivScientist
If I am building using the SE method, I would allow my city to grow as fast as possible by building a farm on each of the 10 grassland tiles. Once my city reached it's full potential in size, I would start using specialists.

Honestly, I never do what you just said. I use Slavery to whip critical buildings.

Also, whether running Slavery or not, I usually assign a scientist before reaching max growth. My personal preference is to have some science coming in throughout. Furthermore, your method would suggest that, once max growth is reached and you assign 2 scientists, you thereafter remove those scientists anytime you whip (rather than reduce working population), which again is something that I personally don't do. I keep the scientists and allow the whipping to reduce the working population. Occasionally there is not enough food income to quickly regrow so I change from 2 scientists to 1, but hardly ever all the way down to zero.

Actually Wodan I agree with CivScientist and hence I disagree with you It is generally superior in a SE to work food tiles before running specialists while regrowing. There are a few exceptions when I would still be working specialists and have farms unworked, while re-growing from a whip, but that would be to do with manipulating the GPP pool rather than simply producing beakers or whatever.

This difference in regrowth strategy for a city is important when you look at the longer term and total hammers and GPPs your city has output. Generally speaking the more food you produce the more GPPs you make over time assuming you extract the same amount of food as hammers by slavery. Let me illustrate this by a simple thought experiment

Take two identical cities in a SE and work one of them as Wodan is suggesting and work the other I as advocate. Let's follow these two cities over say 100 turns of a game. They start with the same size and end with same size as each other and there is the same amount of food in the two granaries before and after the experiment. We will the two cities the same way to produce the same buildings and units so the same amount of food is converted into hammers in both cases.

Now my city will be producing more food since I always work farms before I work specialists while Wodan's will sometimes work a specialist instead of a farm. Since we have only used the same amount of food for growth (we are the same sizes at the beginning and end of experiment) and the same amount was extracted for hammers the only difference is usage of food must come from me running specialists for more turns. Each specialist turn costs 2 food per turn and each turn I worked a farm that Wodan didn't will cost him 1 food = 1/2 a specialist turn in the long term. Therefore although we ran specialists at different periods and for different durations of time we can conclude that over the 100 turns of the experiment I must have produced more GPPs... there is nowhere else for the food to go

I hope that is a convincing argument dealing with a very complicated situation. Total food production of a city over a period of time can be analysed in those overall terms. A properly run SE should attempt to maximise its food production at all times.

EDIT: Revised Wodan's losses as discussed in post 181
EDIT2: Re-revised as discussed in post 181

 

[Dr Elmer Jiggle]

I hope that is a convincing argument dealing with a very complicated situation. Total food production of a city over a period of time can be analysed in those overall terms. A properly run SE should attempt to maximise its food production at all times.

It's a convincing argument if all you're worried about is how many total beakers does one city produce over N turns, but in a real game, that's not the only concern. Your approach might produce more beakers overall, but Wodan's will produce them sooner. That acceleration of your research might translate into all kinds of indirect advantages.

Maybe you found a religion that you would otherwise be beaten to. Maybe you complete a wonder that you would otherwise be beaten to. Maybe you build your advanced military units sooner and win a war more decisively. Maybe you win a race to a bonus technology like Liberalism or Economics. And so on ...

Your approach probably completes buildings earlier (because your growth will come earlier so you can therefore whip earlier), but Wodan's should produce great people earlier, though it might produce fewer great people in total. Which is better? I don't think we can know.

Civ is a game of tradeoffs and in a real game you need to analize the complete situation and decide what approach is best. I would bet that most of the time, some combination / hybrid between your two methods is the true "best" strategy.