Forays Into The Unknown (EJ Builds A New Game Economy (Still))

[EmperorJon]

Another update. About a month and a half or so until I'll be back in full development, but I made some changes today:

Names make more sense and sound less stupid. Mode adjectives used, race names used, etc.
Ship choice for companies now more sensible.
Totally destroyed and began to rewrite the ship-for-station scripts. They now simply use a bastardised mix of current game scripts but paying attention to their homebases needs and other homebased ships.
Ships now jump if they have a jumpdrive and restock energy from their homebase like a player ship would.
Auto-replaced ships by GoD when a homebased ship dies now instantly removed from being homebased then destroyed, before handing it over to the routine to deal with getting a new ship properly.
Began more ideas and planning on outfitting, buying, equipping, upgrading, etc. ships, and also how to implement this into freetraders.
Finished giving all freetrader ships and account for now. This includes dock traders, military supply TSs, and others, for now.

[EmperorJon]

Tonight, "full time" work will reinitiate!

One thing I was thinking over the past few days was how the current economy cheats energy-wise.

I have 3 choices:

- Leave it as it is, and laugh at the player's desperate crystalplexes.
- Remove the player's need for crystals, free energy! (Face it, SPPs are expensive enough as it is.)
- Add crystal need for AI and see if my economic restructuring and growth algorithm convinces the AI that there's a huge shortage of energy... and food, and crystals, and silicon, etc. etc. And see if it all works out nicely.

[jack775544]

[UberWaffe]

Heh. I like this thread. I'm aiming way lower in terms of complexity in my concept for a civilian economy for XTimelines (of which the far-more-talented-than-me modders probably have a very good reason for going military economy only).
Anyway.

[EmperorJon]

Well currently nothing happens, but the selling of assets was planned.

As for everything else, it seems to be more relevant to what you're doing than what I'm doing; don't get me wrong, it's interesting, but it's confusing me how you're jumping between "This is what mine is going to do" and "This is what yours could do".

[UberWaffe]

Sorry, I apologise.

Verbal diarrhea.

[EmperorJon]

I'm currently trying to work out exactly how the new stations system will work.
Using the trade info function I can get necessary information on the storages of stations.
My original idea was to collect the storage information for every station in one race's sector group, so for example it would collect all the resource and product storage space in the core Argon home cluster around Argon Prime.
It would then weigh up these values to work out what factories were needed, for example if there was a total possible storage of Energy Cells as a product of 100000, and a resource storage of 200000, then it knows that if all the SPPs were full they'd only half fill all the factories once it got shipped out.
The issue with this is it pays no attention whatsoever to production times; it's all well and good seeing that the total storage space for Energ Cells being used equals the total storage space required, but factories, especially large weapons factories, use a few thousand cells every few hours... and XL SPP creates a few thousand every few minutes!
Does anyone know if the storage directly matches the factory size EVERY TIME, as if so I can work out the total storage in terms of sizes of S, M, L, and XL, and therefore the total output.
For example, if the total ore consumption is equal to 10x L, then I need 10x L's worth of mines, surely!
I also have to consider the resources needed for the step below. For example, instead of saying "I need ore, I'll find somewhere for a mine" it'll say "I need ore, but ore needs energy, there's not enough energy" etc. and work its was down. It'll also have to notice a surplus and then suggest new factories which can exploit that surplus while making a product that WILL be used somewhere by something... It may involve recursion. Ew.
All in all, it'll work out what factories are really needed where, then add them to a list. The factories on this list can then be built by any relevant company. By this method, the companies and economy grow building their own factories, and the fact that it's actually centrally pooling data to find where to tell anyone build is hidden and irrelevant. It just simulates companies deciding to build wherever they want in a dynamic way.

[Killjaeden]

[EmperorJon]

Just checking. Also, is there a command to get the production size of a station?

[Killjaeden]

don't know , sorry
But i remember that lucike had a function in his CAG script where you could check your complexes and it showed the excess/deficits of all wares in the complex. Perhaps taking a look at how he did it might save some time.

[Aydan]

What you're missing is the role the stations are playing in your economy. You're looking only at stations and ignoring one of the major gameplay parts. Stations producing a ware are only bricks in the middle of nowhere They do nothing but converting items into another form. If they have no more ressources, or their storage is full of products, they simply stop. They can not produce anymore. If they don't get supplies, they are useless, regardless if there's a demand only 20 metres away.
/edit: This post is about dependencies of traders and stations, and the behaviour has to be adjusted to the changes/scripts of EmperorJon

The most limiting factor on stations are NPC traders and THEIR behaviour.
It's NOT only a question of just placing them, i'll explain that later.
First, you need to calculate the average speed of a given trader job. That means you need to look at the jobs file and identify the ship type and calculate its average speed. That means you need to get the total amount of ships used to transport the ware in a given race, the range of variants and their possible max speed. You can then calculate the possible average speed. But the tricky part is that the equip libs are giving them only a limited amount of engine tunings, so you need to have a look at the equip libs also, to get an image of how many engine tunings they get in average. Only then you can calculate the real average speed, and believe me, it's far slower than you may expect.

Take a very close look at the flags on the traders, if they trade only in race sectors, if they are allowed to cross enemy territory and so on. The distribution of traders is a very tricky thing, and it'll take some days to get the distribution over all for just a single race.

Now, the speed is so important for you, because it may take longer to ship the products from one station to the next, because the way is longer than the actual cycle time on the station itself. That means that you may need multiple traders doing the same trade run to keep the consumer supplied. Otherwise the production will stall. A NPC trader with a max speed of 70 m/s is only flying at 35 m/s, because it's flagged to use only 50% of its max speed in the jobs file. It it is supposed to fly 35 km.... well, let's say it'll take a while, and a 35 km wide sector isn't that uncommon. Now add to that that the gates have to e used, because the station is in the next sector, and the cycle times of a station are getting obsolete, because the traders are the limiting factor with their flight time.

And now, here comes the hard part. NPC traders are independent and act autonomous. They decide on their own needs, and you have to place the stations accordingly, so you get a running economy between producers and consumers. NPC traders are rating trades by the amount of money they can earn in a given time range. That means they do look for the best trade and consider producer, consumer and the time they need to get there. You may have often wondered about traders with only 3 or 6 ore, although they could transport thousands of them. It was the best deal they could come up with in their range. And that may be only 6 ore. They calculate their current position to the point where they will buy for the best price. The ware is randomly picked from their job, but the calculation is not random. If the station is in the next sector, they calculate "Current position->Rendezvous point gate->gate->Rendezvour point gate->Rendezvous point station->station" That's the calculation for a single way. Stations and gates with extensively long docking ways (rendezvous point->station) have a decent disadvantage, because the speed is greatly reduced on docking. So TOAs and giant terran stations can get ignored, because the flight time can be longer than crossing an entire sector. Just measure the speed and time It's really true.

If the ship has reached its destination, it checks for the demand and buys accordingly to the demand. The same distance check is performed for the best sell run now. This is the point where you need to consider that the positioning of your station is critical, as the consumers must be near by. If it's located at the other side of a 70 km sector, but a gate is nearby, the way through the gate to the next sector my be much shorter, and all the ships will probably be heading there, because it's only 30 km through the gate, but 70 km in sector. This is the tricky thing behind NPC traders and distances, according to their behaviour.

Now the second trickiest thing is demand and the NPC trader behaviour. Demand is CREATED by traders. Only if wares are traded, there's a demand over all. If the product storage is full of products, there's no demand for supplies. The station can not produce at all. Only if the products are sold, it will start to eat up ressources again, creating demand of ressources. So if no goods a re sold, there's no demand, and the demand is strongly tied to NPC traders and their ability and behaviour to trade. Once again you need to have a look at the job files and equip libs to find out how large the average storage room for a given job in a given race is. Then you need to adjust the size of the stations on the map itself, to create a lot of demands. The thumb rule is, the larger the station size (XL and up), the worse the economical performance is.

The goal is to spread the traders as far as you can to have roughly the equal chance to have a trader ready to do a trade run, wherever it is. If you place a XXL SPP somewhere, the storage and producing capability is so strong, that after a while all NPC EC traders will gather somewhere around this SPP. And they will all calculate their trades from there, resulting in a monopolistic trade behaviour. The storage room is so large, that the price is never high enough to make other, smaller SPPs attractive to them, simply because they're too far away. The price only rises if a good chunk of Ecells is taken away, and this is calculated as a percentage. If you only have 3000 storage on a trader, but several 100ks on the SPP, and the producing capability, it never falls. So after a while the outer regions of the races sectors are stalling, no matter what you're doing. The storage room on the NPC traders is the absolute limiting factor. If not, the distribution of traders across the sectors is completely messed up. The best example are the paranid sectors south of Loo'Manckstrats. With only exchanging the size of some stations (and adjusting their position of course), you can stabilize the whole region. It's a lot better to place M and L factories, especially M ones. This keeps also the original God engine from turning them down. It's design flaw, as the vanilla map is a complete mess, and most of the helpless tries to get better results are changes made on the map itself and not in WareTemplate. So it gets so clunky that the reasons aren't obvious. The last try to fix this where the company traders, but they failed due to not taking in consideration that upping the jump range also creates a massive flight time for a best buy/sell run.

Anyway, if your just looking for demand and supply on stations, you're missing the fact that the wares have to be transported first, and the ways and storage capabilities are a lot more limiting the economy than you may think. Only in case of existing complex tubes you can calculate without these transportations. Consider placing stations, size of stations, amount of race traders/dock traders for a given ware, and their average speed/storage

[EmperorJon]

Hmm, I believe you're missing something too...
Each station belongs to a company. Each company hires dedicated traders. If a station needs more wares brought in, or more sent out, then it will get itself another trader.
Traders that upgrade themselves to fly as fast as a player trader, use the jumpdrive, and so on.

[Aydan]

I know.

The problem is, that you are creating centralised structures. If you assign a trader to a station, and he cares only about the problems of this single station, he has only a limited pool of suppliers and consumers. Theoretically you can get a map to work with this, but a single accident, some pirates blowing up a station which is a key element, and a complete region stalls. A single build mission somewhere placing an L Mine or an XL SPP and your map doesn't work anymore, because the dock trader needs to fligh further away, or the former consumers are supplied a lot faster than the actual station is able to do. /edit: If a key station is stalling, the whole error is cascading down (or up)

You will realise that the cunsumption of high end tech wares and weapons/shields/missiles is limited by your GOD engine eating up the products on EQ Docks and Trade stations. You can raise the efficiency artificially, but your system will be a lot more vulnerable. Relying on a chance based system and customised trader scripts is a lot more effective and the whole thing will be more flexible. You can, for example lower the priority for your GOD engine to respawn destroyed stations, you can reduce the time for the needed checks if your assigned traders are still existing, without the fear that a branch of sectors would stall.

[EmperorJon]

Er no... not quite. A single station is destroyed, and the game notices it needs to be rebuilt. The economy in that area maybe suffers whilst those who required that product need to go further, but well, that's what happens.
I don't quite see what the issue here is that you're finding.
For example, you say it's much more effective to use customised trader scripts... which is what I said I was doing in the last post. The general basis of this is already hammered out, though I need to talk to some of the experts in this area.

[Aydan]

I do not want to change the outline of your project I want to point out that there's an inbuild weakness you may be able to work around, as you do not divide between stations and ships (traders).

Just an example, if you have a station somewhere near the south gate of CEO's doubt, and it needs E-cells, then your station assigned trader will probably for the nearest source of e-cells from this specific point. It never changes, and if there's a bigger (XL) SPP in Cardinals Domain, he'll jump/fly there, every time, as the price of e-cells is always the lowest. If there's an enemy cruising around, he'll jump there over and over again, being destroyed every time. So you must include something that reminds a newly spawned trader that he was destroyed there. Kind of blacklist. All the other stations with similar distances will prefer this single SPP. Your trader losses will be relatively high, while other SPPs, although not that far away, won't be considered. Instead of this, if the starting point is random for a trader, the chances that this specific SPP is chosen as a e-cell source is considerably lower, so your losses of traders, because of this incident are a lot lower right from the beginning. Also the fact that other sources of SPP are contributing to the e-cell supply is keeping sectors from stalling. A station sleeping for only a single hour may result in a delay for several other stations which is a lot higher, because you have for your station : flight time to supplier+flight time back home+cycle time+flight time to consumer. If the station is supplying two others, the problem essentially doubles, resulting in a double amount of time for their consumers. This is cascading up, and it'll stall from a certain amount of time build up. Think of the war zones in AP for example, or highly frequented pirate sectors, race sectors with a fix pirate patrol route. Your ships would need to fly faster than the actual cycle time for your product to lessen the limiting factor of NPC traders. Even if you use the CAG strategy and jump your ships to 0,0,0, there's still amount of time to fly.

Now you can raise the efficiency artificially by letting your traders fly faster, letting them jump closer to their destination and whatever else. The resut is that the trading delay isn't limiting production anymore, and you will find the the consumption of high end goods is the limiting factor. The vanilla God is pretty slow when it comes down to consuming goods, and you'll see EQ docks filled to the brim. You'd end up adding additional trade stations/EqDocks.

ONE solution, which i'm suggesting, is to handle the over/under production on the side of the traders and not on the stations, as this is the original idea behind the X series trading. It creates a lot less static economical progression. If you have a massive amount of goods in the bellies of trader ships trnsporting them, a single failed station (destroyed, straving, whatever) is not that important, as the traders are spread out, instead of using a centralised command. Centralised is more predictable from a scripts point of view, but it's also not able to handle problems flexible. Especially overproduction is a lot more hreatening then under production, as overproduction means only god removes goods from the economy cycle. Instead of this you can reduce it by having randomly destroyed NPC traders, delay through slower transportation, and so on. Have you calculated how much your god is consuming over all?

It's not that i'm insisting on this I'm trying to point out that there's a weak point in your plan, and you may save a lot of scripting time/load of scripts running, nothing more Streamlining the trader scripts and jobs gets a better result overall. But that's not a MUST, it's only a suggestion I'm not trying to force you to do so, even if it sounds like this. It's not intended