Thoughts on technology

[haravikk]

I'd post this in the design board but I'm very likely going to step out of the intended line of discussion so I'll just throw a load of thoughts about what I've read into the tech tree thus far.

Generating RP's
This should be fairly standard, labs generate RP's, research agreements with races add/subtract RP's (depending at which point in the agreement you are) and the ability to put some extra cash into research would really help too.

Distributing RP's
This would be done by a slider bar method allowing you to choose how many RP's get spent in each category. An additional two categories however would be beneficial:
Refinement - How many RP's to spend on refining tech
Breakthrough - How many RP's to force to become surplus, RP's put into this category can significantly increase your chances of early breakthroughs. These are in addition to RP's left over from each category due to there being no projects to spend them on

Researching things
I like in MOO3 how the AI automatically researches things that become available to it, however with this system it may not be able to research everything that is available (not enough RP for example). So if an AI is to help then it would probably make sense for it to automatically research the maximum number of possible things at once, mainly because this opens up more advanced techs than later techs would (if any are available). Cancel and start project buttons would be beneficial for the user if they see a more advanced tech that they would prefer, they can cancel a few lesser ones and start this later one instead. The AI should ideally be able to make decisions based on the dependency of techs, if a later tech would open up lots of lesser ones then it may choose this instead of a few other lesser ones. Or if something in another category may soon require an advanced tech then it should develop this as the need for it is likely to be greater.

Basics
I like the idea of having X RP's per turn for X turns for researching a project, however I do not think that it should be used for all tech areas. I am a fan of the idea of Theory, Application and Refinement as separate entities and I think they should all be done a little differently:

Theory
This one would use the RP/turn system with a few little differences. Each theory project would have it's RP cost per turn, and a 'total' RP cost. This is used to work out how many turns it will take to complete and is not a definite value. I think that when a project hits the 80% done mark it should enter a 'breakthrough' period at which point you start getting a higher and higher chance of suddenly completing a project. This percentage chance would depend on a few things, firstly a race attribute (creativity?) and secondly on your surplus of RP's. The greater your surplus, the greater the chance of an earlier breakthrough.
Without excess RP's to cause a breakthrough a theory could continue a long way beyond it's minimum number of turns as the percentage chance of completion only slowly trickles upwards, a tech that takes longer than expected will decrease the minimum time for the next tech in that category by half of the excess time it took. e.g if a tech was meant to take 40 turns but took 60 (50% more) then the minimum time for the next tech in that category would be lowered by 25%

Application
Applications would be different. Once you have the theory complete it is usually easier to decide how long a project is likely to take. In this way an application would again have an RP/turn cost with a minimum number of turns method to completion. However, different from theory an application cannot complete early except by a few methods:
Event - if a special event hurries it up a bit
Better understanding - if an application is completed then it will decrease the research time of all other applications on that level of that category (ie the application may have components useful in another that now do not require researching).

Refinement
Refinements are made on a curve basis to all technology at once. For example, level 1 of refinement will reduce weapon sizes by 5%, increase trade income by 5% etc. The curve simply increases based upon how many levels of refinement you have achieved, at its simplest each level could simply cost twice the previous one. Unlike other techs refinements have a fixed cost, once the required number of RP's are reached then the refinements are done and become available.
Refinements could very easily follow tech development in that level 1 of refinement is only applied to applications on level 1 of the tech tree (in all categories). This way, when you develop plasma cannons later on then they will be at normal size until you refine that level of the tech tree (which could take a long time). If this method were used then the refinements would have to be bigger, say 20% unless you research a tech that improves refinement.
Another possibility is to combine the two, so for example when you reach refinement level 20 you not only refine all applications on that level, you also further refine all applications on all previous levels.

Relationships for techs
The relationship between theories and applications is a complex one. On each level of a branch of the tech tree there can be any number of applications and/or theories. Theories can be researched as soon as you reach that level of the tech tree, however applications will require a number of theories before the project can begin (it may not even be visible until the theory is complete), this could be a theory on the same level, or a theory a few levels back, or both.
Some applications may also rely on other applications from the same or another category. For example if you are developing enveloping laser in physical sciences (ie it may require a special type of coil component) then you obviously require laser cannon to be developed under the energy category.
Further still is that some applications may require theories several levels in advance in order to begin. This could be a good way of allowing empires to catch up on technologies they earlier missed. For example if I missed light armour because I had not been able to research 'lightweight materials' theory then when I research 'reflective materials' which allows for medium armour then this could unlock 'lightweight materials', allowing me to research those older (and probably much quicker to research) techs.
A note on this is that if a level is all applications that are unavailable then tech will progress to the next level, skipping these applications until required theory becomes available.

Progressing through technology
Technology progression should be done on a level system. Each category has loads of levels and each category may progress through these at different speeds depending on research allocation. Progressing to another level requires first that you have completed all tech available to you in that level. Once you have done this the next level is treated like a tech with a few differences;
It has a fixed value of RP's that it requires, the more RP's it gets the quicker it can progress to the next level.
It has a minimum time like a theory.
It is subject to breakthroughs like theories, but it may not take any longer than it takes for it's RP value to be reached.
Once it is done research into the next level may begin.

Balancing technology
A big issue is with balancing technology. Often a race that is good at building things is no match for one which is big on research as the research heavy race will quickly develop technology that outdoes the other race's building ability, leaving them behind with little way to catch up.
Which got me thinking about balancing technology in such a way that a race may catch up or at least remain competitive if they are getting left behind;

Espionage
This is a good way for an empire to quickly catch up, a well funded spy program can get you theories and applications far more advanced than your current ones, allowing you to skip whole levels in your tech tree. This however is pretty risky with the chance of provoking war, or being entirely wasteful if the target has a good counter-espionage program.

Diplomacy
Research agreements with more advanced empires will give you a lot of extra RP's that can really help out. It can also increase the chances of an early breakthrough for you if you are researching something that the other empire also has.

Scaling techs
The simple scaling of tech and the minimum time to complete each one is a good balancing factor as research heavy empires will start to slow down over time where less advanced ones can catch with their relatively quicker to research technology levels.
When a level becomes available it might be a good idea to have the costs for the techs in that level calculated based upon the average tech level of the rest of the universe. If you are a bit behind then the costs/times will be quite low, if you are way ahead then the costs/times will be large.

Research 'quality'
The quality of your scientists (not necessarily their intelligence) could be another factor. If your empire is overall very happy, and your research projects have generous amounts of funding then your scientists will generally work much harder. Further to this, if your empire is further behind then your scientists will very quickly become more motivated to work harder to catch up, this is the more important factor to quality.
The quality of your scientists would further effect the chances of breakthroughs in theory, if your scientists are unhappy or poorly funded then any amount of surplus RP's from labs etc will make little difference to your chances of a breakthrough.
This idea results in empires that are further behind having lots of breakthroughs, if they work hard on lowering unrest and getting a good bit of funding into their research then even a low RP empire can start developing things quickly and catch up.

[Aquitaine]

Checked the 'design archive' board for the public review of the tech tree (part I) and the very last post. Some of these topics are already decided (but several of them aren't).

[spiff]

just a small contribution to the tech ideas. i noticed this document:

http://www.freeorion.org/index.php?modu ... ology.html

"MOO2: The Good:
-- When a new item became available it was a big event, with cool graphics and animation.

MOO2: The Bad:
--Techs run out in the late game.

MOO3: The Bad:
-- Advancement is too fast and happens without any fanfare and with sucky art"

and so on; so maybe freeorion could merge the two? how about:

- the usual techs presented with fanfare because they're a genuine breakthrough
- interspaced with small "advancement" techs which might, say, improve certains things (e.g. weapons, hull sizes) by a small % without fanfare, just a little announcement.

this way freeorion has major breakthroughs and constant development - putting more money in tech not only gives you great new paradigm shifts now and again, but constantly improves your empire. freeorion could then spread the major techs so theres tech late into the game.

i understand the tech ideas are largely settled, but this may require very little change, if any to the design? just a thought.

[guiguibaah]

Hopefully some sort o randomization will still occur wih the techs. One problem of Moo2 is that everyone would beeline for Automated Factories, Automated Farms, Research Labs, etc... and do it always every time they played the game (unless they played as uncreative) never choosing the other techs (like heavy armor or biospheres).

[Aquitaine]

There will be no tech randomization.

-Aq

[Satyagraha]

Hopefully some sort o randomization will still occur wih the techs. One problem of Moo2 is that everyone would beeline for Automated Factories, Automated Farms, Research Labs, etc... and do it always every time they played the game (unless they played as uncreative) never choosing the other techs (like heavy armor or biospheres).

in a well-balanced game, this shouldn´t be an issue. if someone can come up with a good strategy it´s only fair if he can apply it whenever he wants.

[Krikkitone]

Hopefully some sort o randomization will still occur wih the techs. One problem of Moo2 is that everyone would beeline for Automated Factories, Automated Farms, Research Labs, etc... and do it always every time they played the game (unless they played as uncreative) never choosing the other techs (like heavy armor or biospheres).

in a well-balanced game, this shouldn´t be an issue. if someone can come up with a good strategy it´s only fair if he can apply it whenever he wants.

Not necessarily, after all many games have random maps and in a game like Civ you adapt your strategy to the situation (bee line for heavy ground troops tech isn't the best idea if your on an isolated island) This also works with games where you know everything to begin with (Risk 2210 randomly eliminates some positions making standard strategies useless if they are not flexible.

A random tech tree (whether the player knows what has been changed from standard at the beginning of the game or not) can make games more interesting.

However there is the point that techs are very powerful and anything beyond slight randomization can change the balance between different Types of strategies (military v. researcher, etc. which changes the value of race picks.. although this could force more generalization... but it does make the game more of a random event)

Which would be why we apparently are not having tech randomization

[Sandlapper]

Krikkitone wrote:

A random tech tree (whether the player knows what has been changed from standard at the beginning of the game or not) can make games more interesting.

I made this very argument before the final review, but to no avail. Even if there is perfect balance among all techs, a player is likely to perceive a preferred tech route over other techs, because some will fit their style of play better(prefers carrier attacks over heavy weapons, etc.). These same techs will be picked, time and time again, because of perceived superiority.

As stated earlier, there will be no random techs.

[guiguibaah]

Actually I wasn't referring to blind random research. It was more a critique (albeit not a good one) of how Moo2's research system was exclusive. One was not able to research something (like biospheres) because they chose Hydroponic farms. As a result, in multiplayer everyone would beeline for automated factories, but because the system did not allow for dual research in the same field (unless you were creative) you'd never see 'heavy armor' in the game.

I would say it was more of an oversight on the moo2's dev team. Nobody could predict after thousands of people playing it that automated factories would be so more useful than heavy armor or fighter base.

Of course, if you do want a challenge in Moo2, play as uncreative

P.S. - I don't want to irate the dev's about game issues that have already decided.

[haravikk]

In MOO3 the order in which technologies appear seems to be randomised, for each empire, could something similar to this be done?
ie certain 'base' techs are on specific levels, however some techs leading off from these could appear on one of several levels, sometimes making for levels with loads of tech or empty levels.

Just to break the tech tree up a bit

[Sandlapper]

Too late, this discussion has already been discussed, and decided on. Any desire to change this will have to be in a post release mod (after version 1.0) at the earliest, if anyone wishes to code it.

[Aquitaine]

The inherent problem with this suggestion is that it operates on the premise that it's just impossible to balance the tech tree, and so the best solution is to make it go totally bonkers and randomly remove/re-route pieces of it.

In my book, this is the worst possible solution to the stated problem.

If you look at our myriad of tech discussions, you will see that our tech tree has more to do with HoI than it does with MOO2. And yet this suggestion is trying to fix a problem with MOO2 -- not with ours. In HoI, there is no equivalent to 'rush for automated factories' - also because there's no building.

I see randomization of the tech tree as abandoning our responsibilities as designers. I hated it in MOO3.

-Aq

[Kostik666]

darn it just when i wrote a whole essay on building management... the lead designer says not... well mother f$%^er... oh well past that


for the tech tree, i dont like the idea of having tech 'levels' and in order to reach a next one you need ALL tech on that level.

i personally like going for the heavy stuff - reseach a big gun take a few extra planets and then get all smart and uppity. i like plasma cannons and i dont like farming. so i say you need to have a portion of technologies to reach a new level, just that some other section (like farming) can lag behind a bit and get better later, while big guns and armies can get strong.

and there could be randomization but only on the same tech level, as in the max you can have is 3 technologies from a level of any given section, but in total there are more than 3 techs there, which other empires will get and you wont, thats why youll have to lie cheat steal and kill...