Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TrashMan]

The problem with ships costs is directly tied to the problem of NO COST for the AI, whose fleets you're expected to fight.
If BOTH sides have to pay for their fleet, even in some simplified, abstracted way, then there is balance and consequence.

[Chaos Blade]

Range was in one of the 4 prior iterations of the economy. It was removed because it was too complicated. People complain already that they have to babysit their colonies. It would be hell if you had to figure trade range into the market as well

True, but some kind of range mechanic would be interesting. at least it would make the expeditions more sane looking, specially if said expedition's parent faction is at war.

Hmmm... maybe they could tie likeness of expedition to lack of war (against factions other than the pathers and pirates, obviously)

Other than that, there should probably be a whole redone of the accessibility number. Still I'd love to see some sort of trade range mechanic for non core colonies, it would make colonying more of a challenge and make your faction more of a regional one.

Hmmmm.... I really think shipbuilding needs to be divorced from HI, and maybe need extra requirements for you to build hulls at your colony.
Maybe increase the production tics of some elements? (like Hulls) that way the market "stock" waxes and wanes
How? say hulls are produced every X months, with the whole sector surplus production is added to a pool (shown in the economy interface)
So, say Colony a produces 5 hulls, B 3 and C 10, a total of 18 hulls, every three months, but there is a demand of X hulls/monthly, that might fluctuate a bit and respond to other factors (demand, war between factions, pirate raids (successful), pather attacks (successful), so even if the average demand is 6, the concrete average should be different

while the whole idea might be a pain in the ass, the idea of fluctuating dedmands due to events is something that I'd like to see, though, and could help a lot to make the market more interesting

[speeder]

The old economy handled distance right in my opinion...

But did something wrong too that made it so crappy.

In the old economy each market would detect trading partners, and they would trade between themselves ONLY. It was kinda silly.

Right now Alex went to the other extreme, everyone trades with everyone else at same time...

But this introduces other issues:

1. No distance mechanics... meaning you can put a tiny colony on the extreme far corner of the universe and corner the market anyway.

2. As consequence of the first... fuel costs are not an issue. Traders will happily trade across the whole universe somehow, people won't consider distances when choosing where they buy fuel and supplies and so on.

3. As consequence of 1 and 2... the game has tactics, but no strategy, supply lines, choke points, and so on, are not a thing for AIs... Also if your colonies can sustain your fleet there is no reason to build near the core.

[Hiruma Kai]

Actually, the new system does have distance mechanics.  Under accessibility, there is a "Proximity to other colonies" bonus.  Perhaps that just needs to be tuned way up, so that its like -200% at the edge of the sector.

Perhaps there needs to be an accessibility for the core worlds, and a separate one for your own faction colonies in case you put down a few colonies near each other and so they exchange goods with themselves but not the core worlds.  Still simpler than an accessibility between every pair of worlds (which is something Alex wanted to avoid).

Or perhaps accessibility should still be 1 number,  with no distance penalty on the primary number, but to each other colony it falls by 5% or 10% for every light year between you and the other market its comparing against.  That is kind of what its doing now, except to the center of the core.

Perhaps a tool tip that lists all the colonies by distance and combined accessibility for the top 10, kind of like the exports tool tip.  And then add a colony tab that takes the Custom production screen's select a colony button on the upper left, and combine it with the intel screen's planet tab to list the accessibility combination from the upper left colony to all the colonies in the list, and list information that way.  As it is right now, while you know your own planet's accessibility, you still currently have to manually click on a different colony to see their accessibility and know if you're effectively trading with them.

[nomadic_leader]

Is anyone an economist? Surely some of this stuff has already been successfully turned into algorithms and simulated for academic journals.

[Goumindong]

Is anyone an economist? Surely some of this stuff has already been successfully turned into algorithms and simulated for academic journals.

I am an economist and I suggested against doing that for computational, understanding, and functional reasons. You’re likely to get a better simulation for the player facing side by doing something incredibly simple (like this) than doing something complicated.

[Eji1700]

Is anyone an economist? Surely some of this stuff has already been successfully turned into algorithms and simulated for academic journals.

I am an economist and I suggested against doing that for computational, understanding, and functional reasons. You’re likely to get a better simulation for the player facing side by doing something incredibly simple (like this) than doing something complicated.

Seconding this.  I know alex is getting a ton of flak on the economy system right now, but it really is a great improvement over the previous versions.

 I think there's a few things that would help (forcing colonies to be specialized would be a good start, right now all of mine look identical), but this is so much better than trying to realistically model a real life economy (which in a game that's trying to make the player feel important is often even easier to break).  There's lessons to be learned from understanding economics and how they apply, but the absolute most complicated you'd ever want to get in a game is something like offworld trading company or maybe EVE, and even then you're looking at games that are 100% built around them.

[Chaos Blade]

Is anyone an economist? Surely some of this stuff has already been successfully turned into algorithms and simulated for academic journals.

I am an economist and I suggested against doing that for computational, understanding, and functional reasons. You’re likely to get a better simulation for the player facing side by doing something incredibly simple (like this) than doing something complicated.

Seconding this.  I know alex is getting a ton of flak on the economy system right now, but it really is a great improvement over the previous versions.

 I think there's a few things that would help (forcing colonies to be specialized would be a good start, right now all of mine look identical), but this is so much better than trying to realistically model a real life economy (which in a game that's trying to make the player feel important is often even easier to break).  There's lessons to be learned from understanding economics and how they apply, but the absolute most complicated you'd ever want to get in a game is something like offworld trading company or maybe EVE, and even then you're looking at games that are 100% built around them.

I think part of the issue is that accessibility is too... soft, I mean, yeah distance would be a pain in the ass, but, perhaps a distance to the core could work, given the matter that colonies, new ones, would be in the periphery, that mauls would probably insure the core kept its primacy over some new piece of land.
Would even play to the core world mentality of the sector, to see goods from elsewhere as... less than.

Another idea is some develoment value that would modify the value of goods (sorta like a quality of the item) the value would float, and be related to things like how fast the colony grows, open port, and maybe some support buildings.
The faster a colony grows, the faster its development decreases (as the new people are in more precarious accommodation, for instance), so that could also work in making colonies grow slower (or you go for the option of making a ton of cheap ass knock off products, like the Song Tredees, or the Mountebank time pieces or the Poshe sport cars!)

[Drone_Fragger]

Is anyone an economist? Surely some of this stuff has already been successfully turned into algorithms and simulated for academic journals.

I am an economist and I suggested against doing that for computational, understanding, and functional reasons. You’re likely to get a better simulation for the player facing side by doing something incredibly simple (like this) than doing something complicated.

A good example of this is how hitpoints and armour work in rimworld v Dwarf fortress, which have pretty similar gameplay styles and essentially attempt to simulate the same thing. One does it by tedious mechanical analysis of various layers of material against various layers of material and how sharp said layers are compared to eachother, and the other rolls some dice and does some coin flips to decide which part gets hit and how badly.

Both produce the same essential output, and to the player there's no real difference in them. Parts come off, armour deflects hits, blood, hitpoint loss, etc.

[Cik]

fair points all i think

what is really needed is just a more concrete handling of ship production generally. once the AI can't pull infinite ships out of it's hat things will be a little more sane.

but yeah, i think range has to be added back in somewhere.. the sector is supposed to be in a death spiral so the fact that trade is so omnipotent and untouchable is a real problem.

we will likely see some significant improvements to diplomacy and state interaction going forward which will make the AI polities behavior less insane and more bounded by actual resource limitations one hopes. trade (outside well-protected convoys) should be quite rare, and only for cases in which there is simply no other way of producing that good at a place that desperately needs it.

[Baro]

I've always thought Alex's designs were generally always perfect, but the economy is the first time that it seems he's really missed the mark.  Even in the blog posts early on I never felt like he was tackling this from the right mindset.  His attempts to abstract the economy to such an extreme degree ends up complicating it and making it extremely unintuitive.

I'd rather the resource production and consumption of buildings be real pooled numbers with stockpiles being shipped off and sold based on demand.  If I produce 5 food, I'm producing 5 food and can supply 5 food worth of demand, not an infinite number of >5 demand markets.  Basic understandable supply and demand.  If I build a 2nd farm world, I'm increasing my production by that amount and it adds together, nothing unintuitive about highest-only.  I know we can't have a literal agent-level simulated market where trade fleets physically have to move all the resources around and figure out where to sell them, but there's got to be something a little more understandable and realistic than this.

 Then again look at Railroad Tycoon 3, I think that's a great potential model for industries and shipping.  In RRT3 it's a game mostly about trains and moving products around to make a profit, but what that game did that many other similar games didn't was fairly realistically model prices and shipping costs.  Every building in the game produced and or demanded something and would pay a price for their demands.  If they didn't meet their demands, the price for that good would tick up a little bit for that particular building/town.  If it was over-supplied, the price would tick down.  If a town with a steel mill demanded 5 units of ore and 5 units of coal, you could build a railway to supply those goods, but the value of that coal and ore would go down quickly as its demand was met.  If that steel mill also didn'thave customers for its steel, its demands would drop, and thus price.  The whole economy needed to keep moving to keep your trains profitable.  But, you could also directly own and build industries as well, making money not just from transport but also the profits from the industry.  Sometimes it was well worth it to run a train barely breaking even if it meant providing fast cheap shipping for a very profitable industry.  But the most interesting factor was that the whole system automatically balanced its self.  Goods could move around on their own to a degree,  and high demands/prices acted like gravity pulling products towards them, automatically sending goods to where they were needed/demanded.

So imagine we have 3 cities,  or in Starsector's case, planets.  Sindria produces tons of fuel, but how does the game know how to distribute that fuel?  Well, every planet would have a demand for fuel and if not met, the price and thus "attraction" for fuel would increase.  An AI merchant fleet loads up 500 fuel from Sindria and takes it to Corvis to sell,  Corvis only has a demand for 300 fuel though so the price of fuel in Corvis goes down, making it unattractive for future deliveries.  But that 200 extra fuel is still there, and a neighbouring system has a highprice for it, so a merchant spawns to move that fuel over to the system where its demanded once the prices are different enough to make it "worth it" to transport.

Goods would be zipping all over the sector, automatically going from where there is an excess to where there is a demand minus shipping costs.  There could be a bunch of pre-set trade routes working like a railway in RRT3, plus random traders to make up for any slack.  As a player our planets would get hooked up to these trade routes, which would calculate the shipping costs and hook us into the system just like everyone else.  We won't really need to know that our ore we mined ended up going from our planet to Corvis and then from Corvis to Sindria and then from Sindria to the Luddites, as long as someone's buying our ore it doesn't really matter.  What would matter though is that we're not over-supplying a product, and if we are, to make sure we can eat that reduced global price or even handle the fact that our over-supplied local good might not even be worth transporting until prices go up.  

This could also open up the ability for player-owned trade routes. Build a fleet of cargo ships,  assign them to a user-created route, and they'll automatically buy low and sell high as they ply their route.  Ore is $50 at your colony and the next destination in the route is paying $55?  The trade fleet fills up on ore then.  It sells all the ore at the next colony,  the price goes down to $54, but the next colony over is paying $65 for ore so it essentially re-buys the remainder that wasn't grabbed by local industries and sells it onto the next market again.  Every stop in the route is only compares the prices of goods between its current stop and the next, filling up its cargo holds in order of profitability.  This system works perfectly because even if a mid-point system doesn't need that ore,  the next system over does, increasing the price and thus increasing the "pull" for that good down the chain.  

It's hard to fully explain but it's not at all theoretical.  Railroad Tycoon 3 came out in 2003 and could handle massive maps with this simple but extremely robust (and fun to work with) system.

[DeltaV_11.2]

I think the issue Alex hit with goods based economies was that each market was checking against all the other markets to try and find suppliers/purchasers. That meant that the economy had huge computational costs which weren't sustainable. You had lots of little trades happening(along with checks about success of shipments, overhead from Java, etc) which clogged the system up. RRT3 makes a really big simplifying assumption: each tile only sees the 8 tiles around it. That makes the RRT3 economy prone to slow adaptation to change(it takes a long time for prices to travel) and theoretically profitable industries going bust. Which is good for RRT3's design, because it lets the player react to economic changes and get ahead of the market.


I think any reversion to a goods based economy would need to use a similar sort of simplifying assumption. Markets shouldn't be free to search the whole sector for the best deals, not just for computational reasons.


Having said that, I think that the current sort of single-market economy is workable, but the way it is now doesn't really work very well. Having a summed supply/demand model would seem pretty possible now- each planet generates supply and demand curves for various goods and then those get turned into real prices and quantities at each market, using availability to modify local prices.

[Mordgier]

I've always thought Alex's designs were generally always perfect, but the economy is the first time that it seems he's really missed the mark.  Even in the blog posts early on I never felt like he was tackling this from the right mindset.  His attempts to abstract the economy to such an extreme degree ends up complicating it and making it extremely unintuitive.

I'd rather the resource production and consumption of buildings be real pooled numbers with stockpiles being shipped off and sold based on demand.  If I produce 5 food, I'm producing 5 food and can supply 5 food worth of demand, not an infinite number of >5 demand markets.  Basic understandable supply and demand.  If I build a 2nd farm world, I'm increasing my production by that amount and it adds together, nothing unintuitive about highest-only.  I know we can't have a literal agent-level simulated market where trade fleets physically have to move all the resources around and figure out where to sell them, but there's got to be something a little more understandable and realistic than this.

 Then again look at Railroad Tycoon 3, I think that's a great potential model for industries and shipping.  In RRT3 it's a game mostly about trains and moving products around to make a profit, but what that game did that many other similar games didn't was fairly realistically model prices and shipping costs.  Every building in the game produced and or demanded something and would pay a price for their demands.  If they didn't meet their demands, the price for that good would tick up a little bit for that particular building/town.  If it was over-supplied, the price would tick down.  If a town with a steel mill demanded 5 units of ore and 5 units of coal, you could build a railway to supply those goods, but the value of that coal and ore would go down quickly as its demand was met.  If that steel mill also didn'thave customers for its steel, its demands would drop, and thus price.  The whole economy needed to keep moving to keep your trains profitable.  But, you could also directly own and build industries as well, making money not just from transport but also the profits from the industry.  Sometimes it was well worth it to run a train barely breaking even if it meant providing fast cheap shipping for a very profitable industry.  But the most interesting factor was that the whole system automatically balanced its self.  Goods could move around on their own to a degree,  and high demands/prices acted like gravity pulling products towards them, automatically sending goods to where they were needed/demanded.

So imagine we have 3 cities,  or in Starsector's case, planets.  Sindria produces tons of fuel, but how does the game know how to distribute that fuel?  Well, every planet would have a demand for fuel and if not met, the price and thus "attraction" for fuel would increase.  An AI merchant fleet loads up 500 fuel from Sindria and takes it to Corvis to sell,  Corvis only has a demand for 300 fuel though so the price of fuel in Corvis goes down, making it unattractive for future deliveries.  But that 200 extra fuel is still there, and a neighbouring system has a highprice for it, so a merchant spawns to move that fuel over to the system where its demanded once the prices are different enough to make it "worth it" to transport.

Goods would be zipping all over the sector, automatically going from where there is an excess to where there is a demand minus shipping costs.  There could be a bunch of pre-set trade routes working like a railway in RRT3, plus random traders to make up for any slack.  As a player our planets would get hooked up to these trade routes, which would calculate the shipping costs and hook us into the system just like everyone else.  We won't really need to know that our ore we mined ended up going from our planet to Corvis and then from Corvis to Sindria and then from Sindria to the Luddites, as long as someone's buying our ore it doesn't really matter.  What would matter though is that we're not over-supplying a product, and if we are, to make sure we can eat that reduced global price or even handle the fact that our over-supplied local good might not even be worth transporting until prices go up.  

This could also open up the ability for player-owned trade routes. Build a fleet of cargo ships,  assign them to a user-created route, and they'll automatically buy low and sell high as they ply their route.  Ore is $50 at your colony and the next destination in the route is paying $55?  The trade fleet fills up on ore then.  It sells all the ore at the next colony,  the price goes down to $54, but the next colony over is paying $65 for ore so it essentially re-buys the remainder that wasn't grabbed by local industries and sells it onto the next market again.  Every stop in the route is only compares the prices of goods between its current stop and the next, filling up its cargo holds in order of profitability.  This system works perfectly because even if a mid-point system doesn't need that ore,  the next system over does, increasing the price and thus increasing the "pull" for that good down the chain.  

It's hard to fully explain but it's not at all theoretical.  Railroad Tycoon 3 came out in 2003 and could handle massive maps with this simple but extremely robust (and fun to work with) system.

I think you are on the right path - I think standings should also play a factor. For example look at the real world where we have tariffs and trade embargoes.

Those would go a long way towards messing with the player economically rather than sending raids.

"Accessibility" is a weird way of setting up trading preferences, but really it should be standing. A faction that does not like me should try to avoid trading with me, and it can do so via tariffs that would lower the 'value' of my goods to them or an outright embargo.

What good is making all this fuel if nobody is willing to buy it from me? As is factions that dislike me, fund my war machines by buying my goods.

Giving me basic market control on how I price my goods would also be interesting, for example saying "sell fuel at 10% under market" would result in more purchases - on the other hand if I notice that I dominate the market already I could push my prices up - you're going to buy it anyway if there is just not enough fuel.

Sadly right now the goods and demand are highly abstracted with oversupply and shortfalls being impossible. I come into an existing market, build 7 colonies all with abundant farms, and yet there is demand for all of it! Was there a massive deficit before? Who is eating all my food?!

The end result is that right now we can build ALL the buildings we can every single time, max them out and make a fortune every single time. Demand is never saturated and production is always profitable. There are no market downswings. A peace treaty does not suddenly cause demand in ship parts to drastically drop nor does a war cause a spike in demand like it should.

[Alex]

Just wanted to chime in with a few details

"Accessibility" is a weird way of setting up trading preferences, but really it should be standing. A faction that does not like me should try to avoid trading with me, and it can do so via tariffs that would lower the 'value' of my goods to them or an outright embargo.

Standing factors in; being hostile with more factions reduces the accessibility of your colonies and therefore their market share.

Giving me basic market control on how I price my goods would also be interesting, for example saying "sell fuel at 10% under market" would result in more purchases - on the other hand if I notice that I dominate the market already I could push my prices up - you're going to buy it anyway if there is just not enough fuel.

That sounds like something where there's always a "correct" answer that maximizes income; if that's how it ends up, why make the player micromanage it?

Sadly right now the goods and demand are highly abstracted with oversupply and shortfalls being impossible.

If you press F1 on a commodity tooltip, you'll see locations with both oversupply and shortfalls for most commodities.

In a broader sense, oversupply and shortfalls are also possible. For example, if all Fuel produces except for Sindria were gone, Sindria would make much more export income, both in total and per unit of Fuel it produces. Likewise, if there were 100 Fuel producers, each individual one would make less per unit.

I come into an existing market, build 7 colonies all with abundant farms, and yet there is demand for all of it! Was there a massive deficit before? Who is eating all my food?!

Not exactly; all of those colonies are simply increasing your market share, but as your income is capped by the total market value, additional colonies producing food get diminishing returns on income per unit exported. E.G. if you have 20% market share, doubling your production will result in 33% market share. (20 / 100 vs 40 / 120). You'd be selling more units of food for less per unit, which is what you'd expect to happen in a supply/demand situation.

(There's a detail in that adding more colonies also increases the total market value for Food, but that's not going to keep pace with the diminishing returns.)

The end result is that right now we can build ALL the buildings we can every single time, max them out and make a fortune every single time.

That's true, and I'd like to add some benefits for specializing. But I think that's a separate topic from the economy.

[Goumindong]

Oversupply and shortfalls are possible. You just have to create them!

Starsector isn’t an economy simulator.. it’s a spaceship shooting semi-arcade game. The structure around the that is there to propel you organically into the combat and exploration content. Not to produce a number of spreadsheets as you attempt to optimize trade.

The only problem that the game probably has is that “good” planets can spawn too close to the core, which produces both high accessibility and possible production in multiple products