Sacrificing the Depressed

[Um the Muse]

Well, it's that time again. Seems like there's a new wave of topics on slavery yet again.

Anyway, a recent discussion got me thinking about whipping unhappy citizens away. Traditionally, people whip 2 citizens such that they grow the next turn and then take 10 turns to grow again. That costs you a certain amount of food for the growth, but the biggest cost is the opportunity cost--you're working with one less population for 10 turns (more or less, depending on how much you MM and how careful you are).

However, if you're only whipping unhappy citizens, you don't have to pay the opportunity cost. Instead, you have to pay an additional 2 per turn. On the one hand, if 2 per turn is worth more than the opportunity cost, then why did you grow that big to begin with? There's no need to worry about the happy cap.

On the other hand, if 2 is worth less than the opportunity cost, you should whip only after you reach the cap.

Does this sound reasonable, or have I forgotten something?

 

[vanatteveldt]

Suppose your happy cap is 6. What I would generally do is whip on (or around) the turn reaching size 7. If you have a 2-whip, you now have pop 5 and happy cap 5, and plan to grow back to six in time with the whip anger wearing off. So, at any time before or after the whip, you are working at happy cap.

If you want to whip every 10 turns, you would grow back to size ~7 by the time the whip anger wears off, so you would be working one above happy for ~5 turns.

 

[Zx Zero Zx]

I don't really pay THAT much attention to happiness and anger. I usually whip for the production overflow, or because I really need the extra production of units for a war. The Globe Theater is #1 NW for that reason. It gives you 1863824314321431 extra hammers. I go into whipping cycles, I don't whip everything always. There are circumstances that allow you to whip effectively, and there are other times when you should whip. You just need to play the game more, and understand the different situations you can get into. There is no one way to play Civ4, and no one strategy that is the best.

 

[MyOtherName]

However, if you're only whipping unhappy citizens, you don't have to pay the opportunity cost. Instead, you have to pay an additional 2 per turn.

Eh? You have one less productive citizen whether it's because you don't have that citizen or because he's angry. The strategy you seem to be describing requires you to bay both the opportunity cost and the two penalty.

 

[VoiceOfUnreason]

However, if you're only whipping unhappy citizens, you don't have to pay the opportunity cost. Instead, you have to pay an additional 2 per turn. On the one hand, if 2 per turn is worth more than the opportunity cost, then why did you grow that big to begin with? There's no need to worry about the happy cap.

On the other hand, if 2 is worth less than the opportunity cost, you should whip only after you reach the cap.

I wouldn't expect the opportunity cost to be better than 2 very often. Remember, as a rule we are whipping citizens off of our worst tile(s), so the opportunity cost is going to be low.

To some extent, you can approximate an unhappy citizen by thinking of it working a 0/0/0 tile.


But lets step back a bit. Consider a city at 35/36 - that's size 8, and let's set the happy cap there. Food surplus is +5. Because this is an illustration, rather than an example, we'll pretend that there's no granary.

Q: if we again grow at +5 food, where does the food bar end up?

A: 4/38.

In other words, although we have an unhappy citizen now, he hasn't actually eaten any food yet. There hasn't been any waste - waste only appears when you hit end turn with an unhappy citizen.

So you can work 8 tiles on turn 0, and then 7 tiles for 10 turns (assuming you can keep growth in check - training a worker or a settler or something).

The "traditional" approach works 6 tiles on turn 0, then 7 tiles for 9 turns... Aha, there's one more turn here where the old approach can work an 8th tile where you are stuck working 7 tiles.

Net profit - one 7th best tile?

Net loss - it's sort of hidden here, but it's basically the lower of your two food, and the value of the food gap if you don't grow a turn early. Example - At turn 9, both strategies are at 33/34. The early whip approach doesn't need to worry about the unhappy, so it can grow to 4/36 on T10, then 9/36 on T11. The late whip approach either works an unhappy citizen (4/36 then 7/36) OR parks an extra turn (33/34, 4/36)... it's 5 food "behind" but presumably it got some compensation for this.


One way where things have the potential to get very interesting is the situation where you are going to be whipping out workers or settlers. To avoid the penalty, you'll of course want to invest one turn of work into them before the whip. Growth will be stagnant on that turn, etc.

BUT... unhappy citizens don't eat food (EDIT: when training a worker or settler). So you would still have 5F contributing on that build, and you'll get it back in overflow hammers. So you can beat the game, kind of.

(Of course, the early whip approach can use the same trick - I don't think you come out ahead)


The other hidden cost that you need to pay attention to is the starting point - You need to grow, so the minimum food bar to start with (again, assuming +5F) is 31/36.

The earlier whip doesn't need nearly so much - he doesn't need 36 food to grow to size 9 right away, but instead 32 food to grow to size 7. So that approach can start as early as 27/36.


With a granary in place, the analysis gets messy, because you have two extra food to throw away! Growing from size 8->9, you get 18 food from the granary; growing from 6->7, you only get 14 food. So you can break even on food, by working the same tiles as the early whip case, and come out a full 7th best tile profit ahead.

Unfortunately, you get badly rooked on the fact that you have to start at 31/36, which basically means that you need 14 food after you grow to size 8, where the early whipping cycle needs only 10 food to get back to the lower sweet spot.


A complete analysis is more work, as you have to have a good answer to the question "what am I doing with tile assignments to assure that I'm not wasting food", but I hope you'll find that the simplifications here clarify the problem.

 

[Um the Muse]

I had to re-read what you said a couple of times there, VoU, but if I understood you right, you're talking about what Iranon advocated: whip on the turn you grow unhappy. I like that method because BUG will remind me that I'm about to grow, allowing me to put one turn into the build and then whip. However, that's not what I'm talking about here.

Let's give an example. If you have a happy cap of 5, you have +5 food surplus and you're at size 4. You have the choice to work a plains hill mine or a floodplain farm (just go with me on this). After growth, which will give you more production? Well, once you have a granary, food gives you more production, but working the floodplain farm will make you grow too fast.

So, what if you work the floodplain farm and grow to size 7 (or until you're just shy, whichever method you like better) and then two pop whip?

 

[MyOtherName]

Um: I don't understand what you are hoping to achieve. It's clear you're paying a penalty with the unhappiness... but I don't see what you think you are gaining.


Aside: in various circumstances, with a happy cap of 4, it is worth deliberately growing to size 6 before whipping -- but the main reason there is because it make a 3 pop whip possible.

 

[vale]

BUT... unhappy citizens don't eat food.

It has been a while but I'm pretty sure this is false. Maybe I am misunderstanding?

 

[ben-jammin]

Unhappy citizens not only consume food but contribute to maintenance costs. They only don't consume food if you're building a settler if I recall...don't remember if that only works for 1 unhappy citizen or more.

 

[vale]

Wow, I honestly never knew this. It also works for workers it seems too. I don't think you actually gain anything out of this but it is definitely counter-intuitive.

 

[Um the Muse]

I was trying this out on a game using Vicky as the leader. In the game, my more powerful tiles were giving me more food than I could deal with and plenty of room to expand into, so the question for me was, "should I build settlers and workers normally, or should I whip?"

Obviously, I don't want to stop working resource tiles. Also, Vicky is IMP, which changes things a bit. Still, it seems like I ended up with a stronger position when I let my cities grow such that I could 2 pop whip angry citizens instead of pulling off productive citizens.

The advantages to doing it this way were the continued usage of strong tiles without needing hills and if my happy cap happened to lift while I had angry citizens, those citizens become immediately available.

On the other hand, maybe I just had a really weird situation where in this single case it was better, but it was mostly a fluke. Or maybe I just happened to play better or got luckier the second time through. I was hoping that it was something that I could use in other games, too.

 

[vale]

The situation you are describing is extremely unlikely to be optimal. I can't guarantee it without seeing a save, but supporting unhappy citizens for any meaningful period of time really hurts production in most cities. If you were still needing to produce settlers and workers (which give you a convenient way of stagnating growth without working sub-optimal tiles) it would be very hard for me to imagine that this method was optimal.

What tiles did you have in this particular city?

 

[Um the Muse]

I'll post screenshots tomorrow. I'll just mention that I needed these food-heavy cities to also produce a few troops. So just stagnating wouldn't be good, I don't think.

 

[vale]

Well it isn't just stagnation, it is stagnation with big overflow, large population settler whips every ten turns that would optimize production in big food cities. Every turn that you have an unhappy population in which you aren't building a settler or a worker, you are losing 2 production per unhappy population. But yeah I'll have to see screenshots.

 

[VoiceOfUnreason]

It has been a while but I'm pretty sure this is false. Maybe I am misunderstanding?

Yes, but only because my writing there sucked (there was a lot going on). Updated to say:

BUT... unhappy citizens don't eat food (EDIT: when training a worker or settler)

For completeness: that's from a simple in game test, not from a code check.

 

[VoiceOfUnreason]

I had to re-read what you said a couple of times there, VoU, but if I understood you right, you're talking about what Iranon advocated: whip on the turn you grow unhappy.

I think we didn't quite manage to communicate. I think I've proved that for two pop whips, whipping before the growth happens rather than after is the superior play, assuming the rest of the micro works out.

As you correctly point out, that's a really big IF.

Let's give an example. If you have a happy cap of 5, you have +5 food surplus and you're at size 4. You have the choice to work a plains hill mine or a floodplain farm (just go with me on this). After growth, which will give you more production? Well, once you have a granary, food gives you more production, but working the floodplain farm will make you grow too fast.

So, what if you work the floodplain farm and grow to size 7 (or until you're just shy, whichever method you like better) and then two pop whip?

I don't think these conditions are specific enough to help clarify the situation. Let's get a picture of a representative city and go to work.

But I think we're going to find that, unless you are content to train workers and settlers, a food heavy city is going to have to whip sometime, and it's likely better early than late.

 

[cseanny]

Anyway, a recent discussion got me thinking about whipping unhappy citizens away. Traditionally, people whip 2 citizens such that they grow the next turn and then take 10 turns to grow again. That costs you a certain amount of food for the growth, but the biggest cost is the opportunity cost--you're working with one less population for 10 turns (more or less, depending on how much you MM and how careful you are).

I typically run little to no cottages and emphasize food and hammers in my cities so I need to rely on building research, wealth, and a limited amount of specialist so.....I find its generally best to whip at happy cap (bigger items and wars are different). On the whole this allows my citizens more time to build research and wealth plus abuse the benefits of que chopping. If my math is wrong here someone please tell me so I can fix it!! Just to throw in.....this is with the assumption that a granary is already in place so that regrowth comes quick enough that I get more hammers in the long run by doing so.

 

[MyOtherName]

Obviously, I don't want to stop working resource tiles. Also, Vicky is IMP, which changes things a bit. Still, it seems like I ended up with a stronger position when I let my cities grow such that I could 2 pop whip angry citizens instead of pulling off productive citizens.

This is the part I don't get -- whipping (when happiness limited) always pulls off a productive citizen due to the anger. What you describe doesn't make sense.

The advantages to doing it this way were the continued usage of strong tiles

Doing a 2 pop whip at size 6 already lets you do that. Growing to size 7 first doesn't change anything.

(And with a food glut and granary, a 3 pop whip at size 6 or a 2 pop whip at size 5 only deprives you of a single strong tile for 1 or maybe 2 turns)

without needing hills

Hills are usually good tiles, especially with a food glut. They're often neglected because slavery is often better and you need food to support using slavery as often as possible. But you describe a city that doesn't have that problem -- one with too much food to manage even with slavery.

and if my happy cap happened to lift while I had angry citizens, those citizens become immediately available.

Happiness cap lifts are usually predictable, and you can plan to be ready for them no matter what your strategy is.

 

[vicawoo]

1. If you grow before whipping, you lose 1 food with granary, 2 food without, for each population that you whip off.

So whipping early, you trade 1 tile turn for extra expense of growing.

2. Growing to size n+1 (1 above your happiness) isn't a continuous strategy, in which case you're always at 1 anger or 2 anger.

3. And these have your food bin at exactly the right time strategies, I think they're better off considered in the margin of error, because you often have to work suboptimal tiles to get there. In general you're not forgoing the chance to work a higher food tile (unless it's a normal farm over mine), so you're slowing down your growth to hit that point.

4. Growing past unhappiness, it depends how fast you grow. If you grow so fast that working a more hammer heavy tile (mine) won't generate enough hammers to lower your whip size, then sometimes it's an efficient inefficiency.

 

[JammerUno]

Growth is 20F +2 times your pop. If you whip before growing you don't need the 2F per pop you just whipped. I'm not sure if it's growth before production of production before growth, but either way that population point needs to produce 6F minimum to break even with whipping before growth (since you reduce growth costs by 4F by whpping 2pop). Not including the additional upkeep you're carrying.

In short, unless you have some riverside pig/corn you're not working at size 4+, whipping before growth is going to be best. So don't grow into unhappiness before whipping unless you need the pop for the amount of pop you want to whip, in which case it's fine.