So I recently posted an idea where we could construct facilities I call Dojo's where players can train and become better at the game. But searching for this sort of facility in search, I found that only I had ever referred to them. Looking under the search term "training" I also found no such idea... It seems a suggestion has appeared, and I had wrongly assumed someone like flatfinger, cornflakes, or thymine came up with it long ago
Anyways, Dojo's. Josh has already hinted at the idea of being able to train workers, by having them do certain tasks while they statistically increase their skill levels. I am suggesting that the same sort of function should be applied to NPC's and the human player.
I would suspect that all NPC skills and intelligence will have some sort of optimal state for decision making, such as perfect accuracy, safe and quick pathfinding, optimal price beliefs, formation rigidity, etc. Most NPC's will also have some sort of pollutant in their code which artificially skews their perfect accuracy or optimal price belief so that NPC's have a wide variety of skills, ranging down from perfect, to horrible decision making. Training at a Dojo would however remove that pollutant, allowing their accuracy to increase, to more closely follow a fleet formation, to have a more optimal price belief, to be better at frequency identification, or any skill that can be increased. Additionally, training at a dojo could have a small effect on your personality, shifting you slightly towards the personality of the owner every time you use it.
For workers, training is done simply by doing something for a random amount of time, but for the more complex NPC's, this could be done by completing "Dojo Tasks" that would take place in specially designed testbeds. For practical purposes, it's largely the same if an NPC uses a dojo: they are temporarily removed from the game, and come back with a statistical increase in whatever skill they paid for and a very slight personality shift.
If the human player enters the training testbed, the below happens.
This is a training simulation, so even if LT has a dead is dead approach, it would not apply here. Failure of a Dojo task should instead cost a certain amount of credits (unless otherwise deactivated), and each player would be only willing to spend so many credits on training. If so, this would partially account for the statistical nature of success, but wealthier players could afford more failures, and so perhaps account balance would shift the statistical probability of skill increase higher.
Josh is already working on testbeds to create the actual ingame AI, but such testbeds could equally serve as scalable, semi-scripted tutorials for the human player. For each aspect of the game which the player can learn, there would be an initial, basic setup. For combat, it might be "kill this one super easy target" for mining it might be "Successfully mine 100 units of ore" for trading it might be "Find the good with the highest price difference at station A vs station B, and make a profit" For exploring it might be "Use your scanner to find the wormhole and travel through it" for piracy it might be "immobilize this ship without destroying it"
From there, the simulations could grow ever more advanced, where combat simulations start to incorporate smarter enemies, more advanced formations, when to call in reinforcements, semi-random environmental obstacles, semi-random objectives, fleet command, when to send in reinforcements, semi-randomized combat positions/numbers, non-hostiles in the combat zone, etc until you are throwing pretty much any combat situation that you can come across in LT at the player.
Mining simulations may grow from being mine X units of ore to "Find and mine X credits worth of ore" or setting up remote mining operations, hiring out your own miners, setting up a mining operation in a virgin system, dealing with setbacks to your project such as broken equipment or pirates, new competition from smaller, equally sized, or larger competitors, etc.
For traders/manufacturers it might grow to making x credits of profit on a simplified market, making x credits of profit on a dynamic market, delivering x units of goods within a certain time, creating goods of a given quality, hiring employees, playing the markets, managing competition for resources or employees, building infrastructure, building a small commercial empire, dealing with setbacks such as pirates, rogue employees, damaged infrastructure, doing business in a warzone, vertical integration, etc
Explorers could grow from learning how to use their scanner, to being able to isolate particular frequencies from the noise, discovering the most valuable parts of a virgin system, gathering information about hazardous zones/natural dangers, have competition come in and start reporting on your finds, scouting out hostile territory, gathering information while in the middle of a battle, gathering information while under attack by pirates/hostile forces.
For pirates, it is much like combat, except you tend to play in the opposite role, and your goals are more to destroy ships like fighters while only immobilizing freighters/barges to get the goods. But it would also be partially business oriented, because a pirate needs to turn around and re-invest their ill gotten gains into their continued survival or expansion. Pirates also need to learn how to hide from those who are hunting them, and how to stalk their prey until it's safe to strike.
Mini-LT, The ultimate tutorial would in fact be a tiny contained LT universe, with maybe a dozen systems where a variety of semi-randomized situations could be called up, and various dojo-tasks issued, depending on the particulars of what form of training is being requested. Your simulation may require you to have an account balance of 1 billion credits for economic training, or it may require you to destroy an opposing faction, or it may require you to go and steal 50 million credits worth of cargo, or it may require you to discover every discoverable object in the contained universe (which may be easy, or extremely difficult) Such simulations don't really matter on what form of training you are looking for, they just present you with a goal, that might not even be possible to accomplish within those constraints.
Dojos in the Game World
Dojo's as presented would make for an interesting addition to the game imo , but they raise some interesting issues and opportunities. Though the testbeds may be isolated from the unbroken game, their effects of increasing player skills are not. An idea to consider is how much of a net skill increase can a dojo provide to a region? This is important for LOD, as well as for the value/expense of creating and running a Dojo. If I can build and own Dojos, why not just build a dozen, and not charge my own subordinates to use them, so that I can have an army of highly skilled players working for me?
Should any random Dojo be able to teach any and all skills possible in LT, or should they have their own levels/required resources to train in a particular area? Should dojo's specialize in a specific form of training or be able to teach everything?
In reality, I would imagine such a place could teach anyone any skill they wanted, but for balance purposes, there needs to be some sort of limitation.
Should Dojo's require special resources to run a training simulation? If so, this special commodity which only dojo's consume will need a manufacturer, could dojos manufacture such simulations themselves? what are the required inputs? could specialized dojos sell this training simulation to other dojos? Would this resource have quality tiers?
As an initial thought, very open to criticism I would suggest
Dojo's cost money and resources to build
Dojos create specialized, consumable training resources(TR)*
Dojo's can sell their TR to anyone, but only dojos can use them to train players
TR costs resources and/or credits to produce
A Dojo owner can bite that production/purchase cost or they can charge for training as they so choose
If a Dojo has no TR, it cannot train NPC's
TR specialties focus on every aspect of LT where the AI is handicapped
NPC's with no handicap do not seek training in that area
...And lets just throw in there, that dojo's also train workers without TR, but it still costs credits (and maybe you could just pay it to upgrade your workers)
Perhaps I have a fundamentally flawed idea as to how the AI will be handicapped, and this wouldn't actually work. However, if the AI is being artificially handicapped with less effective algorithms, or lowered variables, this sort of mechanic could really add a lot to the game in that it could be a way the Skill of particular areas will be greater than those without. Perhaps Dojo's will be the sort of place only intellectual players/cultures will build, and will become icons of civilized space as apart from the wilds of space where skills are hard earned and not bought.
And on top of everything, it would offer an entirely optional tutorial for the player to use at their own discretion or ignore entirely.
So, thoughts? suggestions for improvement? alternatives?
Thu Oct 09, 2014 11:43 pm
Challenging your assumptions is good for your health, good for your business, and good for your future. Stay skeptical but never undervalue the importance of a new and unfamiliar perspective.
Beauty may not save the world, but it's the only thing that can
Beauty may not save the world, but it's the only thing that can