Whipping vs working hammer tiles

Elandal

King
Joined
Oct 24, 2006
Messages
870
Location
Finland
While I'm not allergic to the whip, I'm a bit averse at whipping when I actually have hammer tiles to work (and to stagnate on). Even if I have excess high food tiles that thus aren't worked. My reasoning relies mainly on the idea that whip gives 3hpt / pop and if you use whipping cycle where you whip as soon as previous unhappiness has been forgotten then you effectively work the city at -1 happy cap.

I usually whip in the very early game (I improve food resources first for any city to start growing, only then other tiles) and in cities that don't have high hammer tiles at all. But I really don't think whipping pop out from working the mines is a good idea.

I read VoU's "Vocum Sineratio: The Whip" (is that latin and does it mean something?) again, but didn't go searching for the old discussions regarding whip efficiency. I didn't see any such calculations that would clearly highlight whipping.

So, I was thinking of coming up with some examples to clarify when working hammers is better and when whipping is better. This does not consider the strategic issue of whip-hammers being available immediately in one clump - that is generally needed in emergencies only.

The examples should consider hammers only as commerce isn't the goal where whip is used. That means that discarding differences in eg. working coastal fish (2 coins) vs. riverside mine (1 coin) is done - the tiles are considered for their FP value only, not C value.

I use epic speed, so when calculating anything I end up using 15 turn whip cycle and 45 hammers per pop. These should be 10 turns and 30 hammers on normal speed. Also, city growth on epic speed is 30 + 3/pop food - 33/36/39/42 to grow to sizes 2/3/4/5. On normal speed this should be 20+2/pop for 22/24/26/28. All values should scale so that if whipping is more efficient in some case on epic speed, it's the same on in same situation on normal speed (and the other way around too).

Consideration for multipliers can be given where appropriate. It may be that working hammer tiles yields suboptimal values in some cases - eg. assuming forge OR organized religion it's best to get hammers in fours.

Example #1:

Early game, no happiness sources available, thus happy cap 4 (and enough health). City has both high food tiles and high hammer tiles - best of both kinds so fish/pig/corn (6F0P) and mined plains hill (0F4P). City tile is normal 2F1P tile.

As the food tiles are higher FP total and whipping ends up using them more, it does tip the scales a bit for whipping. But that's how things are - having higher than four hammers per tile from more than one tile early game is rare (quarried stone on plains hill, plains copper mine, Tin event in BTS are examples of >4P tiles) where food resources are known immediately.

Working hammers at size 4 means working one 6F tile and 3 4P tiles for 13hpt for total of 15 * 13 = 195 per cycle.

Whipping 2pop every 15 turns without granary, growing ASAP, working hammers while waiting for whip unhappiness to be forgotten yields:
- 90 hammers from whip
- working 2x 6F tile for 4 turns grows us to size 3 with 4 food in granary so 35 more needed. As our "stagnation" means working 2x 4P tile and 1x 6H tile, it yields food surplus of 2. 11 turns of this gives 22, or 13 below required. 13 means working two more 6F tiles for one turn, one 6F tile for one turn (this means surplus of 5 in the end). We get to work one 4P tile for 10 turns, another for 9 turns. City tile obviously gives a hammer each turn. Thus, 15 hammers from city tile, 40 from one mine, 36 from another
Total comes to 90 + 15 + 40 + 36 = 181 hammers per cycle.

Conclusion: working hammer tiles yields 14 hammers more per 15 turn cycle. Whipping is not efficient enough, but difference is small. Whipping for 2 pop is around the same value as working the mines.


Similar examples would be welcome, highlighting different happy caps (for larger whips), different food tiles, different hammer tiles. The goal of the exercise is to understand what conditions make whipping clearly superior, clearly inferior, or very similar to working hammer tiles.
 
I started a thread with a similar goal in mind, inspired by a very good post by Roland Johansen.

EDIT: oh, btw:
Vocum = voice
Sine = without (sans)
Ratio = reason
VoiceOfUnreason ;) Not sure if it's proper latin or not, though, as i haven't touched latin for something like 8 years ^^
 
Factor in a granary and the difference become quite big though... Also if you have a higher happiness cap it is obviously quite different...
 
@Percy
Read that thread. Good material, but again exactly one example :) Admittedly my ex#1 (and ex#2) are constructed and not necessarily the most common ones. However, they require only 2x 6F and 3x 4P tiles, which is possible with eg. corn + pig + 3x mined plains hill - not too far from reality?


Higher happy cap allows for larger whips, and requires more food to grow back. Larger whips are I believe the more important factor.

Granary lowers the amount of food required to grow back. The minimum loss is the cycle length times the hammers from last citizen, which in ex#1 would be 4hpt x 15 = 60 hammers vs. 90 gained from the whip - whipping 2pop in that example has gain upper bound of 30 hammers (2hpt, also calculated from 4hpt gained by last citizen whipped vs. 2pop whip times 3hpt per pop). Three pop whip where last pop is working 4hpt tile again has upper bound of 75 hammer (5hpt) gain.

IIRC granary preserves half the grain required to grow rounded down, and food above the growth requirement is added as usual. That'd mean that if growth requirement is 39, granary preserves 19 after growth and if the food addition would've brought the total to 42, then end result is rounddown((39)/2)+(42-39) = 19 + 3 = 22.

Example #2:

Calculating situation identical to ex#1 except that the city has a granary. Thus working hammers produces identical result, but whipping is more efficient.

Assuming growth to size 4 in the beginning of the cycle happens so that no food is added above what is required to grow, granary will have 19 food after growth. Whipping two pop, we need 36 - 19 = 17 food to grow to size 3. Working 2x 6F tile for 2 turns yields 20 food, so 36 +1. Growing to size 3 preserves 18 + 1 = 19, requiring 20 food to grow to size 4. This is exactly the gain from 10 turns of working 6F tile + 2x 4P tile. We need to get a three turn set with net food of 0 to compensate for this. This is achieved by working 6F + 2x 4P tiles (F+2) for two turns, then 3x 4P tiles for one turn (F-4). Result is that we work one mine for 2 turns, and two for 13 turns, so we get (13 + 13 + 2) * 4 = 28 * 4 = 112 hammers from worked tiles, 15 from city tile, 90 from the whip totaling 217 hammers per 15 turns (14.5 hpt). This is 36 more than without granary (ex#1) and 22 more than working the tiles.

Conclusion: with granary whipping yields 22 more hammers per 15 turn cycle (1.5hpt) than working hammer tiles. Whipping is slightly more efficient than hammer tiles.
 
It is always inefficient to kill a citizen working a mined grassland hill.

At size 6, it becomes inefficient to kill off a mined plains hill.
At size 6, it becomes inefficient to kill off a plains forest.

At size 10, it becomes inefficient to kill off a mined desert hill.

At size 20, it becomes inefficient to kill off an engineer.
 
In my last game I probably built too few workers, but I still think it gives an example of how the whip can be worthwhile. I whipped granaries in a bunch of cities, and then started whipping units and new buildings. I kept my population small, which lowered the requirement for food to regrow. The other advantages in my game were that the small cities didn't run into war weariness problems and I didn't get that slight raise in upkeep you get when you raise your city's population.

Often my best tiles remaining to work were (for example) grass/hill/forest, grass/flatland/river, or grass/flatland/forest. In this situation, whipping off the workers who were maybe just feeding themselves plus producing one :production:, whipping down from 4 to 2, was quite worthwhile. In cities with more really-useful tiles I didn't whip so much.
 
Thanks Dave - exactly the information I wanted.
 
I do not follow Elandal's calculations, but I assume he is right. The statements of DaveMcW are easier to understand. But is money taken into consideration here? Let us assume we whipped a -2 population; a mined plains hill and a riverside cottage (1 plus gold, and yet another one if you are financial). How much money would we loose?

DaveMcW writes: “At size 20, it becomes inefficient to kill off an engineer.” Does this mean all great people are burdens to our productivity if the cities are less than size 20?
 
DaveMcW writes: “At size 20, it becomes inefficient to kill off an engineer.” Does this mean all great people are burdens to our productivity if the cities are less than size 20?

The value of hammers relative to culture, gold, or beakers is hard to place.

What it does mean is that if you ignore great people points then you're better off whipping than using engineers. The same will typically be true of other specialists when you don't make use of the great people points.
 
This does not consider the strategic issue of whip-hammers being available immediately in one clump

In case anyone missed it. An important qualification.

that is generally needed in emergencies only.

Not sure I agree. I can think of quite a few non-emergency situations where getting something built quicker is worth a certain loss in terms of total hammer/food output.

Quicker/earlier wars, for example. Or settlers to their targets earlier (likewise workers or missionaries). Or earlier specialists, border pops or NWs.

Still, an interesting discussion. :goodjob:
 
The answer is really, it depends. Most dominant factor is how many extra food does that city produce, If its a very high food city like having 3 food ressource. I find it more efficient to do large whips. For example if happy cap is 5, city got up to 6 pop with 1 unhappy citizen, it is inefficient to just whip 1 pop, but if you do a large whip and whip 3 pop, it is worth it.
 
What most on analisys is lacking is taking in account workers turns.
People try to evaluate static situation whe it is essentially dynamic.
 
Sacrificial altar makes whipping of very high food cities extremely efficient choice. If food allows for growth in that half normal time, it's 6hpt per pop and thus nothing can really compete with it.


@Mutineer
How do worker turns enter this? A city that is whipped needs fewer improvements than a city not being whipped?
 
Maybe he means that getting a worker through whipping is much better than getting it through working tiles, because of the improvements he can do sooner. That would be an example of a non-emergency situation where getting a bunch of hammers at once is better than getting them slowly.
 
Good point. In the first post I specifically noted that I don't consider other than hammer counts, but clearly I should. I do that in games anyway.
 
It is always inefficient to kill a citizen working a mined grassland hill.

snip

Dave,

Are these numbers all based on the amount of food necessary to regrow the lost population? If so it would also be a function of the amount of population being enslaved, so are you assuming a 2 pop whip?

Thanks,
Darrell
 
How does the math work out for settlers / workers? Whipping vs using high food reasources directly to produce them?
 
For those wondering about my math:

The whip converts (10 + city_size) food into 30 hammers.
A grassland hill converts 1 food into 3 hammers.
A plains hill converts 2 food into 4 hammers.
A plains forest converts 1 food into 2 hammers.
etc...

If you are planning to whip multiple citizens, use the average city size during regrowth. I assume a granary in all these calculations.

But as others pointed out, this is more of a guideline than an exact calculation. Very often the benefit of building something instantly outweighs the inefficiency in the food/hammer conversion.
 
Top Bottom