Perfectly balanced economics?

[Cholmondely]

Just for information: Black Monks was written by a lawyer, First Finance by a banker!

[Switeck]

Just for information: Black Monks was written by a lawyer, First Finance by a banker!

I guess that'd make them perfectly evil were the 2 OXPs combined into 1!

[Cody]

<chuckles>

[szaumix]

I feel like economic balance in Oolite is the unicorn I care most to catch, and the hardest balance beast to tame.

It's also non-linear because of player skill and methods they use.
Torusing all the way to the main station? Really quick...if there's nothing in the way.
Fighting your way through large clumps of pirates in an Anarchy? Slow and can be very dangerous...with likely damage to your ship which incurs high costs, even if only refurbishment is needed more often.

Sure. But as a working argument, lets assume a vector factorized by only two values:
1. player skill + style
2. desired (rate of) costs and incomes
...then even the level of subjectivity you suggest becomes linear, and easy enough to balance.
My idea to deal with this goes something like Skilled_NPCs: imagine a simple script that just multiplies all equipment values by a value of the player's choice. The only thing left to do then would be to balance the underlying equipment cost values until their costs were more proportional (as with the Rescaling Experiment true realism would likely be undesirable).

If anyone cared enough it could then probably be split by theme -- pylon equipment, operating costs like fuel and maintenance, laser weapons, general ship upgrades -- with little extra effort. I think even multiplying average bounty or bounty range would probably be an easily slapped-on feature in such a script.

[szaumix]

I think this would be a great tool for oxp authors to use to help tweak the economy, I hope you will share it with us once completed.

A dynamic economy would be the most interesting, I believe, but probably the hardest to balance.

Sheeting values is easy and presents good quick perspective. The easiest things to define and categorize are set costs, and dockable shop multipliers. Working out income rate is much trickier -- mining being the easiest. "Bounty income" means average[mean] bounty * kill rate, average milk run profit is by far the most difficult considering the factors and variables at play, and I'm not as clever as the coders here but it would seem to be the hardest by far to tweak either by script or manually. Given that the galactic maps come into play, the only way I can think to determine it is to record trading profits like balance sheet data based on actually playing, probably using roughly the same record system as in real life. It also means any balancing of values should really be done around it; taking trading costs and profits as they already are.

Actually when testing the manifest scanner I found it was fairly easy to be a pirate and maintain a clean status. I've even scooped escape pods from ships I took out, docked with the main station in the same system and collected the insurance with no offender status. Lone traders are like low hanging fruit, nothing like facing a pack of six or eight pirates. Even escorted ships aren't that tough to take out and once the ship being escorted is destroyed the escorts break off and go looking for another ship to escort. I've done this and let the escorts go and still maintained a clean status.

This is my biggest problem with the Oolite piracy model. I am ok with the idea of getting away with piracy/assassination under certain conditions (no surviving witnesses) but if the game code didn't prohibit it, I'd have unprovoked assault lead to offender status on all ships with any witness not just police, and I'd spawn a ton of simple weaponless police drones patrolling the spacelanes everywhere trying to catch pirates in the act! I have vaguely considered that a black box reader could (/should) be present in GalCop stations checking the basic conditions of commanders' kills. I mean kills go to a commander's official elite rating, right?

Blah blah blah, too many ideas not enough time.

[Cholmondely]

I think this would be a great tool for oxp authors to use to help tweak the economy, I hope you will share it with us once completed.

A dynamic economy would be the most interesting, I believe, but probably the hardest to balance.

Sheeting values is easy and presents good quick perspective. The easiest things to define and categorize are set costs, and dockable shop multipliers. Working out income rate is much trickier -- mining being the easiest. "Bounty income" means average[mean] bounty * kill rate, average milk run profit is by far the most difficult considering the factors and variables at play, and I'm not as clever as the coders here but it would seem to be the hardest by far to tweak either by script or manually. Given that the galactic maps come into play, the only way I can think to determine it is to record trading profits like balance sheet data based on actually playing, probably using roughly the same record system as in real life. It also means any balancing of values should really be done around it; taking trading costs and profits as they already are.

Actually when testing the manifest scanner I found it was fairly easy to be a pirate and maintain a clean status. I've even scooped escape pods from ships I took out, docked with the main station in the same system and collected the insurance with no offender status. Lone traders are like low hanging fruit, nothing like facing a pack of six or eight pirates. Even escorted ships aren't that tough to take out and once the ship being escorted is destroyed the escorts break off and go looking for another ship to escort. I've done this and let the escorts go and still maintained a clean status.

This is my biggest problem with the Oolite piracy model. I am ok with the idea of getting away with piracy/assassination under certain conditions (no surviving witnesses) but if the game code didn't prohibit it, I'd have unprovoked assault lead to offender status on all ships with any witness not just police, and I'd spawn a ton of simple weaponless police drones patrolling the spacelanes everywhere trying to catch pirates in the act! I have vaguely considered that a black box reader could (/should) be present in GalCop stations checking the basic conditions of commanders' kills. I mean kills go to a commander's official elite rating, right?

Blah blah blah, too many ideas not enough time.

I think that a lot of all this depends on the view one has of GalCop and the octant.

Is one playing with the strong GalCop of most of Oofiction and the Galactic Navy OXP? They are strong enough to demand that all recording equipment kills are to made available to them, etc., etc.

Or is one playing with the weak GalCop of our Oolite page introduction and HIMSN - or even of the Vanilla game? How can they force these sorts of things upon people? And they have much more pressing concerns anyway.

[Stormrider]

Sheeting values is easy and presents good quick perspective. The easiest things to define and categorize are set costs, and dockable shop multipliers. Working out income rate is much trickier -- mining being the easiest. "Bounty income" means average[mean] bounty * kill rate, average milk run profit is by far the most difficult considering the factors and variables at play, and I'm not as clever as the coders here but it would seem to be the hardest by far to tweak either by script or manually. Given that the galactic maps come into play, the only way I can think to determine it is to record trading profits like balance sheet data based on actually playing, probably using roughly the same record system as in real life. It also means any balancing of values should really be done around it; taking trading costs and profits as they already are.

Well, if you get around to it, it would be cool to see the results.

This is my biggest problem with the Oolite piracy model. I am ok with the idea of getting away with piracy/assassination under certain conditions (no surviving witnesses) but if the game code didn't prohibit it, I'd have unprovoked assault lead to offender status on all ships with any witness not just police

There is no code that would prohibit you from creating such an oxp, but, first you would have to assume that any potential witnesses actually have Scanner Targeting Enhancement installed, second, that the NPC is actually interested enough in what you are doing to target you and the ship you are engaged with and finally that it arrived on scene early enough to see who fired the first shot.

I'd spawn a ton of simple weaponless police drones patrolling the spacelanes everywhere trying to catch pirates in the act! I have vaguely considered that a black box reader could (/should) be present in GalCop stations checking the basic conditions of commanders' kills. I mean kills go to a commander's official elite rating, right?

Total surveillance state? No thanks I think its more interesting to have a system that's more subtle, allowing the player to get away with some illegal activity but increasing the likelihood of getting caught with every offense.
Seems like a lot of trouble to keep yourself from engaging in piracy, anyway, if it is that tempting maybe you should just download the manifest scanner and go full on outlaw to get it out of your system .

[szaumix]

There is no code that would prohibit you from creating such an oxp, but, first you would have to assume that any potential witnesses actually have Scanner Targeting Enhancement installed, second, that the NPC is actually interested enough in what you are doing to target you and the ship you are engaged with and finally that it arrived on scene early enough to see who fired the first shot.

Really? I looked over  Shipdata.plist many times in the last month... and peeked at the AI files, and the way the populator overwrites all police's SCAN_CLASS and a few other police details had me thinking my options were more or less bottlenecked. Like most things I wish I could do, it's beyond my current abilities.

Total surveillance state?

Nah not total, just a little increased risk in higher tech commies/confeds/democs. Imagine two or three drones half the size of a thargon, maybe simple boxes or four-faced pyramids (so as not to waste CPU power) zooming back and forth along the spacelanes at 1000km doing nothing but witnessing every other battle. I'm pretty handy in a dogfight, I still think I could get away with it!

[Cholmondely]

Total surveillance state?

Nah not total, just a little increased risk in higher tech commies/confeds/democs. Imagine two or three drones half the size of a thargon, maybe simple boxes or four-faced pyramids (so as not to waste CPU power) zooming back and forth along the spacelanes at 1000km doing nothing but witnessing every other battle. I'm pretty handy in a dogfight, I still think I could get away with it!

I an see why Galcop might bother if they had the resources. But why would the commies pass the information on to them?

[Switeck]

I feel like economic balance in Oolite is the unicorn I care most to catch, and the hardest balance beast to tame.

It's also non-linear because of player skill and methods they use.
Torusing all the way to the main station? Really quick...if there's nothing in the way.
Fighting your way through large clumps of pirates in an Anarchy? Slow and can be very dangerous...with likely damage to your ship which incurs high costs, even if only refurbishment is needed more often.

Sure. But as a working argument, lets assume a vector factorized by only two values:
1. player skill + style
2. desired (rate of) costs and incomes
...then even the level of subjectivity you suggest becomes linear, and easy enough to balance.

If we're talking about fixed conditions, a linear outcome can only happen with fixed OXPs...and sometimes that means not even updating them, as some have grown more or less profitable.

I either quickly uninstalled or toned down aspects of many OXPs precisely because of the changes they made to game balance and credit-earning.

Plus the NPC changes in Oolite have upped the challenges as well, with OXP makers even pushing that higher with normally unreachable skill levels for the accuracy of ships. So I can't even fully compare my credit-earnings from how I played in Oolite v1.75 vs v1.90!
Cargo contracts are much harder as a result, and they are my late-game big credit earning method.
Even milk runs offer the rare challenge from a big pirate group that previously I could tackle.

"Solve for X." ...assumes the formulas are known and understood.
Sure, I can calculate within 100 credits how much I make doing "uneventful" milk runs of Computers and Liq./Wines in the starting Cobra 3 with a large cargo bay.
But the time part of the equation changes from v1.75 to v1.90 when I have to go further out of my way and risk getting a big ambush on me at the witchpoint buoy as well. Even the loading time for jumping from system-to-system is slower, due to the many complex OXPs I have.

I don't do Ironman challenges because I have game weirdness and bugs kill my ship too often.
...But I did come up with a Broke Adder savegame start.
And I've yet to earn enough money after start to afford a Cobra 3 with it, despite about 200 hyperspace jumps.

[szaumix]

I either quickly uninstalled or toned down aspects of many OXPs precisely because of the changes they made to game balance and credit-earning.

Yeah, me too. I don't even have Commies (and therefore, Factions) any more because what the hell do I want all those dockables around for? If I can balance them all first, maybe. But free, meaningless money is invariably the historical reason I either started a new save with more "honor code" rules or just got the shits and abandoned Oolite in favor of real life for long a while.

Everything else you said can almost be summarized, "give up, Oolite economics is broken deal with it and it's every tweaker for himself." There's sense in that but all you're doing is pointing out theoretically solvable oxp convolution; I still think a "solve for x" solution is possible here along the lines of Skilled NPCs, where either costs or rewards are raised generally or by category.

[Switeck]

If you stick with a limited number of relatively unchanging OXPs and Oolite itself, balancing is possible.
Balancing with "N" OXPs that change the game's economy can go as high as 2^N interactions.

I had little trouble balancing Commies dockable stations relative other aspects of Oolite's economics (poor tho it be!) -- if you can't sell or buy much there and the prices are the same or worse than the main station, it is time-wise more trouble than it's worth to visit them unless you're near-desperate.

The danger of lots of stations in a system is that is part of the basis that populator uses for how many traders, police, and pirates to add...which can result in more and larger pirate groups.
...And Commie systems tend to be semi-low on the TL and economic scales, so they get a pretty big piracy boost from that as well.
Commie systems typically only gain 1-3 stations, most of which are astro-gulags, so it's not really the worst offender in that regard.

A bigger problem might happen in a high-tech confederacy with 5+ added stations...that's one of the reasons I looked into station exclusions -- if there's a Superhub, chances are a Sothis station is unnecessary.

[Cholmondely]

If you stick with a limited number of relatively unchanging OXPs and Oolite itself, balancing is possible.
Balancing with "N" OXPs that change the game's economy can go as high as 2^N interactions.

I had little trouble balancing Commies dockable stations relative other aspects of Oolite's economics (poor tho it be!) -- if you can't sell or buy much there and the prices are the same or worse than the main station, it is time-wise more trouble than it's worth to visit them unless you're near-desperate.

The danger of lots of stations in a system is that is part of the basis that populator uses for how many traders, police, and pirates to add...which can result in more and larger pirate groups.
...And Commie systems tend to be semi-low on the TL and economic scales, so they get a pretty big piracy boost from that as well.
Commie systems typically only gain 1-3 stations, most of which are astro-gulags, so it's not really the worst offender in that regard.

A bigger problem might happen in a high-tech confederacy with 5+ added stations...that's one of the reasons I looked into station exclusions -- if there's a Superhub, chances are a Sothis station is unnecessary.

Or, if you play with Stranger's World, when the factories are thrown into deep space away from the sun, and a 30 minutes real-time ride from the witchpoint, these things tend to self-balance. Even if someone comes up with a tweak to put them where they should be, it is still a long trip to get to them.

[Stormrider]

Really? I looked over [EliteWiki] Shipdata.plist many times in the last month... and peeked at the AI files, and the way the populator overwrites all police's SCAN_CLASS and a few other police details had me thinking my options were more or less bottlenecked. Like most things I wish I could do, it's beyond my current abilities.

You would have to have a script that checks the status of all ships in scanner range of the player, probably with shipAttackedOther, then use shipBeingAttacked to check if the player is doing the attacking, if so use Player.bounty to set the player's legal status. The Bounty System oxp already does a lot of the work of checking the status of ships so you might start by checking out those scripts.

The danger of lots of stations in a system is that is part of the basis that populator uses for how many traders, police, and pirates to add...which can result in more and larger pirate groups.

I believe setting the allegiance can help balance stations in this respect, although I have set Distilleries and Saloons to chaotic, so they likely make systems more dangerous.

[Switeck]

The danger of lots of stations in a system is that is part of the basis that populator uses for how many traders, police, and pirates to add...which can result in more and larger pirate groups.

I believe setting the allegiance can help balance stations in this respect, although I have set Distilleries and Saloons to chaotic, so they likely make systems more dangerous.

I also don't want a lot of redundant stations due to not making economic sense for them -- if they have competition, they likely get less traffic.