A pattern governing supply and demand changes

[samson]

One thing I have noticed in all my games is that the Germans are always wanting wine, much more so, on average, than other nationalities. Demands may be linked more to individual civs (not just colors). I know you are testing for this, but are all civs being included in a systematic way?

No. I've stumbled across a couple of Civ-related associations. Silk supply is higher for the Chinese. Wine supply is higher for the French. But with 21 different civs, it would complicate things enormously to do systematic testing of all civs. It's possible, though, that each civ has its own commodity that it favors for either supply or demand.

[Elephant]

I would have thought Wine for the French; the Germans should be Beer... and the Russians, Vodka...

My mind boggles at the amount of time you two have put into this issue in the last few weeks. The rest of our measly contributions are but 1% of the volume you guys have cranked out. A hearty thanks and "salut" from all the rest of us with families, kids, and jobs...

[solo]

Elephant,

Thanks! As my username suggests, I am not married, nor have a family to distract me, so a lot of my spare time goes into these investigations, because I get a lot of fun out of finding out new things about the game.

On the other hand, Samson is just real good at doing this. He probably only spends a half hour or so every day working on these problems!

Samson,

Another oddity with the city Washington on a cycle turn. This was the city that allowed me to build a dye caravan (without dye in supply) on a cycle turn earlier in the game. This time no commodities were changed, but the demand for salt (also the demand wildcard), became satisfied without having any salt delivered to the city.

Looks like some smuggling is being simulated, too!

[William Keenan]

IMHO there is a different formula for each of the S/D slots. The first slot being strongly influenced by technology, the second having a strong regional influence (Continent Number?, Map Quadrant?). The third I have no idea.

Consider also that city improvements contribute to S/D. Building Superhighways causes oil to be demanded.

[samson]

William,

See the thread "How Supply and Demand lists are determined." The formulas are tied directly to the commodities themselves, not the slots. The slots are just the ranking of demand or supply quotients for the commodities, except for the wildcard phenomenon which is also explained in that thread.

As for testing for city improvements affects on the formulas, that's an excellent idea! I'll try it out. Thanks.

[solo]

I thought I might bring things up to date here:

After testing the formulae developed by Samson that predict which commodities will appear on supply and demand lists, the reason for the gaps appearing in the cycle pattern has become apparent. Each cycle turn, supply and demand values are calculated for each commodity, using a unique formula for each one. Commodities are assigned to supply and demand rosters, from which the supply and demand lists are filled. If the ranking of commodities in the top three positions of each roster have not changed since the previous cycle turn, the supply and demand lists will remain the same, and a gap in the pattern will appear for that city on that particular turn.

Many commodity values can be affected by dynamic factors such as changing city size, techs totals, specific techs, and/or by terrain changes. Other commodity values are based more on static values such as city geography and terrain. In addition, Samson discovered that “wildcard” commodities, which are based on a city’s grid coordinates, will also appear on commodity lists. More details can be found in his post. It is enough for us to know here that if a city has gone through enough dynamic changes affecting the calculation of the commodities at or near the top of its rosters since its last cycle turn, a change in the supply or demand list(s) is very likely.

One more thing I would like to add is a list of cycle dates (deity level) for the first city leading each group of 16 (cities 1, 17, 33 etc.) on the chronological list:

BC years

3250
2450
1650
925
525
125

AD years

220
540
860
1180
1500
1660
1764
1796
1828
1855
1871
1887
1903
1919
1935
1951
1967
1983
1999
2015

Since cities are processed in reverse order, the year preceding each in the list above would apply to the second city in each group of 16 on the chronological list. For example, on the 1918 turn, cities 2, 18, 34 etc. would be processed. In 1917 it would be cities 3, 19, 34 etc. Moving forward in time from 1919, for the turn 1920, it would be cities 16, 32, 48 etc. and in 1921, it would be cities 15, 31, 47 etc., working down a notch each turn to form the pattern illustrated in previous posts.

Since many maintain a list of “oedo” years, this can be used to obtain the same list as above, by just circling every 4th date. Intervening years on oedo lists can be used as a convenient reference for every 4th city appearing in the 16 turn cycle.

As for finding a city's position on the chronological list, the demand lists displayed by the Trade Advisor list cities in this order and the various lists can be used to piece this information together. If one of a city's cycle turns has been identified, using Samson's formula is another way of obtaining the city number. (Earlier in this thread, the use of the "Find City" command was suggested, but further investigation confirms that cities on this list appear in alphabetical, and not chronological order in MGE games. "Find City" does provide chronological lists in 2.42 games, but it's a good idea to trade maps with all the AI to make sure the list is complete and up to date, before using "Find City".)

A final note. Using up a caravan by making a delivery or by contributing to a wonder is another way of triggering changes to the supply and demand lists of cities. Again, if there have been enough dynamic changes, commodity values that are calculated will change, resulting in changes to these lists. To predict what changes will occur and what they will be, refer to Samson’s formulae, which can be found in this thread.

http://apolyton.net/forums/showthrea...threadid=67754

[SlowThinker]

Maybe I didn't read carefully enough, but...Destroying of any city in the world - does this cause changes in the pattern?

BTW, I think destroying of any city causes changes in happiness. Seehappiness quirk...

[solo]

Destroying a city does not change the pattern or the position of other cities in the list. The next new city entering the game will fill the empty slot and it will have the same cycle years as the city that it replaced.