A pattern governing supply and demand changes

[duke o' york]

Originally posted by William Keenan Yes, occasionally, cities supply oil before the discovery of industrialization. I've seen it, we've all seen it. And it sux when it happens because nobody wants the stuff!

I wouldn't say that it sucks at all!

I'm sure that everyone has seen the preponderance of coal supplying cities in the early game, when the demand is quite low, but when you arrive in the industrial age then far fewer cities supply it and everywhere wants it.
I find it useful to build the coal caravans when you have the chance, and then store them until they are demanded later on. The actual trade bonus, from which you get all the beautiful techs and cash, is calculated based on the state of the two cities at the time of the trade and so you'll get a very large bonus indeed. Obviously you will have to weigh up the pros and cons of delivering the freight soon after it is produced, for a smaller but maybe more valuable bonus, or waiting until the cities are larger and you can choose between more, for which the bonus will be enormous. The length of time involved also means that you can transport the caravan to the other end of the map to get a bigger bonus, but try to avoid getting it killed!
Similarly with oil. Build the caravans while you can and then leave them until they are demanded after Auto. You can of course do the usual maximising of trade in both cities to get the largest bonus, and always have the chance to throw one into a vital wonder if it looks like you'll be pipped to it by the AI.

FYI, I'm not a newbie as such, although this is my first non-OT post here. I regularly visit to keep an eye on the valuable work done here, like in this thread, and then go back to CFC to claim it as my own. j/k

[La Fayette]

SG,
I bet you should have a look at the sig of the duke.
A man with such a sig cannot be so bad as he claims

[samson]

I may be close to cracking the supply/demand question, not quite ready to claim that yet, but I'm getting there. I have identified or confirmed a large number of factors that affect the supply/demand lists of cities. They are:

1) Terrain -- the most important factor, but also the most difficult to understand.
2) Resource specials
3) Rivers
4) Roads (?)
5) Location -- the actual X,Y coordinates.
6) City Size
7) Number of Techs acquired.
8) Specific Techs -- Pottery, Industrialization, NF, many others.
9) Region - proximity to the Poles, perhaps other effects.
10) Continent #

This is like an onion, layers within layers.

[Scouse Gits]

samson...terrain is a difficult puzzle. I thought that the initial commodities gained from any tile didn't change whether in game play mode or cheat mode.

However, there is a very temporary glitch. Open a trade start game and activate - Cheat - Reveal entire map - create unit (settler) - build city. Note the supply and demand items when you first see the city screen after founding. Then return to the game but immediately go back to the city screen. The commodities have changed. Perhaps you already know this, but it seems to me that the reveal map cheat momentarily fools the game that all squares have been explored but then it makes an adjustment when the count for revealed tiles is verified as very few.

Another variable could be the amount of exploration.

---------------------

SG(2)

[Grigor]

If exploration is a factor, the amount of exploration might very well be handled in the same way that Civ Size is computed in the Demographics. There could also perhaps be an influence from relative position on the civ size ranking.

I think that the anomalous way the computer handles hides (the idea that it might be assigned the index number of zero seems very attractive) could very well be the converse of the way it handles Uranium.

The compelling and elusive nature of this general problem is illustrated by the difficulty of devising testable hypotheses. This month promises to be as exciting for lurkers as last year's string of discoveries about the key civ and tech gifting. Keep it coming!!

[solo]

SG,

I also believe revealing the map will lead to some commodity changes in the supply and demand lists, when they change. I agree about exploration being a factor since it reveals more of the map. I put trading maps in as a factor, too, since this also reveals more of the map.

However, I think any changes brought on by cheat mode may be altogether different. I am now believing that any use of it distorts supply and demand, although I think using it does identify the factors that may be involved.

Samson,

I hope you are right about being close. I thought I was quite close to the cause of the gaps, but not any more.

With more trades my test game is showing two more changes in supply and demand lists to two cities that delivered demanded cargos to targets on different continents. In both cases, new changes appeared in the lists just a few turns AFTER 16-cycle changes had occurred. In both cases, hides were traded out of the cities' supply lists. Both cities were capitals, too, and this was the third change so far, observed for both. A lot of coincidences!

For one of them, activity has been limited to just 8 different commodities, with hides and beads jumping from one list to the other. Dye, wine and copper are the only others to appear on its demand lists.

For the other city, 9 commodities are involved with salt joining hides and beads as a commodity having appeared on both lists. Again dye, wine and copper have been the only others on its demand lists.

As more changes occur during my test game, I think a method to this madness may be revealed.

[Elephant]

Something wierd happened in one of my tests that sounds similar to things you have faced:

I am using a game SG created for his starting tech investigations where the English start with Trade. At one point in the 3000sBC I checked my capital city screeen first after moving and noticed a S/D change, which I logged. I then went to Trade Advisor to check the rest of the foreign capitals against the last turns log and noticed that my own capital showed up the same as before. I then went back to my capital city screen and the S/D was back to the same lists as the turn before. So some things can "change" with no other triggers, or the Trade Advisor is some sort of "fixer" to other changes.

Samson, in looking at your lists I note that some variables may be considered dependent on or extensions of others. Specials are extensions of terrain. Perhaps there is a "terrain#", with modifiers for specials, rivers, and roads, sort of like that long binary string that military units use for special abilities. X/Y location and Region are dependent. Number of techs and specific techs are dependent, or at least related. The specific techs may "activate" Demand and Supply separately, much as we suspect Oil can be Supplied early, but Demanded around the time of Auto. I was starting to fool with City# in addition to CitySize after some of Solo's findings, and I threw Civ# in as well but have not investigated that thoroughly.

I was also wondering if the specific tiles a citizen was working could be a factor, along with GameTurn, MapSize and DifficultyLevel. Another possibility is "number of revealed tiles" - that could explain how using Cheat to reveal the whole map could affect S/D, as does trading maps. Have not investigated any of those yet, just wondering what else the game could be tracking as trigger possibilities. I remember Xin Yu talking about unknown state variables he had noticed in the save files when he was debugging to create his City Manager utility. I wonder if some of those could be related to the variables we are looking for.

Just a few more thoughts to muddy the soup. Civ On.

[solo]

Elephant,

Questions:

Did the weird thing happen on the city's 16-cycle turn?

Did you try to repeat it using a reload?

I had this happen once when using the cheat menu while conducting earlier tests. The demand list of a city changed, but went back to its prior state after checking demand with the Trade Advisor. On a replay, I was able to deliver a demanded commodity that was new to the list, but it disappeared when checking the city screen right afterwards. If I checked the city before delivery, the commodity was no longer demanded when the delivery was made.

I agree with you that the variety and limitless combinations of factors now known to affect commodities chosen for these lists make it hard to isolate the cause(s) for each change that occurs!

[Elephant]

I have not mapped that round out for Solo-Cycles yet, so I don't know the answer to your first question. I was not saving every turn, so I am unable to try it again. I will try to do so in the future - it just takes more time, and I am playing these test rounds on "lunch break"...

[William Keenan]

There is a way to review supply and demand of all cities without going into cheat mode. The CivCity utility allows viewing and editing of supply and demand in a saved game. At the end of the turn, save the game, load it in CivCity and page though the cities.

[solo]

Elephant,

Testing does take time, especially when it seems no progress is being made! However, the infrequent discoveries are a rush and make it worth while.

William,

Thanks for the tip. I have downloaded the utility, which I found at CFC, but it wants a Visual Basic .DLL to run, which I do not believe I have, since I do not have VB.

One question. Are supplies and demands for cities shown before their civs have learned Trade? If so, I just might get a copy of VB in order to use civcity. Otherwise, I do not feel handicapped at all, using my hot seat approach for viewing all lists, since this does not require reloads every turn.

[samson]

Solo -

Re: 'gaps' -- The 16-turn cycle is a 'trigger' event not a cause of suppy/demand changes. Latent changes are made visible when a city's turn comes around. The explanation for gaps in your record is that those cities simply had no supply or demand changes during the time period between cycles.

[solo]

samson,

I have suspected for quite a while that the 16 turn cycle is just a trigger, and have been looking for underlying causes to explain the gaps, such as slow city growth, etc.

The only other triggering events I have noticed so far in my test game are commodity deliveries, which sometimes trigger a change in the Source or Destination city.

Have you noticed any other triggers?

From what I have seen in your new thread, it appears that, once finalized, your quotients and wildcards will be able to predict all commodity changes, and all the causes leading to them. A very nice piece of work!

[samson]

Solo - Other triggers?

The only one I'm certain of is the Reveal Map command in the Cheat menu. After the map is revealed, clicking on a city display will update the supply/demand list immediately.

I'll see if I can think of some tests to catch others.

SG - I don't think the Reveal Map command changes supply/demand by simulating "exploration". I think it causes city displays to be updated immediately when you enter. I think the 'glitch' you mention is a real bug.

Every time I have seen that 'glitch' it involved the commodity Dye. I suspect there's a Dye-related bug which causes an inordinate number of cities to demand Dye when they are founded or when Trade is first discovered. This false Dye-hunger seems to disappear after the first Solo cycle.

[solo]

samson,

Without exception, the first cycle observed has always placed or retained dye in the first position of city demand lists. So far, in my hotseat test game, where many cities have had a second or third change, none of these dyes have been dislodged, yet!

In the other short SP test game using SG's start to verify the use of the cycle, I observed this same behavior with dye.