Using slavery effectively?

[VSarius]

Slavery is undeniably powerful but there seems to be this break point where people realize how powerful it is, but start to overuse it, and then the point where you actually start to use the whip efficiently. Generally, if you can grow unto financial cottages/grass mines or specialists, you shouldn't whip past the opening except in the opening/basic infrastructure. If you are whiping almost everything, you didn't specialize cities smartly.

 

[Kid R]

If anyone wants a good example of when slavery might or might not be ideal, play 50 turns or so of the current Deity School game. The starting area is riddled with riverside grass hills and can probably be played close to optimally with no whipping at all.

 

[Mantic0re]

IMO where people make a big mistake in these calculations is *opportunity cost* - specifically the potential gain from working a different and much better tile somewhere else. Planning for potential yield from all cities and potential cities in tandem is probably the hardest part of this game.

I think this is due to the fact the game requires managing shifting priorities in multiple time horizons. These types of analysis tend to make calculations with static priorities for practical reasons. They don't help players answer questions regarding how these priorities might change or how to manage the available decisions when they do.


There is a missing "likelihood function".

 

[VSarius]

If anyone wants a good example of when slavery might or might not be ideal, play 50 turns or so of the current Deity School game. The starting area is riddled with riverside grass hills and can probably be played close to optimally with no whipping at all.

Good example. Though, a whip for a settler to claim a particular site may be worth it. Though I agree, the level of production makes slavery very meh.

 

[ben-jammin]

A lot of grassland mines (riverside or not) don't make whipping much less effective in the expansion phase. It can be about getting the hammers sooner, not necessarily getting more of them over a certain number of turns.

 

[TheMeInTeam]

A lot of grassland mines (riverside or not) don't make whipping much less effective in the expansion phase. It can be about getting the hammers sooner, not necessarily getting more of them over a certain number of turns.

This is a misconception. You are whipping in the expansion phase so that you do get more hammers over a certain amount of turns, and with expansion that is typically both short and long term.

Without maintenance cost as a consideration brainless spam-expansion onto as many strong tiles as possible would win out routinely, and anything that speeds up that process would be a winning move.

Would you work a grassland mine over a pig? No. Then why would you delay whipping a settler to get a pig just to work the mine?

 

[blunderwonder]

i dont know about the rest of you, but i whip all the time, even into unhappiness, its a way to quickly get things out, and i dont know the exacts, but its a very good way to get production when land is not ideal... If you're worried about whipping away cottage tiles or gold THEN you specialized wrong. But whipping in a production city, well... thats what its there for, and i dont mind 2 pop whipping for 60 hammers when ill only be missing 1 or 2 grassland mines for 5 or 10 turns. With proper amounts of food, granaries and tile improvements, its pretty difficult to overuse the whip, every 10 turns, and enough food, and i have to whip so much that it sort of becomes tedious. The only time im not whipping at least every 10 turns is when im building wealth or failgold.

by the way, in the recent diety game, theres 2 irrigated corns, it'd be difficult to get even close to the amount of production whipping provided from jumbo hills.

as for mines being riverside, well farms usually are too Oo.

/e, and also, not using the whip can be outright silly, having 2 unhappy citizens that can easily be whipped away just makes for maintance costs. i will sometimes whip those guys away even taking the penalty for having no hammers invested.

 

[ben-jammin]

This is a misconception. You are whipping in the expansion phase so that you do get more hammers over a certain amount of turns, and with expansion that is typically both short and long term.

Without maintenance cost as a consideration brainless spam-expansion onto as many strong tiles as possible would win out routinely, and anything that speeds up that process would be a winning move.

Would you work a grassland mine over a pig? No. Then why would you delay whipping a settler to get a pig just to work the mine?

I'm talking about one city, not your overall empire-wide hammer output. Even if it means no net increase in hammers over the next 10 turns in the city in question, you might want to whip a settler because it finishes the settler 3-4 turns faster. Which often means getting the spot you wanted and maybe blocking the AI. That's what I mean. There's no mathematical gain if you're just calculating how many hammers that city can produce in the next 10-15 turns, but you still want to whip. And in a realistic scenario you would actually be gaining hammers anyway because you can work a higher yield (food) tile instead of working a mine to avoid growing over the happy cap.

You can't just ignore maintenance, getting your early settlers out faster wins out even with maintenance costs if you're blocking land that you can later settle at whatever pace suits you best.

On re-reading I'm not sure if we're even disagreeing or not? I've had a few beers.

 

[TheMeInTeam]

There's no mathematical gain if you're just calculating how many hammers that city can produce in the next 10-15 turns, but you still want to whip.

That's just not accurate. There is a huge mathematical gain, but it's too complicated for people to want to bother quantifying. The difference of getting a city or not, and the potential yields from that city, definitely amount to much more than "no mathematical gain".

And in a realistic scenario you would actually be gaining hammers anyway because you can work a higher yield (food) tile instead of working a mine to avoid growing over the happy cap.

That's what I was saying. Eventually, the ability to whip and get a better tile goes away, but in the expansion phase there tends to be a better tile.

On re-reading I'm not sure if we're even disagreeing or not? I've had a few beers.

Mostly agreeing. You and Fippy are saying that single-city efficiency calculations will lead to poor empire wide decisions, and I am pointing out that it is silly to do single-city efficiency calculations but that it IS possible, though tedious beyond my will to bother, do to them empire wide and arrive at the correct conclusion on whether one should whip.

Conceptually, it's a simple "yield from doing this vs opportunity cost" consideration, just a difficult one to execute and still consider this a game and not an assignment .

 

[Iranon]

Done wll, whipping tends to win on the per-city level as well until double-digit sizes.

However, that may involve tedious micromanagement for small gains in stable cities at their happy/health cap.
A more relevant advantage: In cities below their caps you grow faster and work more tiles, turning everything into hammers only on the last turn.
The same may also apply when whipping cap raisers.

And of course, it's golden if your regular production options suck (desert hills, forests) or you don't even have those.

 

[Manco Capac]

Best detailed, analytic explanations about how to exploit slavery I have read:

Sorry for O/T and wall of text. I recommend skipping if you're not into simple mathematics.



With which city is it better to whip?
A

Spoiler :

B

Spoiler :

Well, obviously city B. You can see the amount of turns to grow are significantly less even though the population is higher.

That was my point about infinite 3F tiles...

~~~

There's a way to quantify your expected production with whipping.
All numbers assume NORMAL speed. It's a simple exercise to convert for different speeds, although the result is the same.

All multipliers are ignored as they're multipliers... they will affect everything in the same fashion and not actually change the results as well.

Consider the following:
1 population = 30H through whipping

This is the actual yield from whipping 1 population, before any multipliers.

There are caveats as we all know... and loss of tiles.
Higher population costs more upkeep. More civic upkeep as well... for the sake of keeping things simple, I will only consider food and hammers and ignore everything else. Keep in mind that in reality it is not so; you'll have to develop your own model to make the right decisions.

That being said, the whip converts food to hammers. If you know the amount of food required to grow, then you have a way to determine the efficiency of food:hammers ratio.

Food to grow = 20 + 2x
Where x = city size

Spoiler :

For example a size 5 city requires 20 + 2*5 = 30F to grow.

The F:H ratio is thus a simple equation: 30/(20+2x)

Spoiler :

As you see, the ratio is better at smaller size, which is why people say "whipping at small size is better".

How does the granary factor in?

Basically, it doubles the value of food... because it stores 50% of the food. The F:H curve is the same, but the numbers doubled:

Spoiler :

Food to grow becomes (10 + x) instead of (20 + 2x).

Without a granary, whipping becomes inefficient at size 5.
With a granary, whipping is efficient all the way to size 20.

That is, the value of 1F is greater than the value of 1H.

~~~

This brings me back to city A and city B example above. While the last section would appear to contradict the general feeling that city B is better for whipping (after all, you're converting food to hammers at a ratio of 2.7 vs 1.5), you should, in fact, consider the city yield and not the ratio itself.

The expected production of city A is:
(F:H ratio)*(food surplus) + (hammers per turn)
(30/11)*(3) + (1)
9.2 H

The expected production of city B is:
(F:H ratio)*(food surplus) + (hammers per turn)
(30/20)*(12) + (1)
19 H

~~~

This simplified model allows you to see the production worth of various tiles as well:

take 1 grassland hill mine, for example. 1F3H on screen.

What's the actual city yield of working this tile?
-1F+3H
That's right, you're losing 1 food by working this tile over a food neutral tile of 2F.

What's the worth of a grassland hill mine?

Spoiler :

That's right. A grassland hill mine is actually WORSE than a grassland forest until size 5... if you intend to whip.

Here's the chart for grassland farm vs grassland hill mine:

Spoiler :

Basically, a 3F tile is better than a mine, for whipping, up until size 10.

A lot more can be said and done with this model, but this should give you something to ponder about.

~~~

DISCLAIMER: As I've said earlier, this is a SIMPLIFIED model and ONLY takes into account and . Everything else is neglected. Results may vary with use.

 

[blunderwonder]

Thats interesting, though i suspect those numbers are a bit strange, you cant whip at size 1, so (A) its a moot point anyway. That said, its an intersting take on the idea that its always better to whip at a lower size, because its quicker to grow back. If you have 6 grassland pigs, it may well be easier to grow from size 3 to 6, than to grow from size 1 to 4. Ive never really bothered to calculate the amt of food to grow, so the 20+2x is the most valuable piece of info there for me, it means to me that a granary makes the value of food much more than double early on, and for quite some time because it splits the amount of base food needed, not just the additional amount. For arguments sake say the 20 was a 100, thats my point, it helps alot more than double early on.

at size 15, it cost 50 to grow, 15 farms is 47 food with the base tile, thats +17 food, which is 3 turns to grow, am i doing that right? with a granary, 1 and a half turns? That math doesnt seem to right to me somehow, but then i dont typically farm so many tiles Oo

As far as the hill graph goes, im not sure i understand it, not sure what production value means.. it appears to be the same as FH ratio, its 2.6 early same a previous graph. It appears to me some funny math, as when you add GF+GHM = 3 always, i doubt that graph is accurate for some reason Oo Its just funny to say the value of a ghm in terms of hammers, when food is also involved - im finding it difficult to explain but it seems to more appropriate an example to use PHM's instead, and use farms/ghm's to balance out the necessary food. I just mean that you cant work all GHM's at size 11, theres not enough food, so it gives food a bad break.

Anyway, thats a wonderful presentation you put together there Oo, i always find when people throw the math at you its annoying because its not explained Oo, yours is nice though.

now if you can make a presentation about the battle odds thing that would be great

 

[Nials]

Thats interesting, though i suspect those numbers are a bit strange, you cant whip at size 1, so (A) its a moot point anyway.

The number for a given city size is not the efficiency for whipping at that particular size, but rather the value for the average city size during regrowth.

As an example, if you do a 1-pop from size 2 -> 1, it's the effiency for a city size of 1 that you want to be looking at. With a Granary, it takes 11F to grow to size 2, so we get a : conversion rate of 30/11 = 2.72.

For a 3-pop whip from size 6 -> 3, we want to be looking at the value for a city size of 4 since this will be the average city size during regrowth - 90/(13+14+15) = 2.14. Whipping 1-pop at size 5 similarly gives us 30/14 = 2.14.

 

[blunderwonder]

good call, i really just ignored the first one all together because you cant whip at size 1, i suppose those values are really for a size 2 city though Oo.

and shouldnt it take 22 food to grow, he lists the growth thing as 20+2x...?



I think a graph detailing farms would be neat though, its clear that the second city would be better to whip, which is sort of counterintuitive the food/hammer ratio is higher in the first, but theres alot less food/tiles, in the 2nd because the 3F tiles add up so well early on and so many, its more efficient to whip in a size 6 city with all farms than with a size 2 city, idk its strange..

 

[Nials]

and shouldnt it take 22 food to grow, he lists the growth thing as 20+2x...?

Only if you discount the Granary