Cottages!!

IMO those assumptions are so far from real-game that it's hard to say if the approach is truly better. Worker turns aren't infinite in any practical game scenario that matters. Food surplus = happy surplus is a pretty good "rule of thumb", generally.
You have to make some kind of assumption... even at the "rule of thumb" Food = Happy you have to farm and cottage a lot... Both get the "best" out of the city and needs x number of worker turns.

In a city with a happy cap of 15, like in this sample both Dave and I advocate growing ASAP, this takes the first 34-ish turns so lets ignore the first 30 turns where we both farm the heck out of this baby whip the granary and grow grow grow.
After that the strats start to differ a little.

But first... lets start by admitting there was a little miscalculation in my spreadsheet. To calculate all this I turned the cottage growth up side down.
On turn 130 Cottage on turn 0 Turn 120 Hamlett etc... This while turn 130 itself never actually got counted. As this affects both calculations exactly the same (i.e. commerce for turn 130 not counted)

Now 1 picture paints a thousand words. See below graph.

Because Dave starts cottaging earlier than I do, he builds up a head start which at its worst/best is 45 commerce on turn 55. Because Dave starts working them earlier they mature earlier, thus you get the sea saw going forward and back a bit in total commerce.

Because at turn 45 I start working the cottages vs Dave's turn 35.
Turn 45 I work 7 vs Dave's 7
Turn 46 I work 13 vs Dave's 7
And from that point on I work either more or eventually at size 15 equal number of cottages.
Turn 76 is the last turn Dave is in a small advantage situation, after that I start to run away with the total commerce.

I am not a mathmatician, so I cannot put a formula to it... But for (atleast) this case the formula as stated by Dave Food surplus = Happy surplus is NOT the best solution
 

Attachments

  • CottagingGraph.jpg
    CottagingGraph.jpg
    24.6 KB · Views: 3,472
Another intresting thing is that when things start to matter... Mature towns... I work more villages and more towns faster...



That means if you extend the timeline my version though more worker intensive (more farms and more farm > cottage conversions) will yield even more commerce than the "Food = Happy" rule.

Obviously the area's where Dave gains ground are where his cottages mature while mine are still the lower category.
 

Attachments

  • VillagesTowns.jpg
    VillagesTowns.jpg
    38.9 KB · Views: 3,431
To explain namliaM's example: The message I remembered from the original thread on "How to get a new commerce city up to speed best" was "Farm everything until Basic food surplus = happiness surplus", then cottage everything. Here Basic Food surplus is the surplus without any farms at all. In his new example (without rice) that would be 2 (from the city square), so optimal would be "farm until until size 13, then cottage".

Missed NamliaM's bigger explanation above before posting (it was on page 2), and didn't actually do any new calculations, but if it is not too much work I'd be interested to see if he can simulate my proposed course of action as well.

PS Nice article Dave.
 
@ Oyzar,
Sure I can do the calcs for Farm to size 13 and cottage...

@Dave and "infinate workers"
Presuming you have the original farm in place (advanced farming inside the culture of another city?!
2 workers can farm enough and pre-cottage every farm untill such time you need to replace them. The trick comes where you switch from all farms to all cottages where you need 3 more workers to transform the farms to cottages. 2 workers isnt quite as infinite is it?? Though one more than "average", then again one of your rules is "when working inimproved tiles, whip or build a new worker" so 2 workers shouldnt be that bad...

Though my approach would require 115 worker turns vs your 81 (+34 turns or +42%) it is gradual enough... In fact your approach is earlier a little more worker intensive due to switching farms to Cottages earlier. The only place where 2 workers dont work for my approach is the (mass) swap from farms to Cottages


Again I am ignoring the first thirty turns because farms all the way and whip granary is best.

Fact remains that if we have 2 such cities you require 162 worker turns vs my 230 if all that work is required to be done inside of 80 turns. You can make due with 2 workers where I would need 3 workers. Say the worker costs 1 GPT but I gain 180 commerce vs the 80 gpt cost. And offcourse the 60 hammer cost for the worker. It remains very iffy if it is actually better... but... raw commerce it is...
 

Attachments

  • WorkerTurns.jpg
    WorkerTurns.jpg
    30.2 KB · Views: 3,416
^^ One huge factor to remember is that one commerce :commerce: is worth more early than late game.

10 :science: at turn 30 are worth more than 30 :science: on turn 90.

So although mathematically you may gain a few commerce in the long run to (farm all) and then (cottage all), in a game situation I prefer to cottage early and get the commerce when it is worth the most.
 
^^ One huge factor to remember is that one commerce :commerce: is worth more early than late game.

10 :science: at turn 30 are worth more than 30 :science: on turn 90.

Aha, this is what we thought on the linked thread too... now... there is a BIG difference between cottaging all the way and Farming a lot...

I made 4 different options from worse to best
1) Cottage all the way
2) Farm and whip the granary then switch to all cottages
3) Dave's "Food = Happy"
4) My Mathmatic maximum
And calculated the differences

A comparison of each vs the Cottage all the way.

At the lowest point I am at -133 commerce at turn 45, at which point Granary whip is at -81 and dave is at -94 all compared to Cottage all the way.
Yes at this point you do have ~100 commerce head start... in 45 turns = ~2 commerce / turn.

Notice though that all versions of "not cottaging all the way" catch up FAST.
Cottaging right after the Granary gets Break even on turn 74, I catch you turn 60 and Dave catches you by turn 58

By turn 80 Dave and I are fighting for the honors both at about +380 commerce over the Cottage all the way and about 350 over the Cottage after the Granary result, which is "stuck" at around +30 commerce

By turn 80 it is -on average- +4 commerce per turn (over the 80 turns) for Dave and me over the Cottage all the way solution. Now one commerce this turn = more commerce next turn that may be true... but at
1053 vs 1081 vs 1422 vs 1442 total commerce on turn 80 or
13.16 vs 13.51 vs 17.78 vs 18.02 commerce per turn
That is almost +5 commerce or +36.94%

The difference simply explodes from there on out... And I do mean EXPLODES.

Same graph but extended beyond the first. At this point 130 turns into the future, while you enjoyed your 100 commerce lead... I now am stretching ahead by almost 1500. Again 1 vs 2 vs 3 vs 4, worst to best.
2977 vs 3257 vs 4331 vs 4422 total commerce after turn 130 or
22.90 vs 25.05 vs 33.32 vs 34.01 commerce per turn.

That is +48.54% worst compared to best even the inner two, Whip Granary and cottage vs Dave's 'limited' farming strategy gets a difference of 8 commerce/turn or +32.98%

I will grant you that the 133 commerce I am behind cottage all the way is a lot, particularly given that I am only making 10 commerce at that time (13 turns, assuming the rest of the empire doesnt generate any commerce.)
However turn 45 I am working 13 cottages vs 2 villages a hamlett and a Cottage on the first and start to catch up REAL fast.
By turn 60 (in just 15 turns) I have cought up making 37 commerce/turn vs only 19 for cottages all the way.
At turn 100 it is 54 vs 36, that is +50%

Now tell me again how much is that early 133 extra commerce worth compared to the +1445 after 130 turns (that is 30 turns delay for the Cottage all the way version, again assuming the rest of your empire doesnt produce any commerce)
 

Attachments

  • CottageComparison.jpg
    CottageComparison.jpg
    33.3 KB · Views: 3,329
  • CottageComparison2.jpg
    CottageComparison2.jpg
    29.8 KB · Views: 3,371
I will conceed (to Dave, but I think I did that already) that his way is way more worker friendly and probably way easier to do (less MM)... than having to pool all my workers together on those 2 turns. A more gradual change from Farms to Cottages will probably make things a lot easier. Though I think the Food=Happy can be improved upon by taking the middle road of our two ways.

Now the mathmatical way gets 4422 commerce vs Daves 4331.
Using 2 workers, assuming we have 1 farm waiting we can build enough farms and cottages, also switch 2 farms to cottages every turn.

Using the 2 workers as a premise.. i.e. switch 2 farms to a cottage every turn ...
Farm everything to size ... + ... turns then add + ... cottage, get ... commerce total.
13 +0 +2 4355
12 +1 +1 4367
12 +1 +2 4375
12 +0 +1 4387 <= Grow to size 12 add a cottage tile, start replacing the turn after.
12 +0 +2 4390 <= Grow to size 12 add a cottage tile and change 1 farm to Cottage
12 +0 +3 4389 <= Grow to size 12 add a cottage tile and change 2 farms to Cottages
11 +0 +1 4392 <= Grow to size 11 add a cottage tile, start replacing the turn after.
11 +0 +2 4390 <= Grow to size 11 add a cottage tile and change 1 farm to Cottage
11 +0 +3 4382 <= Grow to size 11 add a cottage tileand change 1 farm to Cottage
10 +1 +1 4375
10 +1 +2 4372
10 +0 +1 4361
10 +0 +2 4346

As you can see it drops "rather quickly" and with only 2 workers beeing the optimum somethng of 4392 out of the Mathmatical 4422 for 30 commerce saving you the headache of flashing in 4 workers to flash all the farms to Cottages making MM much easier... I think I will opt for the 2 worker option all the way. Then again in this case taking the hit for -0.7 commerce (see my previous post) per turn on average (-2.03%) to the Food = Happy "rule of thumb" is not that big.

Now it is up to the Mathmaticians to come up with a formula that says something like "Farm up to size 11 or 12 then start working cottages" -in this case-
I am allmost sure it will change if we re-add the Rice.
 
So this city has 5 plains tiles, but it seems like a good cottage city to me, since the food surplus from food resources exceeds the -1 from each plains tile. What do you think?

 

Attachments

  • Civ4ScreenShot0022.JPG
    Civ4ScreenShot0022.JPG
    138.3 KB · Views: 3,345
@Tephros: First of all you should get some more sleep (or correct your computer's clock ;)). Secondly, this is real early in the game, what is your happy cap? You don't want to work those plain tiles for a while. In particular, working fish, horses cow, deer, and several grassland cottages, you simply have not enough people to work those plains tiles (even then specialists are probably better (for the GPP)).

Moreover as you are financial you are allowed to work the river plains tiles according to the guide, which leaves one plains tile, so what is the problem? (PS I definitely don't see this becoming a production powerhouse soon, it seems it could only be a mediocre GP farm so I'd cottage it up as well, just leave the plains for a while.
 
That city is actually -3 per Dave's guide. You get +1 for the livestock and +2 for deer, but -1 for tundra and -5 for plains. Actually, that city's food surplus is suspect in general.

IMO, if you wanted :hammers:, 2W or 3W would have been superior. If you wanted commerce, fitting those flood plains in one way or another would have been a better move.
 
That city is actually -3 per Dave's guide. You get +1 for the livestock and +2 for deer, but -1 for tundra and -5 for plains. Actually, that city's food surplus is suspect in general.

IMO, if you wanted :hammers:, 2W or 3W would have been superior. If you wanted commerce, fitting those flood plains in one way or another would have been a better move.

I ignored the tundra, but you ignored the fish. -5 for plains, -1 for tundra, +2 for deer, +1 for cow, +4 for fish. So it's +1 and the city could reach 20 pop without any farms at all. The fish will drive early growth quite nicely, and two production resources without compromising growth. But I should put a farm somewhere so that the wheat resource to the east can be irrigated eventually. This means I can work at least 9 tiles as cottages with a further food surplus.

This was my second city, so I wanted horses asap to stop building warriors, and partly to establish trade routes with chariots, especially to get a religion. I had intended from the beginning to make this a cottage city that has okay production, to minimize pop rushing. A decent city can still be established 5W or 6W, though you can't see that whole area. The area to the East is seriously production poor either way.

@Tephros: First of all you should get some more sleep (or correct your computer's clock ;)). Secondly, this is real early in the game, what is your happy cap? You don't want to work those plain tiles for a while. In particular, working fish, horses cow, deer, and several grassland cottages, you simply have not enough people to work those plains tiles (even then specialists are probably better (for the GPP)).

Moreover as you are financial you are allowed to work the river plains tiles according to the guide, which leaves one plains tile, so what is the problem? (PS I definitely don't see this becoming a production powerhouse soon, it seems it could only be a mediocre GP farm so I'd cottage it up as well, just leave the plains for a while.

I was at work! LOL :)

As far as happy cap, I'm planning on capturing the pyramids from the Carthaginians, as they built it a bit after this screenshot. I have only one happiness resource and will soon be able to trade for a second and possibly 3rd one, and as I mentioned am working on getting a religion. It's on immortal so it's going to be a challenging game.

Also I do have iron, which I got with a 3rd city to the south. So it's not a lost cause.
 
Namliam - thanks for the graphs and discussions. Your graph is not easy to understand - could you please clarify what the two axes are? Does it refer to the (difference in cumulative commerce) in a single all-grassland city managed in three different ways, compared to all-cottages?

It would be simpler to just plot cumulative commerce of the FOUR options, not differences with three... :)
 
NamliaM - my main point is that your example is not conducted in a "game" context.

In a real game, extra commerce does not just sit still getting added up in a graph.

Early extra commerce means early extra beakers, which means early techs that help you boost things further. Those early beakers are leveraged into many times their initial value by allowing earlier multiplier buildings, earlier tile improvements, etc etc.

If (cottage all the way) gets you to - for example - earlier mathematics (50% chop power) or earlier libraries, you can further leverage these into an increased advantage over your slower rivals. This advantage cannot be measured in isolated terms of commerce-per-city.

Your idea may still be correct. But your test is not the right one. I would want to test four Civs on the same land, following the same tech tree routes, building the same improvements, racing for the same wonders, before I necessarily agree. :)
 
The graphs are turns on the bottom, total commerce difference compared to "cottage all the way"

My point was to make the difference clear(er) offcourse making total commerce would be possible for all 4, but the graph gets "disfugured" because the top values of Dave and me get considerably higher 4400 vs the now 1400, while the "bottom" goes up to 3000-ish this will make variations harder to spot.

Since in this sample we can farm any tile we want and build cottages... We atleast have CS, as an extension Writing, Math are also know. Thus we will have other cities producing commerce as well. Also we have a pop-cap of 15 both health and happy. We are into the game already some turns.
Yes the -100 commerce hurts at turn 45. However by turn 60 there is no difference and by turn 75 there is a gaping gap of already +250 commerce for dave and me vs the "cottage as soon as possible"-variant(s)

There are other considerations to make in the early game, among which, but not exclusively
1) Lower caps, everywhere
2) Settlers/Workers to build

Early game when you for a while have a happy cap at 6/7 you just want to grow to what pop 5 and work cottages beyond that. As cap rizes, your cottage cities (slowly) grow with the cap.
The problem with this situation is it is mostly situational anyway because of the other considerations to make... mostly the Settlers/Workers or maybe whip a library which in this case with the 15 happy cap would still be better at 70% science.

@TMIT, Tephros and Celebithil
That city would have been MUCH better IMHO one east getting the fish in its initial 9 and picking up the FP.
How ever intresting that discussion may be, I would like -personaly- not to make this thread about "how do I develop this city best" but try and get a "rule of thumb" kind of thread where we can find out how to develop commerce cities best.
 
The graphs are turns on the bottom, total commerce difference compared to "cottage all the way"

My point was to make the difference clear(er) offcourse making total commerce would be possible for all 4, but the graph gets "disfugured" because the top values of Dave and me get considerably higher 4400 vs the now 1400, while the "bottom" goes up to 3000-ish this will make variations harder to spot.

Since in this sample we can farm any tile we want and build cottages... We atleast have CS, as an extension Writing, Math are also know. Thus we will have other cities producing commerce as well. Also we have a pop-cap of 15 both health and happy. We are into the game already some turns.
Yes the -100 commerce hurts at turn 45. However by turn 60 there is no difference and by turn 75 there is a gaping gap of already +250 commerce for dave and me vs the "cottage as soon as possible"-variant(s)

There are other considerations to make in the early game, among which, but not exclusively
1) Lower caps, everywhere
2) Settlers/Workers to build

Early game when you for a while have a happy cap at 6/7 you just want to grow to what pop 5 and work cottages beyond that. As cap rizes, your cottage cities (slowly) grow with the cap.
The problem with this situation is it is mostly situational anyway because of the other considerations to make... mostly the Settlers/Workers or maybe whip a library which in this case with the 15 happy cap would still be better at 70% science.

@TMIT, Tephros and Celebithil
That city would have been MUCH better IMHO one east getting the fish in its initial 9 and picking up the FP.
How ever intresting that discussion may be, I would like -personaly- not to make this thread about "how do I develop this city best" but try and get a "rule of thumb" kind of thread where we can find out how to develop commerce cities best.

I am convinced that you are correct... in an abstract situation featuring limitless workers, a 24-grassland-tile river city, mid-game techs, and no opponents.

I'm not necessarily convinced that you are correct in a real game :D . I could still be convinced, but not by the example you have given. I do appreciate your hard work with the graphs though! :goodjob:
 
My example is not that abstract ... I find a lot of cities in general that have >10 grasslands (river or otherwize), and regardless... Lots of grasslands = Cottage city. Lots of plains... cannot cottage, so you need Food resource + Grass or FPs.

"Infinate" workers... Not true... 2 workers in my last example, that is two workers not infinate... It does however require 1 farm to be present and workable at turn 0. But this is true for the other examples too, having a cottage or farm availabe at turn 0.
Not having it available may 'push' things towards the cottage all a litte. With two workers you can build a cottage in 2 turns, while a farm takes 3 turns before you can work it in the city. Doubt that one turn difference ... really makes a lot of difference.

100 commerce 'lost' the first 45 turns vs PLUS 250 commerce at turn 75 this is TWICE the difference you are complaining about losing in the first 45 turns. Thus invision getting PP a turn sooner or whatever a turn sooner (Optics for caravels or Lib to beat out the AI by only 1 turn)

The big problem in getting a "real" example is estimating the growth rate there are big bursts where growth becomes available.
To name a few HR, Religion (temple(s)), Calander, Currency (market), Metal Casting (Forge) and they are all situational depending on the land given, cities founded, hammers available and game goals.
So you have to make assumptions and have to make generalizations, based upon this example I would say farm and whip Granary keep farming untill nearing the Happy cap then cottage. I am pretty sure that whip Library for 2 or 3 pop is a good thing in this city too.

What kind of "parameters" would you consider a "fair" test case? You set them up, I will test these 3 (I consider Dave and I to be -nearly- the same) different scenarios or what ever other scenario you can think up.

I mean settling your second city when you dont even have pottery, at that point you cannot even build a granary. In which case farm and not whip but farm to grow and later cottage, when obviously you do have Pottery and you may even consider whipping the granary at that point.
Settling your first coastal city with 2 coastal tiles both contain fish? Workboats!
Need to raise happy? Build/Whip a forge/market/Colosseum/theatre/temple as applicable/available, what ever returns best results hammer to happy conversion.
A lot also depends on the bonus you get from the leader traits, the granary may be cheap or the Colosseum or Forge or etc...
Different UBs come into play, extra happy from colosseum, or +10% gold from a "Forge", food from a Aquaduct,etc
Even events can change things, Parrots or other pets (can remember them off hand) for +food or +commerce. Stables quest for +food.
It all makes it unpredictable and all need to be generalized...

So you propose the city and growth parameters, I run the numbers?
Alterations I want to test to the "all grass" city:
- Re-add the Rice
- Replace Rice by Pig
- Add second Pig
- add third Pig
Upping the food, more food may change the size you grow to before converting farms to cottages. Because this leaves more autonomic growth working the High food tiles without additional farms.
Offcourse all this then again is all generalized assuming say 2 workers, however this does dis-regard the need for further settlers/Workers. Thus you need to get them elsewhere in your (young?) empire. Any test/generalization will assume some techs available to an extend if only BW/Wheel/Pottery/Writing
 
The big problem in getting a "real" example is estimating the growth rate there are big bursts where growth becomes available.
To name a few HR, Religion (temple(s)), Calander, Currency (market), Metal Casting (Forge) and they are all situational depending on the land given, cities founded, hammers available and game goals.

I think your comments above capture my point very well... There is great value in getting to the "boost" technologies early. However I can see that this possibly comes at significant cost to later commerce, as your graphs display.

I suspect it is difficult to measure objectively the cost/benefits of each approach, because (as we both agree) it depends upon situation - one of the beauties of Civ IV! :D

An extreme example of the benefits of the all-cottage approach would be a game where getting to Mathematics two turns earlier (thus boosting chops by 50%) could make us beat the AI to a key wonder, with obvious huge flow-on effects throughout the game.

However if we are talking a regular non-capital cottage city being founded in the mid game (say, after all the jungle is gone and when the happy cap is high) then your approach is very likely much better.

So you have to make assumptions and have to make generalizations, based upon this example I would say farm and whip Granary keep farming untill nearing the Happy cap then cottage. I am pretty sure that whip Library for 2 or 3 pop is a good thing in this city too.

Your discussion has definitely made me re-think the pros and cons of farm vs cottage in the early stages of a cottage city, and I am about to go and start a farm-then-cottage game to put your approach into action in a very concerted way, to see how I like it :)

I can foresee that I will have to build a few more workers than usual, as each tile has to be improved "twice", and in order for the switch from farms to cottages to be efficient, it needs to be done quickly.

What kind of "parameters" would you consider a "fair" test case? You set them up, I will test these 3 (I consider Dave and I to be -nearly- the same) different scenarios or what ever other scenario you can think up.

As I said above, I'm not sure there is an objective test, as it is situational to a large degree.
Spoiler :
I would suppose that a test would be very complicated, and would involve playing the same game four times (once with each approach), up until a certain point (say, 50 turns after Civil Service). You would try to maximise game-total empire-wide commerce in each case via tech selection, buildings, etc.

To reduce variables you'd have to play isolated, no random events, no barbs, and settle your cities in the same place each time.

Even then I'm not sure if the answer is going to provide any more than the "best" way to play THAT EXACT map, rather than producing a general rule. :crazyeye:



One variable I would be interested in seeing if you are going to run more tests on your single-city map is the presence of floodplains. Floodplains (which I love! :D ) appear in many awesome cottage cities. From a gameplay viewpoint key differences (apart from +1 food obviously) are:

- they lower the health cap
- they are always on a river
- they take more worker-turns to improve, and
- farming them increases their food yield by only 33% (rather than by 50% in the case of grassland)

So the arguments for immediately cottaging floodplains are possibly stronger than for grasslands - especially if you only have a few workers.

It would be good to see how having 4 or 5 floodplains in your city would affect the yields.
 
Top Bottom