Does using Slavery hurt my game?

[Stolen Rutters]

I'm seeing that Slavery is not helping. It's either comparative or worse in all cases. Given that there is also a cost to running the civic (we could be running serfdom), this is telling.

Wow, I have seen that slavery only helps in some cases so I don't do it much except at the beginning of the game when I am racing for the good city spots against an AI (and need that Iron/Gem/Cow city, for example). On emperor, sometimes you need that second city more than you need to grow your existing one at the start.

Also, the Granary doesn't help. I think the Granary will be most helpful where Slavery is used to excess... repeated whipping. Interesting though, that a Granary is not a good idea if you don't plan to use Slavery. Challenges the notion that a Granary is always a good first build.

It's not always a good first build but I disagree that it's not a good idea without slavery. The granary is more for the health bonus and faster growth. Sometimes that extra two health will pop that city one more level (well, I admit I frequently whip it). Some games the happy cap is the problem, others it's the health. Depends on the map. There are cases for a granary even if it's not an "always build"...

One thing that complicates things is that we aren't considering either overflow food or the benefits of non-Library buildings. Presumably, we could build temples, theatres, etc, and thus increase our city size, and thus get more future potential. However, conceptually we could use the overflow hammers to build these other things, and thus we would have great research coming in along with the future growth as well.

Wodan

Good job on the test. I have learned not to over use slavery, just from playing the game (it can kill you if you use it wrong) but I can see that trading away pop for hammers can hurt far more than I imagined, if my goal is not aligned with the reason for the whip.

 

[lilnev]

I find that I use slavery the most for my second generation of cities -- after the first war has been concluded, and my empire is up to 8 cities, the five recent acquisitions will have good food (resources are developed, and the AI loves its farms) but no infrastructure. So I stall them around size 3-6 for a while (with, perhaps, a +5 food surplus and a happiness cap of 8) while I whip out the standard buildings: Granary, Forge, Courthouse, Library or Barracks. My empire's initial core cities often grow up to their happiness caps while building individual buildings the normal way -- I don't need slavery, or minimally, and I'd rather be working tiles.

I do think that the hidden cost of not maturing cottages is sometimes underestimated. You're missing out on the one commerce this turn, plus one for the turn that the cottage->hamlet transition is delayed, plus one for the hamlet->village transition, plus one for village->town. Add one more for a financial leader, one if you would have Printing Press by the time of the hamlet->village transition, and two if you would be in Free Speech during village->town, and it could be costing you as much as eight commerce. Admittedly most of those are discounted as future commerce, but still....

peace,
lilnev

 

[Hans Lemurson]

Not working a cottage for a turn pushes the entire growth-curve back by one turn. Pushing the curve back one turn means that you have one turn less of the fully-developed commerce. The true cost of missing a turn is taken away from the future, not the present.

 

[VoiceOfUnreason]

second city starts without culture, with only a handful of tiles improved (or even none) and a worker (or 2, but I tend to go 1 worker /city on average).

Here, you cannot expect to work a good tile with your second pop, for a while. Which in return makes it worthwhile to build a granary to have a faster growth later.

My instincts (already demonstrated to be lacking here) tell me that this isn't such a big deal, as you don't lose all that many worker turns. A farm is what, 5 turns? A cottage 6? So you are working 5 turns at +2 before the farm kicks in +3, then grow 4 turns later (presumably leaving one or two turns on the second farm), then when the second farm kicks in you've got 6 turns @ +4. which is enough time to get a cottage finished. So long as you've got something like a grassland forrest on hand to give an extra hammer when the next improvement isn't ready, I think you'll find the cottages are built in a satisfactory manner.

If you start with a +3 tile (the unimproved rice, for example), then math changes a little bit. In any case, it is an easy enough test to do - simply take the save I provided, enter world builder, remove the improvements and gift yourself a worker (you may also need to gift yourselves the techs).

I'm speculating that in a floodplains city, the math is very different (faster growth + slower improvement rates).

I think you'll also get different results if the happy cap is lifted to seven or so. The extra growth room gives the granary more leverage, and delays the point where you want to start working the cottages on the hill. Essentially, this gives the whipping case more time to catch up.

 

[cabert]

I think you'll also get different results if the happy cap is lifted to seven or so. The extra growth room gives the granary more leverage, and delays the point where you want to start working the cottages on the hill. Essentially, this gives the whipping case more time to catch up.

totally agreed!

but once again, this test is a eye opener, i don't want to imply it isn't a good one

 

[UncleJJ]

I think you'll also get different results if the happy cap is lifted to seven or so. The extra growth room gives the granary more leverage, and delays the point where you want to start working the cottages on the hill. Essentially, this gives the whipping case more time to catch up.

I agree that raising the happy and eventually health cap will significantly skew the results. So even if it is true for the development of a second city that would normally be a special case and the predictions can't be extended to subsequent cities (assuming normal development and research patterns).

Incidently, if I had a city like the one in the test (cottagetown was it?) I would not bother with devloping cottages even if running a CE. Assuming the capital really was tied up building the pyramids (although a bit more competantly than in the test ), then the second city has many better uses than merely growing cottages including; making a worker, another settler, an axe stack (if there was an AI near) and even whipping in a library and running scientists for 17 turns to get an early academy for the capital.

The test situation and premises just don't make sense to me and I am very suspicious of the conclusions. The test needs to be extended with say a happiness and health cap of 10 (or perhaps let the civ research the early techs (e.g. Priesthood and Monarchy) which unlock more options for slavery to benefit the city rather than researching computers which does nothing.

 

[aelf]

I tend to fall behind when the debate concerns such detailed math, but I would really like to see a conclusion to this issue. Proponents of slavery made such a strong case for it in the past. It's interesting to see a revisionist view. Of course the former emphasised production while the latter commerce. Is this a case of warmonger strategy vs. builder strategy?

My own instincts tell me that once there's plenty of food and low happiness caps, slavery becomes a powerful tool. I also whip for macro reasons - foreign policy needs and questions of long-term strategy. But I know there are situations where slavery is counter-productive, and I think that happens when commerce starts to become significant (i.e. when cottages are mature and techs are more expensive).

 

[Wodan]

No time or brainpower today (I'm way short on sleep) but if nobody else jumps on this I'll do a limited test of higher size tomorrow. Basic conditions:

simply grant more techs... no need to research them. Give ability for temples, theatres, aqueducts. See if whipping those buildings to permit larger city size gives better results than the best case determined for the smaller size.

Feel free to comment if anyone has suggestions on conditions or heuristic for test cases.

Wodan

 

[Wodan]

Someone commented about production... there's no argument there. Obviously, if you're whipping Axes or something, then this will both hurt your research as well as give you more production. That's a "duh".

Anyway, please recognize that in the OP I stated my suspicion... the AI does so well on research (in some games) partially because they don't whip the bejesus out of their population like so many of us do. There is an underlying presumption that the city can "catch up" using the building bonuses but the test is showing that is not the case.

Let's see now if the same is true for larger cities....

Wodan

 

[Paeanblack]

I tend to fall behind when the debate concerns such detailed math, but I would really like to see a conclusion to this issue. Proponents of slavery made such a strong case for it in the past. It's interesting to see a revisionist view. Of course the former emphasised production while the latter commerce. Is this a case of warmonger strategy vs. builder strategy?

Many of those debates were referring to 1.52, which had the rounding bug. When fully exploited, the bug made slavery far too powerful. When the Globe Theater/Heroic Epic city could sustain 112 Hammers/Turn in military production by 200AD, basically nothing else in the game mattered. Now slavery is toned down, but still has it's uses.


I have a strong hunch that, post 1.61, it's impossible to increase research by whipping commerce multipliers. When the city is small, the bonus is too small to pay for itself, and when the city is too large, it takes too long to grow back. The window of opportunity in the middle is tiny if it even exists at all.

 

[Wodan]

Well... I think Slavery is always going to be useful on low-production cities. Keep in mind that on our test there are hill mines there. What about coastal or island cities with literally nothing except the 1 hammer from the city tile?

Wodan

 

[Wodan]

Somehow I missed this post.

You are forgetting another good use of a library - hire specialists.

Hmm. Especially if running representation. Good point. Might be another good follow-up case for the small city size. See how it compares.

Granary is NOT a waste here, it will continue to help city growth

The results seem to indicate that whipping a granary helps only if repeatedly whipping multiple items, or if needed to increase max city size.

and this city will continue to whip units whenever possible. Whipping in high-food low-prod cities is almost always good. At least as long as you can afford more military units.

Still, with one food resource your city will quickly grow to unhappiness, and then of course it's just as well to whip something anyway.

I'm not sure I agree. Whipping one unhappy citizen results in one less worked tile. So, again, you're sacrificing some commerce in order to gain production.

Wodan

 

[UncleJJ]

Somehow I missed this post.

The results seem to indicate that whipping a granary helps only if repeatedly whipping multiple items, or if needed to increase max city size.

Wodan

That is stating the obvious A granary effectively doubles whatever food production the city has for growth and slavery. Used under the right conditions it is the more powerful economic building in the game.

The circumstances of this experiment are highly contrived and clearly do not favour building a granary since there is only one other building to whip in (the library) and the happiness cap is low for over 100 turns so it cannot benefit the city by speeding growth.

A more realistic experiment that would allow the combination of slavery and a granary to compete with the Fat Hen strategy (see below ) can easily be designed. Simply add 2 or 3 religions to cottagetown and give (or allow techs to be researched) that allow temples and monastries to be build. Add a source of copper so axemen can be built... and add credits for axemen. That leads to the question how much is an axeman worth in terms of beakers (say 35 hammers and each hammer is worth 2 or 3 beakers). Also allow monarchy to be researched or introduced after so many turns. Let a worker be built and link up the rice for more health and so on. The current test is so unrealistic that the Fat Hen strategy is bound to win. Let's get real.

Facetious Explanation:
The Fat Hen strategy is the default CE strategy of the broody old mother hen sitting on her eggs (cottages) until they hatch into her precious little chicks (towns). All other considerations such as city growth, producing an army, building any infrastructure are subordinate to to hatching those eggs. What poor old mother hen doesn't realise is that not everyone loves her little darlings as much as she does.

 

[cabert]

from now on, I'll call a CE a fat hen strat

 

[aelf]

Many of those debates were referring to 1.52, which had the rounding bug. When fully exploited, the bug made slavery far too powerful. When the Globe Theater/Heroic Epic city could sustain 112 Hammers/Turn in military production by 200AD, basically nothing else in the game mattered. Now slavery is toned down, but still has it's uses.


I have a strong hunch that, post 1.61, it's impossible to increase research by whipping commerce multipliers. When the city is small, the bonus is too small to pay for itself, and when the city is too large, it takes too long to grow back. The window of opportunity in the middle is tiny if it even exists at all.

My guess is post 1.61 proponents of slavery tend to be war-oriented. There's nothing better for war than the ability to whip up an army quickly. There's the draft too, of course, which IMO has received too little attention, but that comes later and only if you have the right techs.

 

[Wodan]

UncleJJ: I both agree and disagree with you.

Regarding using granary/whipping to increase city size (that is what you're implying), the test I'll be doing this morning should compare to the "fat hen". So, basically, this will be comparing a city which whips the builds necessary to grow larger (and thus work more cottage tiles) vs. a city which is smaller but working cottages the whole time.

Regarding Axes or any other kind of production (wonders, barracks, you name it), frankly I feel that is beside the point. If the player is whipping axes or whatever, then he already has other priorities than research. Trying to compare a whipped axe strategy to a pure research strategy is a huge undertaking and isn't what this thread is about anyway.

Wodan

 

[Antilogic]

All right, I'm an advocate of the slavery civic myself, and I don't think your setup is well-suited to slavery. My strategy for using slavery has a simple axiom: never, EVER, whip in a cottage city unless you are being invaded.

I've found that leaving the cottages growing is a good idea...and a great idea is to use a city with at least two, preferably three, food resources. Let the city grow to the 4-6 range, then whip down to 2-3 and work all your food resources to grow back in 10 turns or just under.


EDIT: I should mention that I have only skimmed the last page or so--this thread grew pretty quick.

 

[Wodan]

My strategy for using slavery has a simple axiom: never, EVER, whip in a cottage city unless you are being invaded.

Let the city grow to the 4-6 range, then whip down to 2-3 and work all your food resources to grow back in 10 turns or just under.

Those 2 sentences seem to me to contradict each other.

BTW welcome!

Wodan

 

[Wodan]

Again, col 2 is #turns to complete Computers, thus lower is better. See notes at bottom of this post. col 3 is #hammers at end, so higher is better.

Revised Results for Case 1, with Additional Tests
A 142 86 FC auto (Food & Commerce), whip library when 1 unhappy and about to get 2nd and costs 2 pop, then stay on FC until max pop (happens immediately), then manually work all grass cottages, when whip unhappy goes away work a 5th cottage.
B 144 90 FC auto, whip granary when costs 1pop size 4 and about to get size 5, whip Library when 1 unhappy and about to get 2nd and costs 2 pop, then stay on FC until max pop (happens immediately), then manually work all grass cottages, when whip unhappy goes work a 5th cottage.
C 141 88 FPC auto until unhappy, then switch to working 1 mine 1 farm and 3 cottages (only need 1 turn at this state to build Library but would have stayed longer if needed), then all grass cottages.
D 145 57 Work all grass cottages. Whip when size5 and about to get 1 unhappy.
E 146 155 Work 3 grass then 2 hill cottages. No whip.
F 135 215 FC auto until max pop, then 3 grass and 2 hill cottages. No whip.
G 136 231 FP auto until pop 5 (the happy limit), then 3 grass cottages and 2 hill cottages.
H 138 271 FP auto until pop 5 (the happy limit), then switch 2 farms to 1 grass cottage and 1 plains hill mine (stagnate w/max hammers) until Library built (5 turns), then 3 grass and 2 hill cottages.

Results for Case 2
2A 71 344 FC, whip granary when size 4/5 and about to get 5th (costs 2), build temple normally by FC until 4/4 then work rice/2 cottage hills/1 grass hill (takes 5 turns using whip overflow), then whip 2nd temple when 1 unhappy about to get 2nd and costs 1, then build 3rd temple with whip overflow and working 2 hills cottages, then rice and cottages until whip unhappy goes away, then all cottages.
2B 72 230 FC, build normally: temple, temple, temple, granary When unhappy limit reached, work all cottages (2 hills first). As temples built, temp back to FC to quickly grow to next size. Note: there are periodic periods of unhealthiness, which simply reduces the food income slightly; no big deal though it did mean that the rice needed to be worked instead of the 8th cottage, until the granary was built.
2C 74 338 FP: temple, temple, temple, granary, then all cottages. After first temple built, switch to FC until hit max, then work both hill mines and grass cottages until 2nd temple and granary built.

Note: Selecting FC auto works rice and farms first, and then cottages when size 4 and above. Working all cottages is not possible with automation and must be done manually.
Note: Working all cottages is not possible with automation and must be done manually.
Note: FPC auto works rice and farms, then 2 mines, then cottage with 6th pop
Note: Ending research per turn from test city on case1 is 26 beakers, case2 is 33 beakers.

 

[Wodan]

My comments: The additional 2 tests I did for case1 were mentioned before. G works farms&mines than cottages but skips the Library. H does the same with a Library. A few more turns and case H would have caught up with case G on the research.

Case 2 is a new one, where we add in temples and hook up the Rice (to allow Granary health benefits). I also ran this one to 1000AD to permit the Towns to mature up to size 8, thus end results can be compared apples to apples.

2A got the best research and production too, which was the Slavery case. The production was almost identical to the FP (work the mines) case but research was ~33*3 = ~100 beakers better.

Wodan