Return to “Announcements”

Post

Re: Development Update #12: December 2013

#49
RageQuitKetchup wrote:But of course graphics are not the only things I observed in December's dev update; the feature that excites me the most is that Limit Theory isn't some pre-scripted game with cheating NPC's and AI, but is instead a dynamic game where the NPC's and AI have to follow the same rules that the player has.

But my biggest dissappointment would be if the economy were like the X games where it's all "smoke and mirrors", with everything is spawned from nowhere and with NPC's and AI's that cheat.
I am quite curious about how the AI and the various simulations will work in practice. As I have mentioned elsewhere I have a great deal of respect for Josh's abilities, work ethic, and his vision for the game he wants Limit Theory to be. That said, it has already been revealed that the simulation won't be be uniform throughout the universe (which is only natural due to real-world hardware limitations). In fact, the simulation will get progressively more coarse the further away it is from the player. I am not certain what you consider cheating NPCs and AI, but doubtless, at its least granular, the simulation will likely be little more than vector resolution with an element of uncertainty (via stochastics) to keep developments from becoming entirely predictable.

As for the economic simulation, as far as I am aware, there is little that is certain beyond a universal currency and the existence of markets for trading. With a presumed focus on resource extraction, industrial production and trade of goods thus obtained or produced, I would expect the focus to be on microeconomics. I am not certain if there are any plans for aggregating supply and demand into a universal macroeconomic model (or how that could even be achieved in an infinite universe only a fraction of which would be simulated at any given time). At the same time, for larger economic or political factions you would want a macroeconomic simulation through which to present a functional backdrop for said entities.
Image Which raises the question of just how much processing bandwidth you want to devote to economics simulation. Doubtless, players' opinions will vary according to what they are looking to get out of playing Limit Theory. Regardless, with finite resources and given a simulation with variable granularity there will always be aspects of the game that could be accused of utilizing "smoke and mirrors". I think it only sensible to temper expectations to somewhere within sight of what may be realistic.

In the end, most players will judge Limit Theory based on their overall experience with the game.
I know not what life is, nor death.
Year in year out-all but a dream.
Both Heaven and Hell are left behind;
I stand in the moonlit dawn,
Free from clouds of attachment.
Post

Re: Development Update #12: December 2013

#50
Tom wrote:As for the economic simulation, as far as I am aware, there is little that is certain beyond a universal currency and the existence of markets for trading. With a presumed focus on resource extraction, industrial production and trade of goods thus obtained or produced, I would expect the focus to be on microeconomics. I am not certain if there are any plans for aggregating supply and demand into a universal macroeconomic model (or how that could even be achieved in an infinite universe only a fraction of which would be simulated at any given time). At the same time, for larger economic or political factions you would want a macroeconomic simulation through which to present a functional backdrop for said entities.
Image
If the LT universe is represented at the macroeconomic level, it raises some additional questions for me. Will there be any kind of monetary policy committee (e.g. such as the UK's Bank of England) that controls things like interest rates or employ measures like quantitative easing? These kind of things could lead to changes in the pattern of expenditures of NPC's or the quality of the economy on planets. They could also lead to multiplier effects, since an increase in aggregate demand would feed back into the circular flow of income and increase national income and expenditure by a greater proportion, producing further changes in the behaviour of NPC's and of planets (if their economies are modelled).

Also:

If I as a player become significantly powerful and wealthy, with a personal wealth rivalling that of the entire economies of planets or solar systems, can I expend huge quantities of money in the form of capital investment e.g. by ordering the construction of a new jump gate that bridges two systems? What effect would this have on the economy? Not only might I increase the volume of trade flowing through the system, but I directly increase the wealth of all the people working on the jump gate. They then have more disposable/discretionary income with which to, for instance, put up contracts for other NPC's to do work for them. (this is how the multiplier effects arise in macroeconomics). So can I affect the number, quality and type of economic activity (such as contracting) occurring in a region of space by spending lots of money in it?

I'm guessing that the answer to these things are "no", but I thought I'd ask anyway. But at a more realistic level, will paying an NPC for work that he's done for me change his pattern of behaviour? Will it encourage him to post more contracts himself, now that he has more money?
Post

Re: Development Update #12: December 2013

#53
DWMagus wrote:In regards to the macroeconomic level, I'm still curious to see if/how this AI will handle goals that span more than a single sector.
I don't think the AI really cares where its goals are. Instead, I'm more curious to see if it will stay tethered to a single sector or faction, or if chasing their goals makes them travel the universe.
Beware of he who would deny you access to information, for in his heart he dreams himself your master.
Post

Re: Development Update #12: December 2013

#54
ThymineC wrote:If the LT universe is represented at the macroeconomic level, it raises some additional questions for me. Will there be any kind of monetary policy committee (e.g. such as the UK's Bank of England) that controls things like interest rates or employ measures like quantitative easing? These kind of things could lead to changes in the pattern of expenditures of NPC's or the quality of the economy on planets. They could also lead to multiplier effects, since an increase in aggregate demand would feed back into the circular flow of income and increase national income and expenditure by a greater proportion, producing further changes in the behaviour of NPC's and of planets (if their economies are modelled).
Since Limit Theory universes are essentially infinite, simulated economic models (that adhere to the idea of limited information and connectivity between economic entities) are by definition not possible on a universal scale.

As for money, for simplicity's sake, Josh decided to go with a universal currency. It is not a fiat currency, however. Universal monetary units are backed by a tangible asset(s). The idea of using a specie was raised but discarded due to the improbability of its universal acceptance and set value. I suggested utilizing currency issued by barter exchanges (with the available currency on a given exchange directly correlated to the goods and services offered on it). Alas, depending on just how laissez-faire the barter exchange simulation would be, wealthy players would be able to corner and manipulate markets, potentially in a game-breaking way, though this is likely true with most potential currency/market models.

The focus, as I see it, should be on creating a sound microeconomic simulation. The key to that, in turn, is properly determining elasticity, especially as it concerns substitution and its effect on demand (see cross elasticity of demand). Creating a sound formula for the elasticity of supply is less pressing as it ties into aspects of the game which will get plenty of attention as is (raw materials, production, inventory, etc...).

As concerns macroeconomic modeling one could correlate based on variables such as output, employment, consumption, investment, etc... aggregated from the microeconomic simulations into some sort of coherent equilibrium state. Given the breadth of the game it is my sense that it might make more sense to work backwards and employ dynamic stochastic general equilibrium modeling.
From Wikipedia:
Like other general equilibrium models, DSGE models aim to describe the behavior of the economy as a whole by analyzing the interaction of many microeconomic decisions. The decisions considered in most DSGE models correspond to some of the main quantities studied in macroeconomics, such as consumption, saving, investment, and labor supply and labor demand. The decision-makers in the model, often called 'agents', may include households, business firms, and possibly others, such as governments or central banks.

Furthermore, as their name indicates, DSGE models are dynamic, studying how the economy evolves over time. They are also stochastic, taking into account the fact that the economy is affected by random shocks such as technological change, fluctuations in the price of oil, or changes in macroeconomic policy-making. This contrasts with the static models studied in Walrasian general equilibrium theory, applied general equilibrium models and some computable general equilibrium models.

For a coherent description of the macroeconomy, DSGE models must spell out the following economic 'ingredients'.

Preferences: the objectives of the agents in the economy must be specified. For example, households might be assumed to maximize a utility function over consumption and labor effort. Firms might be assumed to maximize profits.
Technology: the productive capacity of the agents in the economy must be specified. For example, firms might be assumed to have a production function, specifying the amount of goods produced, depending on the amount of labor, capital and other inputs they employ. Technological constraints on agents' decisions might also include costs of adjusting their capital stocks, their employment relations, or the prices of their products.
Institutional framework: the institutional constraints governing economic interactions must be specified. In many DSGE models, this might just mean that agents must obey some exogenously imposed budget constraints, and that prices are assumed to adjust until markets clear. It might also mean specifying the rules of monetary and fiscal policy, or even how policy rules and budget constraints change depending on a political process.

Traditional macroeconometric forecasting models used by central banks in the 1970s, and even today, estimated the dynamic correlations between prices and quantities in different sectors of the economy, and often included thousands of variables. Since DSGE models start from microeconomic principles of constrained decision-making, instead of just taking as given observed correlations, they are technically more difficult to solve and analyze. Therefore they usually abstract from so many sectoral details, and include far fewer variables: just a few variables in theoretical DSGE papers, or on the order of a hundred variables in the experimental DSGE forecasting models now being constructed by central banks. What DSGE models give up in sectoral detail, they attempt to make up in logical consistency.
At any rate, I have already created an economics discussion thread some time ago, even if I have more or less abandoned it due to personal distractions as well as a dearth of new development information relating to economic simulation/modeling in the project.
I know not what life is, nor death.
Year in year out-all but a dream.
Both Heaven and Hell are left behind;
I stand in the moonlit dawn,
Free from clouds of attachment.
Post

Re: Development Update #12: December 2013

#55
Katorone wrote:
DWMagus wrote:In regards to the macroeconomic level, I'm still curious to see if/how this AI will handle goals that span more than a single sector.
I don't think the AI really cares where its goals are. Instead, I'm more curious to see if it will stay tethered to a single sector or faction, or if chasing their goals makes them travel the universe.
That's exactly what I was trying to say. Whether or not the AI can look at "The big picture".
Image
Early Spring - 1055: Well, I made it to Boatmurdered, and my initial impressions can be set forth in three words: What. The. F*ck.
Post

Re: Development Update #12: December 2013

#56
I've just posted a thread over in Suggestions which would make it so you could have infinite/near-infinite variations of the different kinds of components and systems in the game (based on ideas that Flatfingers came up with), and some of these variants would be naturally better (in some or all respects) than others. So I was thinking, "How should the value or worth of a component be calculated? Should it be hard-coded in as a formula in the object templates?".

Then I realised there shouldn't be any need to be that artificial about it: say an aggressive NPC wanted to increase the ability of its vessel to dish out damage during combat. It should be able to understand that a variant of a plasma cannon that has +20% DPS relative to a standard plasma cannon is more desirable than one that has -13% DPS relative to the standard. It should then naturally demand the first variant more, ceteris paribus. Because the demand for the first variant will be higher than the second, this will cause demand to outstrip supply, sending a signal to the suppliers of the first variant to increase the selling price, which encourages more suppliers to enter the market and limits demand until a new equilibrium level is reached (since the second variant is a substitute of the first, it should have positive cross-price elasticity of demand and therefore the rising price of the better variant should attract a little more demand back over to the worse variant too, right?)

In this case, a "better" variant will naturally assume more worth than a worse variant if NPC's are intelligent enough to recognise that better variants will help them achieve their goals better, since there will be more demand for them, and hence a higher price. No need for artificial hard-coded formulas. Sweet!

I can't wait to see how the microeconomics plays out in this game.

It's also interesting because being having components vary in different aspects from one another will probably try to cause the NPC suppliers to pursue blueprints that produce components that are special in some respect - a unique selling proposition. This may naturally encourage different suppliers of a kind of product to specialise their products towards being best in a particular area. In the shields market, one supplier may try to sell shields that tend to have high maximum integrity, while another may try to tailor their shield systems to have fast recharge rates, and yet another may try to advertise based on low prices. You get corporations with a distinct flavour to their products this way - like the different corporations in Borderlands (Torgue, Hyperion, Jakobs, Maliwan, etc.).

Also, since this variance in properties doesn't really apply to raw materials - one unit of axium or bexium is just the same as any other - I'd expect to see a kind of perfectly competitive market appear here, with all NPC's supplying homogenous products at equal prices and making normal profits (particularly in the more crowded systems).

None of this should need to be coded explicitly - these things should just emerge if Josh makes the AI clever enough, I guess.
Post

Re: Development Update #12: December 2013

#57
ThymineC wrote:say an aggressive NPC wanted to increase the ability of its vessel to dish out damage during combat. It should be able to understand that a variant of a plasma cannon that has +20% DPS relative to a standard plasma cannon is more desirable than one that has -13% DPS relative to the standard. It should then naturally demand the first variant more, ceteris paribus.
I'm going to dig into your new concept piece in just a bit, but first I wanted to note here that -- last I heard -- Josh was planning for modifier technologies (which produce installable objects) to have both positive and negative effects.

I've been a fan of this kind of design for many years. It is a thing of great beauty because the instant you implement it, you get much more complex (and therefore interesting) decision-making gameplay opportunities.

Instead of always going without thought for whatever thing has the larger number, giving objects both benefits and costs means you need to weigh the value of a modifier object against a higher-level satisfaction metric than "urrr, BIGGER!" A plasma cannon with +20% DPS is awesome; a plasma cannon with +20% DPS and +20% power utilization is not so obviously a thing you want. Now you have to decide whether the damage advantage fits into your preferred use for that vessel despite the increased power usage, or if the cost is too high and some other modifier object is better for you. (And by "you," I mean both human players and NPCs.)

I think that actually offer more support for your idea of NPC factions having preferences for the kinds of modifiers they prefer, though. It intensifies the value of certain kinds of techs because now they have costs that have to be considered as well -- basically, you'll have to make a choice of specialization to maximize the efficiency of your ships. (And what rational person wouldn't do that? Well, OK, I guess irrational NPCs might not. Doom on you, NPC.)

OK, more things to go read now. :)
Post

Re: Development Update #12: December 2013

#58
ThymineC wrote:I've just posted a thread over in Suggestions which would make it so you could have infinite/near-infinite variations of the different kinds of components and systems in the game (based on ideas that Flatfingers came up with), and some of these variants would be naturally better (in some or all respects) than others. So I was thinking, "How should the value or worth of a component be calculated? Should it be hard-coded in as a formula in the object templates?".

Then I realised there shouldn't be any need to be that artificial about it: say an aggressive NPC wanted to increase the ability of its vessel to dish out damage during combat. It should be able to understand that a variant of a plasma cannon that has +20% DPS relative to a standard plasma cannon is more desirable than one that has -13% DPS relative to the standard. It should then naturally demand the first variant more, ceteris paribus. Because the demand for the first variant will be higher than the second, this will cause demand to outstrip supply, sending a signal to the suppliers of the first variant to increase the selling price, which encourages more suppliers to enter the market and limits demand until a new equilibrium level is reached (since the second variant is a substitute of the first, it should have positive cross-price elasticity of demand and therefore the rising price of the better variant should attract a little more demand back over to the worse variant too, right?)

In this case, a "better" variant will naturally assume more worth than a worse variant if NPC's are intelligent enough to recognise that better variants will help them achieve their goals better, since there will be more demand for them, and hence a higher price. No need for artificial hard-coded formulas. Sweet!

I can't wait to see how the microeconomics plays out in this game.

It's also interesting because being having components vary in different aspects from one another will probably try to cause the NPC suppliers to pursue blueprints that produce components that are special in some respect - a unique selling proposition. This may naturally encourage different suppliers of a kind of product to specialise their products towards being best in a particular area. In the shields market, one supplier may try to sell shields that tend to have high maximum integrity, while another may try to tailor their shield systems to have fast recharge rates, and yet another may try to advertise based on low prices. You get corporations with a distinct flavour to their products this way - like the different corporations in Borderlands (Torgue, Hyperion, Jakobs, Maliwan, etc.).

Also, since this variance in properties doesn't really apply to raw materials - one unit of axium or bexium is just the same as any other - I'd expect to see a kind of perfectly competitive market appear here, with all NPC's supplying homogenous products at equal prices and making normal profits (particularly in the more crowded systems).

None of this should need to be coded explicitly - these things should just emerge if Josh makes the AI clever enough, I guess.
First of all let me note that your prolific contributions to the forums are worthy of praise, ThymineC.

As concerns your suggestion that emergent gameplay (based on NPC AI) could or should replace direct modeling in the economic simulation... well, it is brilliant except for the fact that Limit Theory is a virtually infinite sandbox game the primary purpose of which is to provide entertainment as a space-based simulation. I seriously doubt that Josh desires or is willing to put in the time and resources necessary to seamlessly aggregate AI-based economic behavior into an integrated whole from personal to micro to macro.

It seems to me that it is more sensible to work backward using a DSGE (dynamic stochastic general equilibrium) model(s) that also defines the microeconomic simulation. The AI then merely needs to define the unique values for the economic ingredients held by the decision-makers (or agents) to plug into the model. In this way one can still infuse economic entities with personality (or to be more precise, preferences) while not relying on highly compex and costly vertical aggregation.
I know not what life is, nor death.
Year in year out-all but a dream.
Both Heaven and Hell are left behind;
I stand in the moonlit dawn,
Free from clouds of attachment.
Post

Re: Development Update #12: December 2013

#59
many Joshs died to bring you this update..
May they rest in peace! Their sacrifice was not in vain, the results are worth it!

Your dynamic UI is a fantastic feature! I was thinking of something similar, your setup gave me a few more points about functionality and modularity of such a system. Very well done!

The AI "delegation" is an interesting feature, that will allow the universe to be evolving even if the player does not interferes. How many NPC do you plan to have in a game? And will they always be active? What will be happening in other systems? Will NPC be working in the background? What happens when a contract is not taken? The NPC will just take it and the mission disapear(as in "I'm doing it myself now..." says npc XX21)? Will there be dialogue generated by the missions/activities npc does? For instance, the npc can have a record of events it did during it's carrier, and have summaries for these events, topics, remarks, or even alternate missions like revenge/rescue an enemy/friend for something that happened! So you may have caused harm to an NPC, if it escapes, it will go on a revamping, training, upgrading mission and then look for you and try to get its revenge, or the contrary, save you from a foe because you helped it earlier!

In short it's a dynamic Friend or Foe Interface.

Graphics are great (sorry, I don't pay that much attention to graphic right now [might be jealousy]) I really appreciate the hard work and the Josh's you sacrificed to the GFX god.
Keep it up, 'll keep my eye on your project, and obviously take advantage of your expertise!
Post

Re: Development Update #12: December 2013

#60
Hey guys,
I just wanted to say to Josh that this update (like all the others) looks fantastic! I've been following the project for many months; everything you're doing looks awesome, and as a programmer I look forward to every dev log you put out! I'm really sorry I missed the KS; I found out about it literally the day after it ended! But I'll definitely be buying the game day one.

I had a teensy question: I noticed in the update video that there seemed to be a bit of a "fish-eye" effect going on with the FOV which bothered me a bit, but I know that sort of thing is up to personal preference. My question is: will FOV and stuff like that be customisable in-game?

Once again, keep it up! LT is going to be a highlight of 2014!
Image

Online Now

Users browsing this forum: No registered users and 1 guest

cron