I wanted to get some feedback on the research system I'm using in a game I'm developing. Research is divided in into projects; there are three kinds of projects. Theoretical Application - Another application for a theoretical project beyond the intended outcome. Improvement - improves a completed theoretical project. Each Theoretical project is divided into three phases Initial Experimental Complete. Each phase may have 1 or more application projects, and once the complete phase has been researched then improvement become available. New projects become available when the required theoretical projects and improvement projects have been completed. How research works is like this. There are 5 scientific fields Chemistry Biology Virology Genetics Behavioural Sciences Researchers all have a rating between 0 and 10 in each of these fields representing their knowledge and skill with that subject. Each project/phase covers 1 or more of those fields, and has two values for each field it covers. Threshold – which represents the target number for success tests. Required Successes – which represents the required number of success. In order to complete a project a number of successes equal to the required number of success in each field must be obtained. Successes are cumulative so if you generated 1 Chemistry success yesterday and 10 today that would mean a total of 11 had been obtained. There is also a researcher skill call Data Analysis, each success test in a project generates 1 MB of data; also monitoring experiments would generate a number of MB of data. Then based on the researchers data analysis skill and the amount of data collected, bonus success would be generated for that project and related projects these successes would be divided among the fields in the project with the least number of successes. If there is insufficient computer storage to hold the new data the project is suspended until additional hardware is acquired. Data can also be obtained by having that tactical department raid other facilities and steal the data. Information services department can provided data though covert actions. Personal Basically there are three kinds of personal in the research deparment. Head Resarcher - 1 required - Is in charge of all research at a facility. Provides a bonous to all projects being research there based on their stats and skills. Lead Researcher - 1 lead researcher is required for each project you wish research beyond the first. So 1 lead reacher means you can research two projects at a time. Reseacher - personal assigned to work on projects. You can either hirer or promote personal to the first two postions. promotion to Head Reseacher provides a loyalty boost and raises their salary $10,000 a week, same for lead researcher but with salary increase of $5,000 a week instead. Demotions lower salary and loyalty by the same amount. Applications Currently my plan for applications is to have a random chance involved to determine whether or not the player can research the application project when they complete the required phase of the corresponding theoretical project. This means that the applications a player can get in a game are random and they may not gain any. Applications provided numerous advantages such as sources of legitimate income and access to things that would otherwise be unavailable. Do people see any problems with this? Would it better to remove the random chance so they always get the applications? Or does is make the game better since the player would have to adjust their strategies each time the play based on what applications they gain. One idea I had was to add specialties to researchers, if a researcher on a project had the required speciality then the player would be able to research that application. Lastly should I make some theoretical research require specific applications, this would mean that some branches of research would not be available to the player if they didn’t gain certain application projects. Experiments In order to research past the initial phase of a theoretical project, requires experiments. These experiments generate data, they also inform you of the effects and effectiveness of the project so far. The Lead researcher on the project will inform you when an experiment needs to be performed you can then choose which type of experiment to perform. Volunteer: Performs the experiment at one of your facilities with a few people obtained by your acquisitions department. Provides the least risk of exposure, highest internal risk, and least amount of data. Controlled: Performs the experiment at an independent location in a controlled environment with small group of people obtained through acquisitions. Provides low risk of exposure, moderate internal risk and moderate amount of data. Public: Performs the experiment on a large group of people in the general public. Provides highest risk of exposure, lowest internal risk, and the largest amount of data. Thoughts, Comments, opinions? ----------------------------------------------------- "Fate and Destiny only give you the opportunity the rest you have to do on your own." Current Design project: Ambitions Slave [edited by - TechnoGoth on May 7, 2004 10:59:25 AM]

quote:Original post by TechnoGoth
I wanted to get some feedback on the research system I''m using in a game I''m developing.

Hey!!! What happened to the Samurai?!?!?! (Although, Cyber-Samurai black ops who raid other corporations for research data could be way cool!)

quote:
Theoretical
Application - Another application for a theoretical project beyond the intended outcome.
Improvement - improves a completed theoretical project.

So this is like a tech tree where you have to research theoretical before you get to application, but unlike a techtree in that you can refine new discoveries again and again. This is cool, except that you''ll have to curtail the refinement so that it doesn''t excede the levels beyond it. For instance, if you can refine "bioadaptive neural toxins" to the point where they are more lethal than toxins down the same line of research, you don''t need to advance the tech tree.

quote:
Each Theoretical project is divided into three phases
Initial
Experimental
Complete.


Each phase may have 1 or more application projects, and once the complete phase has been researched then improvement become available.

Any point to just researching just theoretical, say continuing to refine experiments again and again?

quote:
Each project/phase covers 1 or more of those fields, and has two values for each field it covers.
Threshold – which represents the target number for success tests.
Required Successes – which represents the required number of success.

Recommend a dose of statistics here so you''ll be able to determine how likely some paths are. Will the player see the odds, by the way? Again, you''ll need to use stats to simplify, or hope they can make an intuitive guess out of "9 tests at 73%"

quote:
There is also a researcher skill call Data Analysis, each success test in a project generates 1 MB of data; also monitoring experiments would generate a number of MB of data. Then based on the researchers data analysis skill and the amount of data collected, bonus success would be generated for that project and related projects these successes would be divided among the fields in the project with the least number of successes.

This I like because it''s really research points. You could even get a bit more fancy with this, and have MB that correspond to your sciences, and then equipment that was required to store it. Instead of just generic HD space, you''d need autoclaves, centrifuges, electronic micrographs, etc. as GENERATORS of the MB; then you''d use the computer storage as a tank to hold all these points, and maybe even have different systems for backup / protection, with different security levels. Maybe you put all your viriology stuff behind laser grids and guard it by the Red Queen. This can also help protect you from data thieves and nosy activists within your own company.

quote:
If there is insufficient computer storage to hold the new data the project is suspended until additional hardware is acquired.

Good.

quote:
Data can also be obtained by having that tactical department raid other facilities and steal the data. Information services department can provided data though covert actions.

This is also cool, but again I think the data shouldn''t be generic MB, and that there should be different sites to raid that have different tradeoffs. For example, maybe Terragen Corp has great virology, but a nasty spec ops team guarding it; whereas Immunicorp has more generic viral data, but is easier to get at.

You could add another level to this, too: Diplomacy and alliances among corps, so that raiding is more substantial. For instance, you could raid USAMRID for their last remaining bubonic plague samples, but you''d bring down the might of the USA on you. If there''s more nuance, then maybe you could bribe someone for a sample, or compromise them. This could be handled by a covert spy network that simply presents opportunities with a price tag, success chance, and blowback cost. They covert ops department would present more opportunities the better funded they were, simple as that if you like.

quote:
Currently my plan for applications is to have a random chance involved to determine whether or not the player can research the application project when they complete the required phase of the corresponding theoretical project. This means that the applications a player can get in a game are random and they may not gain any. Applications provided numerous advantages such as sources of legitimate income and access to things that would otherwise be unavailable. Do people see any problems with this?

There''s no problem with randomization that I see as long as the computer is bound by the same rules, but know that it will kill some of the strategic flavor of the game. For that same reason, I highly advice against a dead-end research effect. Researching requires effort from the player, there should be some reward. If you want to make it more like gambling, consider making some rewards better than others.

quote:
Or does is make the game better since the player would have to adjust their strategies each time the play based on what applications they gain.

Master of Orion 3 had a "research is blind" game mechanic where you couldn''t choose your path unless you had a special ability. Maybe try that approach? Again, it kills strategy to some degree, so you have to ask yourself to what extent you want to draw strategy players.

quote:
One idea I had was to add specialties to researchers, if a researcher on a project had the required speciality then the player would be able to research that application.

If the player knows it''s coming up in advance, they can plan for it, but planning after the fact would suck. ("As it, dammit, I need epidural bioarmor, and now I need to get a dude that has the dermatologist speciality... and I just spent my budget on hard disks...")

quote:
Lastly should I make some theoretical research require specific applications, this would mean that some branches of research would not be available to the player if they didn’t gain certain application projects.

No, I think pure science should require hardware to run experiments more than specific applications. Think of the atom splitting or cosmology stuff we''re researching, which has no practical application associated with them, just supercolliders and satellites.

quote:
In order to research past the initial phase of a theoretical project, requires experiments. These experiments generate data, they also inform you of the effects and effectiveness of the project so far. The Lead researcher on the project will inform you when an experiment needs to be performed you can then choose which type of experiment to perform.

Is there any depth to a lead researcher, or is she/he just an expense + some skills?

quote:
Volunteer: Performs the experiment at one of your facilities with a few people obtained by your acquisitions department. Provides the least risk of exposure, highest internal risk, and least amount of data.

Controlled: Performs the experiment at an independent location in a controlled environment with small group of people obtained through acquisitions. Provides low risk of exposure, moderate internal risk and moderate amount of data.

Public: Performs the experiment on a large group of people in the general public. Provides highest risk of exposure, lowest internal risk, and the largest amount of data.

What is internal risk and what do you mean by exposure?

I think you have a couple of factors you can play with here:

* Stock value, coming from applications and deals you make with other rival companies
* Public reputation, which affects hiring price, availability of staff, and punishment level for misdeeds (slap on the wrist, citizen protest or Congressional investigation)
* Internal security, which is your defense PER FACILITY against hackers, saboteurs, thieves and spies AND experiments gone awry
* Danger level, which is how far you push the envelope, and results in a gamble of more reward for more risk
* Research time, which is how long an experiment takes

The types of experiments maybe should reflect these:
* Animal is the cheapest, but requires many many experiments and creatures to get any results. Slightly lowers rep and may result in raids by activists.

* Volunteer costs little, but takes longer and yields unclear results because you can''t push the envelope, meaning you get more random data. But it is more likely to raise your rep, and thus stock, because you''re seen as helping people.

* Public yeilds the clearest result, but this can trash cashflow from reputation and stock value, again depending on the danger level. If the experiment goes well, nobody knows, but if it doesn''t then the authorities step in and sieze facilities and force fines. Maybe the player, in a desperation move, can push the level of danger so high that if the win, it can catapult them ahead, but if they lose, it''s game over because the feds arrest everybody.

* Captive is fastest, most clear, largest amount of data, but is most expensive, requires most security. Simply can''t be discovered, or reputation is destroyed, so player has to terminate or bribe leaks. The data yeild more results the larger the sample, but the larger the sample, the bigger the facility needed and the higher the risk of escapes or mayhem.

If you had a loyalty system, captives would require the highest loyalty and the most ruthless scientists (who could have other tradeoffs, like defecting to other companies with your research data)

There could also be versions of captives with different moral costs
- Convicts: Easy to get, average health yeilds average data, but more volatile (likely to escape, cause deadly accidents)
- Urchins: Easy to get, substandard health yeilds poor data but can tolerate more dangerous experiments, more likely to cause leaks because of the moral cost
- Indigent: Easiest to get, little risk, but poor health yeilds poor data
- Nominal: Moderate difficult to get, average health, but more likely to attract investigators / infiltrators
- Elite: Difficult to get, excellent health yeilds excellent data, but powerful friends can damage relations with other companies if subject escapes
- Unique: Rare, difficult to get, unique makeup yields one full advancement on tech tree

--------------------
Just waiting for the mothership...

quote:Original post by Wavinator
Hey!!! What happened to the Samurai?!?!?! (Although, Cyber-Samurai black ops who raid other corporations for research data could be way cool!)

Its on hold for now, since I''m short on resources, I need a game that I could make within my means, this is seemed like a good one. Besides I''ve already got the interface framework built.

quote:
So this is like a tech tree where you have to research theoretical before you get to application, but unlike a techtree in that you can refine new discoveries again and again. This is cool, except that you''ll have to curtail the refinement so that it doesn''t excede the levels beyond it. For instance, if you can refine "bioadaptive neural toxins" to the point where they are more lethal than toxins down the same line of research, you don''t need to advance the tech tree.

There are limits to improvements and some time they take the form of new funcationality. But even a fully improved theory would be infieor in some capabilites to a similar theory further down in the tech tree.

quote:
Any point to just researching just theoretical, say continuing to refine experiments again and again?

Theoretical is the main form of research it is the hardest and longest to discover. the other two forms are just off chutes of varius theoretical projects. Not researching the other forms means that you not getting all you can out of your discoveries.

quote:
Recommend a dose of statistics here so you''ll be able to determine how likely some paths are. Will the player see the odds, by the way? Again, you''ll need to use stats to simplify, or hope they can make an intuitive guess out of "9 tests at 73%"

All paths can be research. I am using the threshold/success system because I thought it was better alternative then research points. It makes it more like actual research where you not sure how much time and effort will be required to complete the project. I''ll Include some kind of progress meter so that the player can gauge how close they are to completing a given project. But as to difficulty I think I''ll hide that.

quote:
This I like because it''s really research points. You could even get a bit more fancy with this, and have MB that correspond to your sciences, and then equipment that was required to store it. Instead of just generic HD space, you''d need autoclaves, centrifuges, electronic micrographs, etc. as GENERATORS of the MB; then you''d use the computer storage as a tank to hold all these points, and maybe even have different systems for backup / protection, with different security levels. Maybe you put all your viriology stuff behind laser grids and guard it by the Red Queen. This can also help protect you from data thieves and nosy activists within your own company.

Yes, that how it works, if you looked at your data files you would see things like. UberMahn Serum 1.3 GB, VitamanX 57 MB.
Then again I could just make it general like Toxicology 1.3 GB, BioChemestry 57 MB.

quote:
This is also cool, but again I think the data shouldn''t be generic MB, and that there should be different sites to raid that have different tradeoffs. For example, maybe Terragen Corp has great virology, but a nasty spec ops team guarding it; whereas Immunicorp has more generic viral data, but is easier to get at.

It wouldn''t be generic, you would see things like your commandos stole 7.8 GB of Virology data.

quote:
You could add another level to this, too: Diplomacy and alliances among corps, so that raiding is more substantial. For instance, you could raid USAMRID for their last remaining bubonic plague samples, but you''d bring down the might of the USA on you. If there''s more nuance, then maybe you could bribe someone for a sample, or compromise them. This could be handled by a covert spy network that simply presents opportunities with a price tag, success chance, and blowback cost. They covert ops department would present more opportunities the better funded they were, simple as that if you like.

For infilitrating, subverting, data theft and sabatoge of an organization please see Information Services.

For purchases, and deals with other companies, branches and divisions please see Acquistions.

Have a nice day

quote:
There''s no problem with randomization that I see as long as the computer is bound by the same rules, but know that it will kill some of the strategic flavor of the game. For that same reason, I highly advice against a dead-end research effect. Researching requires effort from the player, there should be some reward. If you want to make it more like gambling, consider making some rewards better than others.

Only application projects have a random factor to them, and all application projects are optional leaves of the research tree. They are benfical but not manditory. By making them random I felt that it would add more strategy then not having them random. Since if the player relied heavily on a certain application project say VitamaX as an earily source of revenue, and they didn''t discover that project this game, then they would have adopt a new strategy this time.

quote:
Is there any depth to a lead researcher, or is she/he just an expense + some skills?

Oh, I ya didn''t include anything on personal in my post, I''ll fix that up.

Basically there are three kinds of personal in the research deparment.

Head Resarcher - 1 required - Is in charge of all research at a facility. Provides a bonous to all projects being research there based on their stats and skills.

Lead Researcher - 1 lead researcher is required for each project you wish research beyond the first. So 1 lead reacher means you can research two projects at a time.

Reseacher - personal assigned to work on projects.

You can either hirer or promote personal to the first two postions. promotion to Head Reseacher provides a loyalty boost and raises their salary $10,000 a week, same for lead researcher but with salary increase of $5,000 a week instead.

Demotions lower salary and loyalty by the same amount.

quote:
What is internal risk and what do you mean by exposure?

Good question, it is a little vague, If you don''t have my notes with you.

Internal risk is the amount of possible danger associated with performing the experiment to your personal and facilites. For instance testing a new nerve gas in your only research facility the worse case scenario is that the gas fills the facilites and kills you and everyone else in the facility. Actual level of danger will be given to you and that is based on personal and equipment.

Exposure represents the risk of knowledge of the experiment being made public, and thus the danger of it being traced back to you or the company. Which could lead to all sorts of problems such as the SAS storming your facility.



-----------------------------------------------------
"Fate and Destiny only give you the opportunity the rest you have to do on your own."
Current Design project: Ambitions Slave

You should give research scientist stats. By assigning different scientist as head researches, you'd get different results.

For example, you could have Experience, Knowledge, Ethics, and Loyalty, ect. So if you had a scientist that had high knowledge, and low ethics, the job would get finished fast, but there would be a better chance of getting busted for human-rights violations.

[edited by - Onemind on May 8, 2004 2:29:06 PM]

They do have stats, all personal in the game have 11 common stats, and 6 department specific stats.

If your interested they are

Name:
Postion:
Salary:

values 0-100, 50 is the mean value.

Intellegence
Cunning
Charisma
Strength
Stamina
Health

Loyalty stats:

Public Loyalty Company - percived loyalty to the company
Public Loyalty Personal - percived loyalty to you

Private Loyalty Company - actual loyalty to the company
Private Loyalty Personal - actual loyalty to you

Knowledges:

Knowledges are department specific, for the research department they are.

values between 0 and 10, no mean value.
Chemistry
Biology
Virology
Genetics
Behavioral Sciences
Data Analysis

Researcher also have specialties, most reachers don''t have a specilty and the ones that do cost more, a reseacher can have more then one specialty but that the chances of having a specility decrease exponently as the number of specilities increases.

Specialties that are compatiable with the research current project provide the following benefits.

Increased chance of a bnous event
Increased chance of success
Increased number of successes
Increased chance of an application being discovered.

As you can see specilities are very valuable so its worth keeping those scientists with them safe and happy.




-----------------------------------------------------
"Fate and Destiny only give you the opportunity the rest you have to do on your own."
Current Design project: Ambitions Slave