With the subject of delegation coming up this past December, I started thinking about the concept of "fairness" when it comes to contracts and compensation. Josh touched on this in his dev logs, but I wonder if there will/should be some kind of negotiation mechanic.

For example, if the player posts a contract for destroying a whole fleet for a total of 1 credit, it's reasonable to expect nobody to take this contract. But how shall the player know what is a good reward and what isn't? Perhaps NPCs should provide feedback upon rejecting (or accepting) a contract, along the lines of "much less than acceptable" or "better than expected" or even "too difficult". Such feedback might also be useful for an NPC employer who might need to adjust the reward for a contract if it is to have any chance of being accepted. Certainly as a player I would like to have some kind of idea about whether my contract has any chance of attracting a taker.

This ties in with the notion of value in the LT universe, about which I'm sure a lot of words have already been written. If anyone can point me to relevant reading, I'd be much obliged.

Will contracts have a initial fee to prevent someone from taking every contract?

Will contracts have a time constraint so that a NPC calling for help can make a new contract if the one who took the contract didn't fullfil it? (had that example in my mind where the NPC calls for repairing help)

I'm all for being able to post silly contracts. I doubt any NPC will take them. They should also be smart enough to only post contracts that are near the market value.

Take a look at warehouses in MMO's. If you look for a popular item you'll see a lot of people trying to buy it for "cheap", a lot of people trying to sell it for "expensive", but the majority will be around the fair market price. This is something that evolves naturally.

About contracts, I wouldn't mind having to pay a brokering deal to be able to post a contract, and how it's spread in neighbouring systems.

Let the player post 1cr fleet destroying contracts. The NPCs should be able to start become weary of you if you do it constantly that you are of the type to not offer adequate compensation.

But also let the AI do it too. The AI afterall may be shady and trying to skimp the player too.

Lemar wrote:Will contracts have a initial fee to prevent someone from taking every contract?

Will contracts have a time constraint so that a NPC calling for help can make a new contract if the one who took the contract didn't fullfil it? (had that example in my mind where the NPC calls for repairing help)

You answered your own question. If the player wants to be underhanded and take all contracts but not do them? It should make their rep with those NPCs/Factions go down. If you want to be a dick, they'll treat you as such and may not recognize your acceptance of contracts in the future.

But might as well let the NPC repost the contract multiple times if they want. Maybe they're trying to play both sides of the war or stage a war themselves.

DWMagus wrote:Let the player post 1cr fleet destroying contracts. The NPCs should be able to start become weary of you if you do it constantly that you are of the type to not offer adequate compensation.

Yeah, I never meant to imply that I think the player SHOULDN'T be able to post ridiculous contracts, just that there be some feedback that gives you some idea what "adequate compensation" means, as opposed to finding out through endless trial and error.

But might as well let the NPC repost the contract multiple times if they want. Maybe they're trying to play both sides of the war or stage a war themselves.

Heh, I really like that idea.
Mission: Stake out asteroid X for a while and attack anyone who gets close.
*posts 5 times*

It was almost exactly ten years ago that I designed what still seems like a pretty reasonable general contract system.

It addresses several of the questions raised here, plus a few others. For example, it includes a penalty feature that gets activated if you agree to take a contract but fail to fulfill it. This would minimize characters just grabbing every contract. It can also count contracts taken and their results, which could indeed feed into the "faith" component of Josh's delegation AI.

What I called Open Contracts would in LT be missions posted publicly for any character to take. The Personal Contracts, if mandatory for subordinates, would be equivalent in Limit Theory to delegated goals -- an offer you can't refuse.

Dagnabbit, it may take me another ten years but I'm going to find some way to get these ideas implemented!

And it looks like maybe I will get to see some of those ideas implemented.

There's some good stuff on contracts in today's devlog!

Josh Parnell wrote:when there exists no supply, how does the initial demand come about? Basically, what one needs is a mechanism for understanding demand that could exist but does not yet due to lack of supply.

I'm probably oversimplifying (missing necessary constraints), but isn't this mostly addressed by specifying general types of contract resolution criteria, rather than particular things?

For example, suppose the goal is "I need a ship weapon that's good against phase shields." Rather than issuing a production contract for a "Wayne Mk III Heat Ray," a company would issue a contract for a "Wayne Heat Ray" or a "third-generation heat ray" or even just a "heat ray."

The option of creating a contract that is for a general class of thing, instead of for a particular kind of thing, means that the quality of the result can be traded off for increasing the odds of satisfying the true need. A more general contract is more likely to be satisfied, at the risk of a lower level of satisfaction. If only the best will do, and time and money are no object, you can always write a very specific contract.

Allowing contracts for general types of things doesn't directly do anything to stimulate supply. But it does make starting up an economy easier by allowing many more kinds of things to successfully satisfy a demand. That primes the pump for allowing ever-more-specific demands to be attempted.

Josh Parnell wrote:From a company's perspective, I'm not totally sure when one would try to issue one type vs. the other [of exclusive or non-exclusive contract], but I'm sure there are some advantages / disadvantages of each to think about.

The two main criteria that come to mind are parallelization and time-boundedness.

Say I have a goal I want to satisfy through contracts, and that an exclusive contract (an order from a superior in an organization to a subordinate) is the default. If that goal benefits from being split into numerous identical tasks among multiple people, that increases the value of making it non-exclusive. If on the other hand that goal benefits from being (or must be) completed as soon as possible (if the outcome must be guaranteed, as when another urgent task depends on satisfying this goal), then that instead increases the value of specifying a contract as exclusive. If it can be resolved at any time, or continuously, then it's OK to make that a non-exclusive contract (to reserve your employees for contracts that have to be exclusive).

So if you want a lot of people doing the same thing, and they can complete those tasks at any time, then a non-exclusive contract is desirable. An example might be a mining contract to supply a refinery with ore -- as long as you can afford to pay them, you want multiple miners bringing you ore on an ongoing basis.

(This is why in my Player Contracts system I designed contracts to be either one-time or, with suitable conditions for termination, recurring. A recurring contract might be limited to the first 10 takers, or would stop after paying out X amount of cash.)

Parallelizability (multiple people doing the same thing) and time-boundedness (need for guaranteed resolution) are my first cut at how to decide whether a contract should be assigned to members of an organization or put out on the open market: more parallelizable and less-timebound = non-exclusive.

Are there other criteria that would be more effective at letting NPCs make this decision about the type of contract to offer?

Finally, a minor question:

Josh Parnell wrote:Process trees make so much sense for high-level, planning trees make so much sense for low-level...seems like we've got it all covered!

Am I not understanding, or is this backwards?

I imagine planning as a high-level AI activity (what to do) and process-connecting as a medium-level activity (how to do it). (Low-level would be physically doing something.)

Is this just a semantic thing, or is the latest conception of process trees truly a high-level AI function?

[Edited to correct and clarify non-time-boundedness as supporting non-exclusive contracts.]

You could use general contracts to jump-start supply as well. An NPC possesses a cargo ship, a mining ship, whatever - he can issue a general contract that specifies his availability for tasks. "Hire me for 20 creds an hour to haul up to 3k cubic meters" or "Hire me for 30 creds an hour to mine any asteroid." These general contracts may have to lowball the rates, but once one contract is snapped up, the market can begin to adjust towards equilibrium.

One of the great things about reading the dev diary is watching Josh build a universe from first principles. Today was a great case in point - exclusive and non-exclusive contracts are otherwise known as bilateral and unilateral contracts and there is a heap of legal theory that has been devoted to examining the two.

My first post to these forums was basically this same idea. I have reposted below for easy reading.

Perhaps this has already been suggested/confirmed/shot-down, but I would like to see contracts as the primary way to interact with NPCs. For ex:
NPC1 is part of a mining corporation and needs the refined substance transported back to Space Station1. You just so happen to be the only transport ship within range of the space station and NPC1 offers you a contract with something like the following:
1. You will transport 1000 cubic meters of material X to Space Station1
2. You will be paid 15% of fair market value for the material you are transporting.
3. If you are attacked by pirates you will pay 70% of fair market value for any material you lost.
4. You have 1 in-game week to complete this job.
5. If you cancel this job after accepting its terms you will pay 15,000cr
Do you:
[Accept] [Decline] [Counter-offer]

If you counter-offer you can change the values of the above terms, so in effect you are bargaining for the terms of you quest. For example you might change the terms to the folowing:
1. You will transport 1000 cubic meters of material X to Space Station1
2. You will be paid 25% of fair market value for the material you are transporting.
3. If you are attacked by pirates you will pay 60% of fair market value for any material you lost.
4. You have 1 in-game week to complete this job.
5. If you cancel this job after accepting its terms you will pay 20,000cr

And then the NPC can accept, decline or counter-offer. I know this may not appeal to everyone (unless I'm wrong) so you make this a toggle and call it "Advanced Questing" in the options menu.

See: this link for my original post.

pulsar9 wrote:[...]

Looks good but I seriously don't want to see anything saying "...in-game..." in the actual game. Immersion, bro.

Of course. I was just being descriptive. If I had worded it the other way I could see people telling me that it should be and in-game week and should not be dependent on real life time.

pulsar9 wrote:Of course. I was just being descriptive. If I had worded it the other way I could see people telling me that it should be and in-game week and should not be dependent on real life time.

Yeah I guessed so, just wanted to make sure.

Potentially relevant: a discussion of the concept of Trust Cycles

Imagine that some trustees then behave badly and don't return money. Investors who trust will then lose money. In evolutionary terms, they'll be selected against and will die out. Trust will then decline. And it might not be rebuilt merely by state regulation (pdf) of trustees. If it declines to zero, no investor will hand anything over to trustees, so trustees will end up with zero return as well.

But when trust is low, the trustee who can somehow signal his trustworthiness will attract more investors than his untrustworthy competitors. As an example, think of those 19th century food manufacturers such as Bird's, Colman's or Cadbury's who invested in brand reputation to signal that their products were less adulterated than their rivals'.

Trustworthiness will then be selected for, and so will spread. And as this happens, investors who trust will earn bigger returns, and so selection instead of favouring the untrusting will now favour the trusting. Both trust and trustworthiness will then rise. But when trust is very high and investors are handing over big money, trustees will have stronger incentives to run off with the money.

So, we get cycles of rising and falling trust and trustworthiness.

This fits in with the heterodox economic view that the idea of "equilibrium" is a non-starter. Rather than seeing the economy as being an iteration of linear systems tending towards a fixed equilibrium, that it's a dynamic chaotic system and "equilibria" are merely "attractors" around which the economic conditions orbit. Hence we get rises and falls in prices, employment &c. &c.

One thing that comes to mind thinking out loud about all this, is how much truth should be hidden from the actors? How imperfect should their knowledge be allowed to get? Or, in other words, what is an agent's maximum exposure to fraud and deception?

Obviously a model of a 19th century free for all "Gold Rush" economy will result in a bonus to shysters and tricksters willing to debase their reputation in exchange for easy profits. Like quack doctors and snake oil salesmen, reputation doesn't matter if you're gone by morning.

On the other hand, it's simply no fun to get screwed over.

So, what kind of things should agents in the system be able to lie about?

McDuff wrote:how much truth should be hidden from the actors? How imperfect should their knowledge be allowed to get? Or, in other words, what is an agent's maximum exposure to fraud and deception?
...
what kind of things should agents in the system be able to lie about?

This is a great question that deserves its own thread.

I can think of a few kinds of lie that could have serious consequences:

1. False product information. This is why MMORPGs invented the Secure Trade Window: the system makes both parties put what they're offering into a box, and once an offer is made, it can't be changed -- the only options are to agree to the deal (based on what you can see is in the box) or cancel the deal.

Should LT have any kind of equivalent of the Secure Trade Window? Or should it be caveat emptor all the way? Is there anything in between those two extremes?

As suggested (and as I've discussed before), being able to reliably recognize the other party is crucial to allowing cooperative characters to refuse to deal a second time with cheaters. So how well will that work in LT?

Which brings me to:

2. False factional membership. This one lends itself to various kinds of screwage, as in committing some negative act and making it look like some other faction did it.

3. False intentions. In this case, a character has goals they don't actually want to carry out themselves, but instead passes them to another character. This might be handy for nefarious types looking for a catspaw to take the heat for some bad behavior if caught.

So that's lying about what you have, who you are, and what you want. Any others? (This really does need its own thread.)