The Whipping Thread

[Blake]

This thread is for a definite guide to whipping, when to whip, when not to whip and for discussion of case studies and such.

Whipping is an exceedingly useful tool and in some cases it's absolutely better than the alternative - the classic example of this is with training workers/settlers in smallish cities with a granary, it's better to use food to grow population and then whip the population to death, than to use the 1:1 food:hammer conversion you get from training these units normally.

Advise on when to whip varies, advice is often giving like "Every 10 turns" (or after the existing whip anger fades), altough in order to whip so often it is nessecary to adapt your playstyle, this typically involves increasing the governors growth-emphase (best done using Emphasize Food+Production+Commerce), founding cities at sites with ample natural food (pasture, crops or seafood) and building plenty of grassland cottages.

It is nearly always good to whip population which is unhappy, unhealthy and/or working poor tiles. When using whipping as a cure for unhappiness it is critical to kill off 2 or more population by whipping something large.

While discussion has focused on when to whip for a long time, it's also important to understand when not to whip. These rules of thumb provide a guide to both when to and when not to whip.

In short, you shouldn't whip a city unless one of the following conditions are true:
1) The build will increase growth (ie granary, or +health in an unhealthy city, or possibly a worker)

2) The city is already at or above the happy cap - in this case be sure to kill 2 or more population, preferably more if the city grows fast.

3) The city is training a worker or settler, has a granary, and is small. Generally in this case it's better to turn food into pop then whip the pop than to put food directly towards the worker/settler (better in absolute terms).

4) The city's workers are on marginal terrain. Here marginal is declared as forest or worse and remember unhealthiness reduces all yields by 1 food - it's worth working a town in an unhealthy city, it's not worth working a forest in an unhealthy city (1-1-0 or 0-2-0 tiles are terrible).

5) The city has special income from settled Great People, Shrine etc and the building will multiply that income. It may be that the increase in special income will outweigh the income from the tile workers remaining alive for the duration of construction. Trade also acts as special income, to a degree (small cities earn slightly less trade).

6) You just really need it asap (defensive unit, or wonder).

Furthermore, one of the following should be true:
A) There is no existing whip anger or it is about to fade. Most enthusiastic lashers agree that cities should spend every turn "cooling down" but that actually stacking whip anger is a bad idea.

B) The whip anger plus existing whip anger will fade before the city hits happy caps or the build will increase happy caps (this can be from whipped units conquering new happy resources, or just form whipping temple, forge, marketplace etc). Sometimes you just have a lot of happiness.

C) You just really need it asap.

Finally, consider this:

It is not generally worthwhile whipping if it will kill workers on tiles which are 2-2-0, 1-3-0, 0-5-0 or better (remember to consider health). Once you get State Property it is no longer generally worth whipping since a state Property workshop always tends to have at least these yields. Thus whipping does become obsolete.


In a city like a fishless-fishing village which grows slowly, whipping will basically permanently stunt the income from that city. In such a city it is often best to whip in the essential infrastructure ASAP then allow it to grow unmolested to happy cap, at which point whipping can resume, or it can be left stagnant with priest/engineer. There are however exceptions to this which will hopefully be covered in a case study.

[Blake]

This post is for links to previous relevant topics

Everything you wanted to know about pop-rush quirks, and far more.

This thread contains the information on the "Whipping Bug" which is a matter of some controversy. The Whipping Bug means that you can get an unexpected number of hammers per population, for example with a +25% production bonus you can only get 30 or 60 hammers per pop, instead of the expected 37. This can be systematically exploited and the details are outlined in the thread.

The bug is fixed in the Warlords expansion and an unofficial fix also exists for CIV 1.61.
Although this fix is not approved for competitive gaming, discussion will generally assume that the whipping bug does not exist.

A case study on the utility of whipping
This thread contains a number of case studies on whipping - this time whipping Settlers under a variety of settings. Although it's an old thread it is all still relevant and gives compelling examples on why whipping is a very powerful tool.

Micromanagement is alive and well in Civ 4!
In this thread at Civfanatics you get a another perspective on whipping (under the heading "Whip until your hand's bleed"), especially for higher difficulties and embracing the whipping bug, also ideas on the optimal time to whip (these are somewhat contested).

[Blake]

Case study time:

Originally posted by couerdelion
Consider a size 2 city with Cattle 4/2/0 and Copper 2/4/0. It has half-built a granary and should whip this -unless constrained by happiness limits. Although this will lose the use of the copper tile for up to 9 turns (Epic speed), the gain from the granary (around +80% on food surplus) is to huge for this not to be taken now. In general it is a question of what the whipped unit/building will be doing that determines whether or not it is worth whipping. Workers and settlers are also good examples of "high-return" builds.

Now this is an interesting question. I don't automatically agree with the premise that it's pretty much always worth whipping a granary, I mean a 2-4 tile is pretty darn good, my intuition isn't sure which way to go. So this definitely calls for a formal examination.

The best test bed is always the game itself, and thus I create 4 identical cities, but in each one I use a different strategy to build the Granary. After the Granary I queue up something to measure the surplus hammers generated, for tiles above size 3 I use 2-2-0 tiles, which are simialler in value to a grassland cottage, but produce their output as hammers which can be stored per city.

After a number of turns I end the simulation and we can compare where they ended up:



1st Place
Whip! In the Whip city I used the strategy of maximize food and whip the granary as soon as possible. This city generated +10 more hammers than the next best.

2nd Place
No Whip! I simply let the Granary build as god intended, working the cows first. It did generate +2 more food than whipping and it is worth that at emperor this city would not be unhappy at size 3, while the whipped city would be. As such under emperor, no +happy conditions, it would probably be right to not whip. So couerdelion was exactly right that it's worth whipping if not constrained by happiness.

3rd Place
Optimize! I tried to time the granary reaching 30 hammers and the city reaching size 2. Clearly this optimization was a failure and it's extremely hard to go wrong with simply maximizing food.

Last Place
Iron First! In this city I worked the Iron first, such that the Granary completed slightly before the city grew. As you can see it is fully 20 hammers and 8 food behind the Cows first + Whip city. So while the logic of "I want Granary! I max hammers!" might sound good, it just doesn't work out that way in game.


Conclusion? The conclusion is that you really do want to get those Granaries out ASAP, pretty much regardless of the perceived opportunity cost of whipping.

[Cort Haus]

Blake

[Cort Haus]

Interesting that Iron First was the worst option there. However, in low-food cities, might it be different?

For example, a size 1 city with max food surplus of +2 (at 1 hpt), but with a 0/5/0 tile, can build a granary in 10 turns (normal speed) at zero growth. The 20 food lost during the max-hammer build will be recovered 20 turns after the granary completes (@ 2fpt).

To poprush the same granary will take 30 turns @ max food and 1hpt, or a bit less if it gets to size four in that time and two pop can be whipped.

At this point my calculations grind to a halt because I can't remember how much food is needed for each pop level, but the hammer first route looks much more competetive in this low-food scenario than with a 4F tile available.

[fed1943]

I agree, as long "granary ASAP" means built before the next grow.

Not necessary to sacrifice more to do it much before said timing.

Best regards,

[Cort Haus]

OK, low-food case-study :

Scenario is as in my post above - max food surplus is 2, but a 0f5h tile is available.

After 50 turns of max growth --> poprush :

[Cort Haus]

After 50 turns, max hammer first :

[allybhoy]

Blake, could you run the sim on another 10 turns....just interested

[Cort Haus]

A more realistic model would have been a fishing village in tundra with an Iron hill available, but that would have needed a lighthouse factoring in. (This raises another question for - granary or lighthouse first for weak coastal cities?)

The point is though, that max growth --> poprush is not always better than max hammer if available hammers are high and food is low.

[couerdelion]

Low food cities are those that actually benefit least from the granary. This might sound counter-intuitive to some but the granary really only applies a multiplier to the surplus food in the city.

Think of it like a library which you would not probably build in a low commerce city.

As for the question of coastal city with low production, I think I would probably use the iron to build the lighthouses and then decide which way to go.

[Blake]

I'm not going to agree that low food cities benefit less from a granary. They need it to grow to caps faster.
Altough if your caps are rather static (due to lack of ways to raise happiness) then low food cities can easily get away with just working mines.

Each pop costs 20 + Sizex2 pop to grow, halved with a granary.

A granary costs 60 hammers, which could be equated to ~33 food.

It could be put like this - say you have a city with an okay food surplus, like rice (+4 food) and some grassland hill. Your happy cap is stuck at 4, you could grow to size 4 and then whip the granary... or you could grow to size 4 and then soak up all the food using mines. I might do a test along these lines - like a race to size 5 and seeing what else can be produced in the meantime.

Anyway getting back to the Granary, in short the value of a granary depends on how much growing you plan to do, and of course re-growing also counts..

It is safe to say that EVERY city needs a granary (the only exception is like silly desert cities whcih are like 1 oasis, 1 desert iron mine and desert hill...) because all cities grow, and growing from size 3->6 will easily cover the cost of a granary. The question is really WHEN to build it. I really think the answer to that is "Soon after researching Pottery" at least if you want to get any more settlers/workers. It is however justified to delay pottery if your city has some nice mines to work... especially if you want to build a wonder. But it's still usually better to work grassland cottages and whip rather than working mines.

Another case where it makes sense to not build a granary is in OCC where you have unreasonably much food, like quad seafood, the city can cap so quickly that the granary isn't really needed, unless it will provide needed health. Of course OCC is special because you aren't getting any settlers.

Rubbish ice city: Lighthouse & Granary order

This one is easy. Assume the city only has coast to work (not even hill mines). You build whichever is cheaper first, or otherwise you build the Lighthouse first. The Granary can be said to double food surplus (only after a while).

City with neither:
Size 1: +1 food
Size 2: 0 food.

City with Lighthouse:
Size 1: +2 food.
Size 2: +2 food.

City with Granary:
Size 1: +2 food.
Size 2: 0 food.

Since this fishing village is going to spend at least some time at size 2, it's better to have the +2 food ALL the time than +2 only at size 1. Thus lighthouse is easily better.

If you have a mined ice hill to work (+3h) then I believe you should build the lighthouse first, you could then grow to size 2 and maximize hammers until the granary can be whipped. I'll do a test on this though.

In fishing villages with seafood I think the order goes like this:
1) Workboat #1
2) Cheap Granary (Expansionist)
3) Workboat #2
4) Cheap Lighthouse (Organized)
5) Expensive Granary
6) Workboat #3
7) Expensive Lighthouse

[Dog of Justice]

Optimize! I tried to time the granary reaching 30 hammers and the city reaching size 2. Clearly this optimization was a failure and it's extremely hard to go wrong with simply maximizing food.

Um, that's not how you optimize. A freshly built Granary in a size n city will double the first (10+n) food (i.e. half the food bar) you produce at the current population level, and do absolutely nothing for any additional food. Thus, when optimizing, you want about half the food bar filled when your Granary is completed. (If you have significantly more than half the food bar filled when the Granary is completed, you may want to go out of your way to create a large food overflow when your population grows, because every unit of overflow food will be worth 2 food.)

Incidentally, a fairer design would have the freshly built Granary provide 1 food for every 2 food produced, up to the (10+n) cap. But this isn't a big deal.

[Verrucosus]

Oh, more Whipping Math with Blake!

Thank you for putting so much effort in researching and explaining this. I'll try to digest all of this in my next game.

[uberfish]

Nice post Blake, in my experience that order for fishing cities is precisely correct (if you can't send a fishing boat from an existing city.)