Strategic Implications of Understanding Supply and Demand

[samson]

Originally posted by solo under "How Supply and Demand ..."

The knowledge of wildcarding has open a new realm of strategy. At the moment, it appears to me that when a supply wildcard hits its own list, it will be very likely for it to stick there for a long time, especially if it is one of the higher valued commodities, like wine, silver or gems. They don't seem to go through so many adjustments in the earlier half of the game. With dye, hogging spot #1, and when the demand commodity occupys the middle position of deamnd lists, there is only one position the supply wildcard can have on the DQ list in order to show up on the demand list. If it can't show up there, it can not leave the supply list, right?

It looks to me that one would want cities with a high quality supply commodity and with a frequently shifting demand one. If the high quality supply commodity is used to establish the first trade route, then it should become a valuable participant in an alternating supply situation, later on. A volatile demand commodity is apt to set up alternating supply sooner.

Another idea might be to select city sites where hides will be the supply commodity, to take advantage of its unique behavior.

Exactly! I was just thinking that we should start a new thread to discuss the strategic implications of understanding Demand and Supply. Here it is.

[samson]

With dye, hogging spot #1, and when the demand commodity occupys the middle position of deamnd lists, there is only one position the supply wildcard can have on the DQ list in order to show up on the demand list. If it can't show up there, it can not leave the supply list, right?

Not quite. Remember, the demand wildcard can mask the commodity in the 2nd demand position which could also be the Supply wildcard. And as I learned from Elephant's London problem, if both wildcards are the same, it is asserted on the demand list, not the supply list.

However, your idea of founding cities with valuable commodities for supply wildcards is a good one. To make it work though involves understanding the Demand Quotient formula for the commodity you want to persist on the supply list. From the DQ you can choose cities with terrain and other factors that minimize the commodity's demand. This will keep it off the demand list and on the supply list.

[solo]

It's too bad that we are limited in choosing good SSC sites by the necessary specials. However, during past games, these sites have provided decent commodities to work with, anyways.

It will certainly be worth while maintaining complete S/D lists for an SSC and other important cities. Having some control over the timing of new appearances and over which commodities appear will be much more fun than just hoping and waiting for good changes to occur!

[samson]

It's true that SSC sites are chosen mainly for resources. But knowing in advance what commodities it will supply (even before you've discovered trade) can be used to found trading partner cities on other continents with complimentary demands. For example, knowing your SSC will supply Gold would allow you to found a trade partner city whose map coordinates guarantee it demands Gold.

Or take Gems, which can be a supply wildcard and whose demand is also determined by location. By solving both formulas simulataneously, you can derive a list of locations which never demand Gems and which have a Gems supply wildcard. This means that they must supply Gems. Then you can set up the trade routes in a way to provide a continous self-refreshing Gems supply. At least until you reach 32 techs and the wildcard formula changes.

[DrFell]

I could use knowledge of supplied goods in MP games. One of my most often used strategies is the 'caravan rush', building large numbers of caravans and sending them to an off-continent trading partner. Having better knowledge of supply/demand would really help as I could pick out city spots more likely to supply valuable goods. Also, knowing the Chinese supply silk more often than the avarage civ might make them one of my future MP favorites. I'm going to have to test them out more often.

[solo]

Yes, I am liking very much how the precious metals and gems can be guaranteed by picking the right city coordinates. To be able to predict in advance, opportunities for the most profitable alternating relationships, rather than just obtain them by chance, is good news.

[samson]

I think there are two areas of application of Supply and Demand theory to strategy. The first is in choosing the location of cities. Knowledge of what commodities the city will supply and demand, based mainly on static factors such as terrain and location, could be as important to site selection as terrain, resources, and trade potential. Of course to make the calculations of the supply and demand quotients practical, we're going to need a little tool that can derive the lists and wildcards quickly from input parameters.

The second area of application is incorporating knowledge of the dynamic determinants of Supply and Demand (tech acquistion, city growth, trigger events) into long range planning: choice of tech path, scheduling of WeLoveDays, caravan delivery timing, etc. Although complex on a civ-wide basis, such calculations are doable with large impact cities such as an SSC. For challenges like OCC or early landing, the influence of Supply Demand theory could be significant.

On a practical note, it seems that the deliberate founding of cities that supply Uranium can be achieved by determining what combination of factors results in a supply list of fewer than three commodities. After Nuclear Fission is discovered, these cities will always supply Uranium.

[solo]

Yes, clever, get the one you want on, by making sure none of the others are eligible!

[Elephant]

Seems like the SSC can still be founded on a 3 or 4 special group, but we want to be careful in where we found the "caravan feeder" cities in the neighborhood. We will not be able to change the continent or hemisphere early in the game, but we can select or modify terrain and choose exact tile location to try to optimize those nearby cities.

[Elephant]

Practical Application of Supply/Demand Issues:

I'm playing GOTM23 from CFC as an OCC. My city was founded at 18,4 on Plains, with only one other Plains and one Tundra squares in city radius. The Tundra is MuskOx (Fish-special), there are also one Fish and two Whales in radius. The game is on Normal 50x80 map, Emperor level, 2 settlers, #6 civ/Greeks, continent number 3, 2 tiles. At AD1 I am at 6 citizens.

My starting Supplied commodities were Beads, Salt, Oil. I discovered Trade in 575BC. I have built 2 caravans, Beads and Salt, working on a third now. Don't want to make Oil if I can help it.

Questions: Is there any way to force Oil off my Supply list before building the next caravan? Or anytime in the near future? I'd love to get Dye or Hides, but that does not look promising from the formulas.

Also, Samson, have you looked at the issue of getting a repeat Supply, where it does not become unavailable the first time you build it? Seems like Hides often does that, but it has happened at other times.

[samson]

Questions: Is there any way to force Oil off my Supply list before building the next caravan? Or anytime in the near future?

You need to run the numbers to get the SQs and DQs for all commodities. Find another commodity with a high SQ and figure out how to boost it above Oil. Perhaps there's already one with a higher SQ on the Demand list. Then you'll need to lower its demand below its supply quotient to flip it over to the supply side.

Also, Samson, have you looked at the issue of getting a repeat Supply, where it does not become unavailable the first time you build it? Seems like Hides often does that, but it has happened at other times.

If by "repeat supply" you mean it doesn't become unavailable when built, only Hides does that. It's a unique property of Hides that it never becomes unavailable when on the supply list. If you think you've seen other commodities behave that way, I'd love to see you post a save of it.

[Elephant]

I am not confident in my math, but I could not figure a way to FORCE another commodity beyond growing the city. I grew the city to 23 and Gold showed up on my Supply list. I ran the numbers for that and in hindsight can see what I missed. The interesting thing is that I have been able to make 3 Gold caravans so far. I had made and delivered the Oil caravan for a third route previously. The extra Gold openings seemed to coincide with dumping a pile of food caravans into a wonder.

[Elephant]

Salt showed up again after a while, then Beads and finally Gold again. Each of them is () after delivery, but something opens up after a few centuries or a dump of stockpiled food caravans into a wonder. Is this expected behavior? I can attach multiple saves from this game if desired.

Agree that Hides is the only one I have seen not get () upon delivery, but the commodities are losing the () at irregular intervals in this game. And Oil has disappeared...

[samson]

Originally posted by Elephant
Salt showed up again after a while, then Beads and finally Gold again. Each of them is () after delivery, but something opens up after a few centuries or a dump of stockpiled food caravans into a wonder. Is this expected behavior? I can attach multiple saves from this game if desired.

Agree that Hides is the only one I have seen not get () upon delivery, but the commodities are losing the () at irregular intervals in this game. And Oil has disappeared...

I don't think anything unusual is going on here. The Supply/Demand formulas should tell you why commodities come and go from the S/D lists and how to predict when they will.

The availability of commodities for supply (whether it has "()"s or not) is in your control. Commodities become unavailable either when they are in transit (i.e. after built and before delivery) or when they are participating in a trade route. Careful management of your trade routes can usually ensure that at least one commodity remains available (unless in transit). Of course, you'll need to time the delivery of your third commodity to trigger a change in a your supply list in order to get your fourth commodity, but once you've got that you should be set.

[Elephant]

So if I can change the SQ of something somehow so that a commodity that is not in a current trade route or caravan gets on the list, it becomes available despite having three existing trade routes and 1-2 caravans on the way? With an OCC the only things I can play with are number of techs and number of game turns. With new commodity caravans I can control the time of delivery. One thing that might be important is which of the three positions the delivered commodity occupies at time of making caravan and delivery. Have to test that out.

I'd still like some more "practical" advice on how to juggle the SQs over time in OCC to free up new commodities. Someone once talked about delivering in a particular order (2nd, 3rd, 1st?) as a way to free up new ones - does that work?

[samson]

Originally posted by Elephant
So if I can change the SQ of something somehow so that a commodity that is not in a current trade route or caravan gets on the list, it becomes available despite having three existing trade routes and 1-2 caravans on the way?

Correct. What the "()"s around a supply commodity mean is that the commodity is either used in a trade route or is in a caravan/freight supported by the city. If a new commodity comes onto the supply list, it will be available.

Originally posted by Elephant
With an OCC the only things I can play with are number of techs and number of game turns.

Well. there's also changes in city size and terrain modifications. But the first thing to do with the formulas is find out how when the lists will change and plan your trading accordingly.

Originally posted by Elephant
With new commodity caravans I can control the time of delivery. One thing that might be important is which of the three positions the delivered commodity occupies at time of making caravan and delivery. Have to test that out.

What "three positions"? The 3 slots in the supply list or the list of trade routes?

Originally posted by Elephant

I'd still like some more "practical" advice on how to juggle the SQs over time in OCC to free up new commodities. Someone once talked about delivering in a particular order (2nd, 3rd, 1st?) as a way to free up new ones - does that work?

That's a tactic for refreshing existing supplied items to make them available again. It doesn't have any affect on new commodities being added to supply list.

[solo]

Another way of getting new supplies is by forcing changes on the demand list, which can cause a commodity to change from the Supply Roster to the Demand Roster or vice versa.

Also remember that wildcards can be manipulated on and off lists, too, and in subtle ways, often revealing a covered supply or becoming a new one in their own right.

A general rule might be to trade the wildcard supply commodity first, so it ends up as the first trade route. It is most likely to return to the list later, if it leaves, and will be freed up by the 4th trade if it hangs around, which is likely.

[solo]

If all else fails and you are stuck on food, another little trick I forgot to mention is to NOT look at your city during its cycle turn. This often results in supply and/or demand list changes. Same thing goes if you are stuck on food when delivering your last caravan or using it in a wonder. Wait until the next turn to look at your city. These list changes caused by not looking after trigger events occur because of the dye-copper bug.

[Elephant]

Seems to me the greatest likelyhood of doing all the math and analysis would be for OCC. My usual OCC strategy is to rush to get Shakes and max my city size out. From then it rarely changes, so that's not a variable I would want to manipulate casually. I could make a Settler/Engineer and disband it, then WLTPD or foodbox-grow back again, but I'd lose some trade arrows, which I've maxed for research, and OCC is a research bottleneck.

Terrain is another thing I can vary, but it takes a lot of effort and usually requires Engineers, which I would want to be bribed ones but could be the "temporary" ones mentioned above. Unfortunately GOTM23 is not a good test of this as there are only two terrain tiles in the city radius that can be changed. Most of the time I am trying to maximize trade arrows anyway, so I avoid making forests or hills until end-game.

The only other variables that I see when doing OCC are gameturn number and number of techs, with a few like Indust and Nukes in special places.

I think a real useful analyzer for this is going to have to be more than just a spreadsheet - the spreadsheet can tell you what your current SQ & DQ are, but are pretty hard to manipulate to tell you what you can do to the current situation to get a new commodity or make a particular one rise up. Someone who understands the interplay of the variables would have to write this kind of analyzer as a standalone program like CIV2PLAN. There needs to be user options like "preserve city size" or "maximize city trade" for OCCers. And some way to say "show me how to get this commodity to appear". Perhaps it could look at the current situation and say "here is a list of commodities you COULD get, if you did the following things..."

[solo]

Those are all good ideas, Elephant, and I do agree that once a city is full size in OCC, it is very easy to get stuck on food, even with all this new knowledge. Hopefully someone will come up with a good utility taking all this new trade info and using it in the ways you describe.

However, I do have one more idea for you to get a large OCC city unstuck. Being OCC, I'm sure you are building and saving plenty of food caravans. Try using one to start a wonder, (any wonder will do) and then look to see if this use freed up any of your supplied commodities. If so, convert the unwanted wonder back to half of a caravan, and build the rest of it to get the new commodity to trade. This does cost an extra 25 shields, but is apt to be worth while when it works. For a good commodity, it's probably worth taking the chance.

[samson]

Originally posted by solo
If all else fails and you are stuck on food, another little trick I forgot to mention is to NOT look at your city during its cycle turn. This often results in supply and/or demand list changes. Same thing goes if you are stuck on food when delivering your last caravan or using it in a wonder. Wait until the next turn to look at your city. These list changes caused by not looking after trigger events occur because of the dye-copper bug.

Solo,

This is a clever idea, if Dye or Copper are already on your supply list. By purposely letting the D/C bug happen, you can force something else onto your supply list. However, if neither Dye or Copper are supplied, this will have no effect on supply.

I do have one more idea for you to get a large OCC city unstuck. Being OCC, I'm sure you are building and saving plenty of food caravans. Try using one to start a wonder, (any wonder will do) and then look to see if this use freed up any of your supplied commodities.

This won't work. Using caravans (including food) to build a Wonder acts as a trigger to change supply/demand lists, but can not free up currently blocked commodities. Blocking "()" is caused by a commodity being in an existing trade route or in transit. Delivering food caravans to a Wonder does not affect this. Its function as a trigger can only make visible the changes in the supply list that have already been caused by other factors.

Elephant,

I agree with you that other factors in a game, such as growing your city at a particular point in time, will outweigh trade supply considerations. And therefore, few options will be available to actively change commodities when all are blocked. The best use of the formulas is to predict when supply lists will change and to plan your trade strategy accordingly. More important, perhaps, is planning how you will set up your trade routes in order to keep at least one commodity continuously available for supply. It is easier to not get blocked than to get unblocked.

[solo]

Samson,

With what we know so far I would agree, but try the food and wonder trick using this save. After loading 320 AD, hit return to go to 340 AD, which is when Space Flight is discovered. Go into Atlanta and verify that all supplies are blocked. While there, start building Apollo and activate the food caravan just made by Atlanta. Leave the city and use this food to start off Apollo. Check Atlanta again. Bingo! (wine) has become wine.

(Edited after: At this time there was a gems in transit, so our understanding may still be correct, as I believe a gems delivery would have unblocked the wine, too, but I think I still remember doing this earlier in the game, when nothing was in transit. I will go back and check.)

[samson]

Solo,

The mystery in this case, is why is Wine blocked in the first place? It is not on the trade routes list and is not in transit. So how come it's blocked? The caravan delivery merely acts as a trigger to refresh the lists (including the blocking status of all commodities).

If you have earlier saves, see if you can discovery when and how Wine becomes blocked.

[solo]

Here's another and a better example, without anything in transit. Try the same thing with Atlanta. I have to go out for awhile now, but will be back later to discuss this. In the meantime, try it again here.

[samson]

Okay, the answer to this riddle is that caravan deliveries don't always act as a trigger for the source city. They only have that affect when a new route is established or when an existing one is replaced. Or when the year is the city's solo-cycle turn.

Since triggers are the mechanism that updates supply commodities' blocked status, after the delivery of a non-route-establishing caravan the commodity will remain blocked until the next trigger event.

This means triggers are doubly important. Not only are they the events that change the S/D lists, but they can also make blocked commodities reavailable for supply (if they aren't blocked for other reasons.)

And this explains why your trick of delivering a food caravan to a fake wonder sometimes works. Since receipt of a caravan triggers trade display updates, it will unblock any commodities which are in "limbo". As will the arrival of the city's solo-cycle year.

[solo]

If you fool with the 320 AD save a little more, trying different triggers, I think you'll find that the only trigger that unblocks wine is the use of Atlanta's food caravan in building Apollo or another wonder. Delivering Atlanta's food to another city does not work, neither does the delivery of gems to Kyoto, nor the delivery of any helper city commodities to Atlanta. The type of trigger seems to make a difference here.

(By the way, Atlanta's next cycle turn was 360 AD. Also, if you deliver the gems first, both wine and gems become unblocked when Atlanta's food contributes to Apollo.)

So far, my tests show that to unblock commodities in "limbo", you need to build a food with the Source and use it in a wonder somewhere. Otherwise you're stuck, unless there are subsequent S/D roster changes, which any trigger can set off.

There may be some connection here to the fact that after roads joined Atlanta to Rome, the 3 visible trade routes were locked up, putting all subsequent commodities into "limbo".

In my previous game using the other map, I left the first trade route "open" in my SSC to enable alternations, afraid that 3 trade routes to the roaded AI city would block supplies. Then I discovered that food built by the source and put into a wonder worked as an unblocker, too. In my second game, I wanted those extra trade arrows from the road and railroad, and found that I could have them, plus unblocked supplies, by using this little trick.

I think the fact that when Atlanta built a commodity NOT on the supply list (it happened to be food) and used IT as a trigger, it acted to unblock wine, which had been in "limbo".

I am still not sure this explanation is correct. It's just my latest theory. However, I'm a pragmatist first, and doing this works for me in actual games. I will have to experiment more with different saves to see when and how often this trick works.

[samson]

Originally posted by solo
If you fool with the 320 AD save a little more, trying different triggers, I think you'll find that the only trigger that unblocks wine is the use of Atlanta's food caravan in building Apollo or another wonder.

Not so. Any trigger will unblock Wine, regardless of whether there are S/D list changes or not. If you wait until 360AD and merely view the City Display, Wine will unblock. If you deliver Gems to Kyoto in 360AD it will unblock both Gems and Wine.

As I said above, delivering a caravan is not a trigger for the source city unless a new route is established. This is the key.

There were two new (to me, anyway) findings in my previous post. One is commodity refreshment is based on the same trigger mechanism as S/D list changes. And the second is that not all caravan deliveries act as triggers.

There may be some connection here to the fact that after roads joined Atlanta to Rome, the 3 visible trade routes were locked up, putting all subsequent commodities into "limbo".

Kyoto's base trade isn't high enough for freights delivered to it to replace the Rome trade routes with their railroad bonuses.

[solo]

We may be caught up a bit here in our trigger definition or about the turn being considered.

My point was that during the 340 AD turn, the only way I could unblock wine was by using the food produced on that turn by Atlanta to start a wonder. Perhaps it would have been clearer for me to say that the only "action" instead of "trigger", and to limit my observations to just that non-cycle turn.

I agree with all that you say about 360 AD, which was the cycle turn for Atlanta, and agree that not all caravan deliveries act as triggers.

However, it appears to me that food caravans used in building wonders have the same unblocking effect as waiting until a city's cycle turn, for cities have roaded routes, such as Atlanta. With this trick, you don't have to wait until the cycle turn to unblock supplies which are eligible to become unblocked. Also, when using a food to make a wonder you do not need to create a new trade route in order to unblock a supply.

I know that a road connection creates higher base trade, keeping new non-road deliveries off the trade route list. My point was that because of this, unblocking works differently in this kind of situation, and that starting wonders may be a good unblocking strategy for cities with roaded routes in place. For cities with 3 unroaded routes in place, this trick does not work for unblocking any supply commodities appearing on the three trade routes.

[samson]

Delivery of a caravan (food or commodity) to a Wonder is a trigger. It will therefore free up "limbo" commodities, as well as cause any S/D list changes to become visible. This works whether the Wonder is being built in the source city or any other. If you don't want to lose half a caravan each time, you can have a helper city or two in slow wonder-building mode ready to accept food caravans and act as triggers. These "wonders" could be converted to SS parts later on.

If Kyoto's, or another unroaded city's, base trade was high enough it would replace the roaded Rome routes and act as a trigger to free up any limbo commodity immediately.

The trick can work with unroaded routes, too. If three unroaded routes exist, the delivery of a commodity to a city with lower base trade will also create a limboed commodity. The "wonder bread" trick could free this up as well.

[solo]

Yes, I agree. In further tests, I tried wonders in other cities, too, but did not think of keeping them for parts. A great idea that would have helped a bit in my last game!

I have yet to have the base trade exceed that of those to the roaded city, but agree this is possible with a Destination with enough trade arrows. Very unlikely, though for an AI city in early landing games, which move so quickly.

You are right about the trick working with other cities, too, but it is limited to commodities in "limbo", i.e. ones not making it onto the trade route list. In the save I supplied, I tried this with New York, where it worked with a limboed commodity, but as mentioned in my previous post, the trick will not work for blocked supplies that also appear on the trade route list.

I think we have total consensus about all of this now, and have gained some additional ideas for expediting the unblocking of commodities.

"Wonder bread!" Brilliant! I like that!