The mathematics of slavery/whipping

[NFIH]

I'd like to better understand the optimal use of whipping under slavery to produce units faster for an early rush (or to otherwise mass units quickly).

My understanding after reading Siustil's strategy guide is that whipping generates 15 turns of unhappiness (at epic speed). He suggests waiting until that unhappiness has dissipated before whipping again.

If you do this, are you really building units any faster overall? It doesn't *seem* so. But on the other hand, whipping again while there is still unhappiness seems to pile on the unhappiness so much that that strategy isn't worth it either. And then the third issue is that the subsequent units post-whip are being built with less population so the units are building slower anyway, meaning that the total amount of turns to build, say, three units, is no shorter than building them without whipping.

All that said, is there some other timing I should use to whip units?

[wodan11]

There are other negatives to whipping. Every citizen whipped is one citizen not working a cottage or one less scientist. Every citizen working a farm to regrow, ditto. And, to boot, your city stays at small size which means it's a multiplicative negative... e.g., the city stays at size 2-3 instead of max size (5-9 depending on health/happy/game level).

Slavery also has a chance of civic unrest in your capitol.

And, running Slavery means you're not running another civic. If nothing else, Serfdom, which is a vastly underestimated civic.

[NFIH]

But I do find slavery immensely useful for whipping buildings that would otherwise take many, many turns (say 15+) to complete, especially when the city is already at a high pop like 10 or something. So that theater that won't complete for 22 turns gets insta-built for 3 pop, for example. That sounds like a win to me so I do it. And that's not to speak of key Wonders that I whip to beat the AI to them.

Do you not use slavery at all?

The problem with not whipping at all if you're going for an early rush is that you look at the build queue and you see that each axeman (or whatever) is going to take 12-15 turns (or whatever) and by the time you get five of them, it's hardly an "early rush" and it's not too likely you'll have much success against the targeted opponent, all things being equal.

There are other negatives to whipping. Every citizen whipped is one citizen not working a cottage or one less scientist. Every citizen working a farm to regrow, ditto. And, to boot, your city stays at small size which means it's a multiplicative negative... e.g., the city stays at size 2-3 instead of max size (5-9 depending on health/happy/game level).

Slavery also has a chance of civic unrest in your capitol.

And, running Slavery means you're not running another civic. If nothing else, Serfdom, which is a vastly underestimated civic.

[Hauptman]

Whipping does not just finish the queued unit. it adds a certain number of hammers per pop point. so what you may not have noticed is, whip that axe thats half done, next turn select another axe, and low and behold its done in 1 turn without another whip! Used in all of your cities and there is your early rush.

[Theben]

It depends on the city. A city with access to a few luxuries as well as hammers may be better off growing and not using the whip. A city with few or no luxuries and few hammers &/or lots of food would be better suited using the whip to finish builds.

For an early rush it's best to use multiple sources to build the instant army: the whip, chopped forests, and $$ to upgrade those early warriors into axemen/spearmen/swordsmen. On Epic speed I've used all three to build army of 8 Dog Soldiers, 1 worker, 1 settler, and 2 more Dog Soldiers for defense by turn 70 on Monarch level.

While I see the merits of what wodan says, what isn't mentioned is the merits of having these builds done much earlier. Timing various factors will maximize your use of the whip.

[slnz]

But I do find slavery immensely useful for whipping buildings that would otherwise take many, many turns (say 15+) to complete, especially when the city is already at a high pop like 10 or something. So that theater that won't complete for 22 turns gets insta-built for 3 pop, for example. That sounds like a win to me so I do it. And that's not to speak of key Wonders that I whip to beat the AI to them.

The question is: does that city need the Theatre before those 22 turns have passed? If not, you just wasted dozens of tile-working turns for nothing. Also, try to learn approximate times when the AI is completing certain wonders so you know if there's a danger they will beat you to it and a need to whip it. If you have intel, you can also check, for example, if the Industrous AIs have Stone/Marble as this makes a huge difference.

My rules of thumb for whipping (all do have exceptions):

1) Never whip unless you get an instant benefit (needed unit, sci/gold multiplier in an appropriate city, city is already shrinking, etc)
2) Never whip if the city grows beyond the lowered happy cap in those recovery turns (unless whipping for a happy building)
3) Never whip unless the city has at least +6/+3 (with granary) surplus food afterwards

As noted, don't use Slavery just because you have read it's good. With the correct situations, it's very possible and viable to win on Immortal level with never changing to Slavery with an early rush included. Then again, there are situations where your whole nation relies entirely on it for thousands of years.

[rah]

One negative of not being in slavery early is that early DOW by the AI. Most civ players will cheat a bit on their army early to get there workers/settlers and basic infrastructure out first. Most of the time you see the DOW coming a few turns before it happens but sometimes it is a surprise. If you're in slavery you can usually switch all you cities to units, wait one turn so you don't get the penalty for having 0 hammers towards it and then whip a unit in every city and move them towards the incursion. If not in slavery that turn in anarchy sometimes makes all the difference in the world fighting off their attack.

[wodan11]

But I do find slavery immensely useful for whipping buildings that would otherwise take many, many turns to complete, especially when the city is already at a high pop like 10 or something.

Slavery is most efficient when
1) the city has high food (lots of food resources)
2) the city is small size (and regrows for few food)
3) the city is at the max health/happy limit (but if at the happy limit it's tricky to do this so that it's most efficient)

So that theater that won't complete for 22 turns gets insta-built for 3 pop, for example.

Simply getting completed faster is not necessarily good. You have to compare the benefit you get, and what the benefit gets you, vs what you spent to get it. e.g., so in this case you whip 3 pop, which probably removes 3 from working hamlets/villages, and adds more to the happy limit which to some extent compensates for the -1 happy you now have. 22 turns from villages is roughly 60 commerce * library bonus etc. That's a pure loss as you won't have regrown the 3 by the 22nd turn. So from the face of it I'd say you chose poorly. However, we also have to look at what the city can now start building... say it begins building a university which means you get a university 22 turns faster (later in the game). Which results in a bonus of commerce later. But, research now is much more valuable than research later.

It's not a straightfoward thing.

Do you not use slavery at all?

Of course I use it. Slavery is extremely valuable.

The problem with not whipping at all if you're going for an early rush is that you look at the build queue and you see that each axeman (or whatever) is going to take 12-15 turns (or whatever) and by the time you get five of them, it's hardly an "early rush" and it's not too likely you'll have much success against the targeted opponent, all things being equal.

That's one of the best uses of Slavery. OTOH, if all you are doing is capturing cities which maintenance knocks you down to 0%-10% research before you have the techs to recover (CoL, currency, etc), then perhaps an early rush was not such a good idea after all.

[Dis]

But I do find slavery immensely useful for whipping buildings that would otherwise take many, many turns (say 15+) to complete, especially when the city is already at a high pop like 10 or something. So that theater that won't complete for 22 turns gets insta-built for 3 pop, for example. That sounds like a win to me so I do it. And that's not to speak of key Wonders that I whip to beat the AI to them.

Do you not use slavery at all?

The problem with not whipping at all if you're going for an early rush is that you look at the build queue and you see that each axeman (or whatever) is going to take 12-15 turns (or whatever) and by the time you get five of them, it's hardly an "early rush" and it's not too likely you'll have much success against the targeted opponent, all things being equal.

that's my strat as well. If you need an army asap, whipping is the way to go. You do delay some scientific progress etc, but you are gaining some enemy cities (hopefully). And it is nice for buildings that take forever to build, or in high growth areas with little production.

[sorinache]

Slavery ? Beuark !

I almost never use Slavery. It has a higher maintenance cost than Tribalism (which means I need to run 80-90% science instead of 90 - 100%) and there is a chance of Slave Revolt (a 100% chance actually, given Murphy's Laws ...) To what benefit ? That of whipping ... I can't imagine how you are playing, guys, but 12 - 15 turns to build an Axeman ? On Marathon speed maybe (never tried it yet), but otherwise I never experienced such long building times ... Do you use workers to mine the hills ? With a couple of mines you easily have 10 per turn and an Axeman is 35 on normal if I remember well, 50 on Epic if I'm not mistaken ... That's 5 turns, not 12 - 15 ...

22 turns for a theater ??? Wow, I can't imagine how you develop your land. Do you ever build workshops on grassland ? I usually build theaters (which I consider totally unessential buildings) only when I can build them in 1 or 2 turns, even on Epic.

To be precise, I've tried Slavery a couple of times and even tried whipping once, when I thought it made sense (had Three Gorges Dam, whipped Factory in conquered Tenochtitlan to get a +75% boost in hammer production). Still, after watching the time it took the city to regrow and the number of turns a number of developed tiles went idle, I think it was a bad idea.

Disclaimer : I play Emperor level (I might yet need Slavery to win on Immortal)

[rah]

Slavery is best early with high food cities. If you have a granery and a 6 food tile or even two. It does not take many turns to get back to three pop from 1. It doesn't need anything else and it can produce an army in no time. Or getting that worker or settler out 5 turns earlier can make a big difference early on. Chopping and whipping early in the game can be critical in terms of turns saved. Especially were if you let the city grow you will hit the happiness cap to early. It's also useful later in the game if you're on a military roll. There's no way to keep larger conquered cities happy unless you want to keep your army pinned down preventing revolts. Whip a theater and a couple of temples, (which also eliminates a lot of those foreign citizens that will be replaces with yours. And when you're close to eliminating a civ and the weariness is getting high a little whipping can buy you some time to finish them to get relief.

[DirtyMartini]

Common times I whip --
Early, when capital has 3-4 improved tiles but grows fast enough that it is at 5 or 6 pop and at the happy cap. Especially whip that 1st settler. Whip garrison in new cities, overflow goes to monument or granary.

The only time I commonly whip in the AD years is on those annoyingly expensive universities. I've usually chopped all the forests by that time and I need six of the suckers for Oxford. While it hurts to whip population that is working villages or towns, it is sometimes worth it to get Oxford running in the main science city.

During conquest of large developed cities, you might as well whip 'em hard. Many will still be choked by enemy culture when they come out of resistance and starve down uselessly if you you don't.

[Diadem]

Actually the amount of food surplus does not effect the efficiency of slavery. Only how fast your city grows and thus how often you can slave. But that's the efficiency of the city, not of slavery. Other things being equal, slavery is as efficient in a bad city as a good city. But of course in the good city you will benefit more from it, because it's a good city.

Slavery is
- Usually not worth it in a big city (Say 8 and up)
- Not worth it if you have very good tiles to work
- Usually not worth it if you do not have a granary

The basic effect is slavery is to turn food into hammers. Since bigger cities require more food per population point, slavery gives you more hammers per food in small cities. So in big cities, do not bother. And if you do not have a granary, it is 50% less effective, because a granary halves the food needed per population point.

Note that in a way, any tile you work also turns food into useful stuff. Every tile you work costs 2 food, and gives you whatever the tile gives. Working a farmed floodplains turns 2 food into 4 food and 1 commerce. Now that's a deal. Do not whip such a tile (unless you reach happiness of health caps, or need stuff in a hurry). Working a mined plains hill turns 2 food into 4 hammers. That might be more interesting to whip.

Let's use the mined plains hill as an example. That's usually not the best tile to work. So if slavery can not even beat that, it's not useful. So 2 hammers per food.

If I recall correctly slavery gives 30 hammers and one population point costs 2 * (10+city size) food. So a size one city grows for 22 food. Slavery will thus give you 1.36 hammer per food without a granary. Not worth it! With a granary it givies 2.73 food though, and that's a good deal. If the city is bigger the efficiency decreases. At size 5 the city grows for 30 food, and whipping with a granary gives you exactly 2 hammers per food. Same as a plainshill.

Of course there can be other reasons to whip. To get stuff (and thus the benefits of the stuff) earlier. This is most obvious for units to defend against attack. But whipping a granary is usually also worth it, because while you may lose out on the whipping, the 10 extra turns of having more efficient food production more than compensate for that.

[rah]

90% of my whipping happens early when happiness limits are 5 or 6, or whipping conquered cities down to managable sizes instead of watching them starve down. I might as well get a few buildings or troops out of it.

[Theben]

Normally I whip out a monument, archer, or the last turns of a worker. Sometimes a wonder if I feel the AI might beat me to it. In the very beginning from my capital just about anything can be whipped since it normally has a lot of food bonuses.

Once my civ is running comfortably along I leave slavery for something better. And in late conquest I'll buy what I need via Uni Suff- I don't want to risk having one of my good cities revolt.