FreeOrion

<administriva>
This thread will review several discussions relating to our tech tree and allow limited feedback. This thread is about presenting options that have already been discussed and collecting consensus. I will offer several questions that have come up and various answers that have been presented. Most any suggestion that got some support on the game design board made it here;If you were a part of these discussions on the game design board and your suggestion didn't make it here, PM me -- this is a very large topic and so I am liable to have missed something, for which I apologize in advance!

On this thread, though, please limit your remarks to the following:
- Support of one of the ideas listed
- Disagreement with one of the ideas listed
- Questions asking for more information

Please do not post entirely new submissions. These posts will be deleted. Every single review thread we've had so far ends up with me saying 'please don't do that I will really delete it NEXT TIME.' This is such a gargantuan topic, though, that we cannot afford to get side-tracked in this manner. On to the tech tree!
</administriva>

Usually I do a 'best of' the game design threads. Unfortunately, there was so much good commentary on this subject that it would be a disservice to everyone who participated to make a decision without having read it. So please, if you're interested in the tech tree and want to weigh on this, please read the following threads (it's about 10 pages, total). I will try to link to the specific parts of these threads when dealing with the questions we need answered. Generally I will link to two pages (usually consecutive pages in the thread) and the discussion may continue after those two, so you do need to, like, read and pay attention. I'm going to start with what 'assumptions' we want to make about the tree. After this point, nothing has been passed; so anything like 'FO will do this' is generally establishing one assumption to set up the actual proposal. This is the stuff where we need consensus.

http://www.artclusta.com/bb/viewtopic.php?t=412
http://www.artclusta.com/bb/viewtopic.php?t=410

I'll start with the current status of the tech tree with regard to what we may safely assume will already make it in to the game.

The research model was originally based on the one in a game called Hearts of Iron from Paradox Entertainment (http://www.paradoxplaza.com). The following may be assumed about our model:

- Technology is organized horizontally into some number of categories.
- Technology is ordered vertically by theory and then application, e.g. a 'Capital Ship' category might have 'Large-scale propulsion' as a theory and 'battleship engine' as an application. More specifically, theories are required to research their related, child applications; as a general rule, theories will not do very much by themselves, but this is a breakable rule.
- In addition to application and theory, there will be some kind of 'engineering versus scientist' mechanism to refine existing technology. Some ideas on how this might work are in this thread. This covers things like making your current tech smaller/faster/better/stronger.
- Technology has a fixed cost, both in research points and in number of turns required. If an Empire produces 100 RP a turn, it may begin ten separate 'projects' that cost 10 RP. You may not invest 12 RP or 50 RP in one project for a faster result. Each of those 10 projects may have a different number of turns associated with it. The amount of time required is fixed in the sense that more RP will not change it, but it is not fixed in the sense that some technologies may reduce the time requirement for future technologies, or there may be some event that finishes one of your projects. There are questions still open relating to how excess RP should be managed (what if your Empire produces 105 RP a turn and there are no 5 RP projects), and whether or not it should be possible to begin a project with less than the required amount of RP.

For more details on this last point, see the above threads.

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

I'm relying on an excellent summary of the discussion threads by emrys, and will quote him directly. I've edited out some of the questions he had that will have to wait until we're actually building content and assigning costs. My comments and other administriva on the status of the question are in bold.

And now the questions! Meta-commentary or administriva in bold.



This is a lot to digest, but the more I tried to simplify this down into 'please choose A, B, C, or D' as we have in the past, the more I was short-changing the process and people's suggestions. I don't know that we got all the questions we need for round one here, so I may edit and update this review as we progress. But if you read both of the DESIGN: threads and this one, you'll have a pretty good idea where we are. This review will be open for at least a couple weeks (if not more), so take your time!

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

I pretty posted everything I wanted to say and think of, but since nobody has posted anything for the past few day on this important issue of tech tree, I figure I'll bump it up and spam with stuffs that I already said.

Q: Problem with extraneous rp
A: Depending on our econ model, if you do pp -> rp, we can just allow any extra rp be automatically converted back to pp in the imperial stockpile at no penalties. There is no reason why we can't implement this, if econ model passed as such, with my next idea.

As most of you know, my other idea is to have a rp imperial stockpile where you can use it to advance a cat/theory. We can work this idea and solve many other issues such as allowing players to get ahead of research because they have extra rp, how you get technologically developed, and prereq problems. A quick review, my idea is after you research a theory, you can refine the theory again by paying some rp (maybe a modicum amount to start) to up it level which cause all applications stemming from the theroy to recieve a discount in time and/or rp costs. You can keep uping the lvl if the player likes and opens up a short-term research strategy where the player can use his rp to invest on cat/theory so he can get mega-laser and super-amor ultra fast. Refining the theory may cause other new theories to spawn. Prereqs for theories and applications are determine using lvls. So an application may require x numbers of total lvls from this and that theory/cat.

Q: theory opens application
I thought we passed this? Anyway I am in support. As far as refinement is concerned, we can do my lvls idea from above and say you can spend your extra rp to refine an application. By doing so and at certain lvls, you get improvments for the tech. So if you have laser weapon reserached, you get continuous mod at refinement lvl 1, overload at lvl 2, autofire at lvl 5...

Q: Cats should relate to gameplay or particular strategy
Don't really have an opinon on this. I don't have any idea why I would chose one over the other off top of my head. If you mean "gameplay" as in this cat affects all industry aspect of the game and "particular strategy" as in diplomacy techs to be grouped under a cat, then I can't say I have a perference right now.

Q: Research cost will scale with galaxy size
The best I think is have a scale number in the option screen for tech speed customization. There are many players who like techs incredibly slow, and there might be days when ppl want to play a fast research variant game. I think a simple scaling factor be an option should satisfy everyone and whatever preference or style that might want to go for that funky tourney map that they specially customized. Doing this will also solve your galaxy size problem.

If you want to hard code the scaling factor for each galaxy size, I perfer not to use it because the whole point of selecting galaxy size is to have a different play experience. A relative scalar factor will pretty much ruin the flavor.

Q: Blocking out techs
A: I am in support of blocking out non-key and non-requesite techs and the player knows what techs are locked out ahead of the time (maybe at the begining of the game or techs/lvls just before it when the player views what choices he has). This will add more variance to each game, which makes it more interesting and challenging. If you block some techs for some players and other techs for other players, that will enhace diplmacy and the spying game. If we block techs all together for that game, that will force the player to adopt with new strategies with new paths.

Q: Few or many cats
A: I just want something presentable. I prefer not to have scroll bars and go through clutters. I think we should have enough cats so organization makes sense and look nice. If we can, few and simple is best.

Q: Research application without knowing the theory
A: That will work as another mechanism to get ahead in the reserach game on short term. I like my idea better because it kill a few birds with one stone.

Q: Techs that reduce cost and/or time generally or cats
A: I do not support techs that specifically do only that because of cluttering. Workable if some tech also reduce cost/time on top of their normal bonus.

Q: recurrsive techs or eras
A: Yes because 1. simplify balancing 2. Allows research and growth to go on forever 3. No more race to uber weapons and ships and see the same ship types and weapons over and over again 4. Good excuse to put on new skin for UI 5. Further organization of techs on top of cats by splitting into the 4th dimension 6. descrete eras can be used to highligh rps macro-research game. The only con i can think of is extra coding is needed.

Q: Racial bonus or make techs dependent on strategic position
A: Racial bonus may include unique techs and/or bonus to entire cat. I think specific tech is going a bit too detailed.

I don't know how are you going to hardwire making tech dependant on situation. Some crazy programming there.

Q: Balance all techs or have preferred paths. And how to support different play styles
A: Both seems good. I think my lvls idea will support different play style simply and fun

Q: Cancelling projects
A: For UI, the player gets three options to research a tech: click once to start project, click again to put on hold, click a third time to cancel the project. If player cant afford, the tech gets put on hold like clicking the tech twice.

_________________

I agree with the putting of the spare RP into a stockpile which can be used to refine already researched applications, by which I mean improving their stats / reducing their costs etc.

I don't agree with investing in categories or theories to reduce research time or cost, since part of the benefit of the HOI model is place an upper limit on the rate of research advancement with the minimum time requirement.

I'm also not sure where I stand on the idea of the 'weapons mods' being made available by refinement, as compared to them being separate applications made available by researching the weapon and refinement improving stats generally. I'm fairly indifferent to be honest.

Also, did we (or can we now) come to a consensus on whether we wanted strict prerequisites for a tech, OR a given number of levels of refinement between the list of choices, OR a combination of both (i.e. three prerequisites, you MUST have at least researched each to initial level, and also need four levels of refinement between them, but they could all be in one, or spread evenly or whatever).




The question that was asked was a choice between categories that are from the point of view of the essentially arbitrarily named (i.e. PHYSICS, BIOLOGY etc.) OR names that relate closely to game play and particular strategies.



I would say that there should (at least as a default) be scaling by galaxy size, otherwise i think it will be almost impossible to sensibly balance the game. This doesn't mean that there shouldn't be the choice of e.g. 1(slow) to 10 (fast) tech speed, which of course there should be, it means that 5(medium) on a small galaxy should be about the same pace as 5 on a large galaxy.



Again, I think you've missed the point of the question (which means I didn't explain it properly, drat).
The choice was between shaping direction through the tech tree by hardwired racial bonuses, or ensuring any race could take any route, with the choice being prompted only by the situation they find themselves in in that game.


You can't have both, since a perfectly balanced tree by definition has no prefered path. That's why we need to choose.



The workings of the UI are not the point. The game mechanics are. The upshot of the discussion in the design thread seemed to amount to the following:

You could only start a project when you have enough free Research income. If your incoming RP fell below the total of your income, your projects would get put on hold in (descending) priority order, if it rose again they would get activated in priority order (?automatic suffling of order if you only have enough to activate a smaller project lower down?). Presumably we would allow players to reorder their list, i.e trade off active projects for suspended ones. If you had ANY projects on hold you would not be able to start ANY new ones. You could cancel active or on hold projects, but you would lose all the research invested so far.

Are we happy with this?

If we have the left over tech be used for refinement (i.e. advancing the level of a previouisly researched tech), then a player may conceivably want to deliberately suspend a project even though they can afford to run them all, in order to spend the RP on refinement instead. Should they be allowed to manipulate the system this way or not?

It could be the place for refinement, but I really think research should be able to be distributed between research and refinement. Can't we just put it back into the current tech, or the next tech to be researched.


I think few categories, but things like offense, defense, travel, economy, relations, leasure.


I don't like this idea. First, it is obviously unrealistic (don't shoot me!), I mean woo, a galaxies size determines the laws of the universe.

Secondly, players will get confused, because the research will go one speed in a large galaxy and another speed in a small galaxy.

Third, by changing the speed of research, your basically going to make all galaxies gameplay the same. You'd expect a small galaxy to be a quick battle and a large galaxy to be a slow battle (the game will take a while to play), but with this research speed idea, all galaxy sizes will take the same length of time to play. Luckily we're not going to make other things change based on galaxy size, such as build time of colony ships, etc (then playing in a small galaxy would pretty much be the same as playing in a large galaxy).

Also, what is this problem I am hearing about to do with galaxy sizes?


No, only if the race cannot use the tech.


Fewer.


Maybe, but thats like skipping branches, skipping prerequisites. You could but maybe that fancy laser you get is unreliable, or needs to be refined before it can be used.


Yes, I supose it would be a good idea. Maybe something like anti-spill/anti-break test tubes eliminates the overhead of accidents, therefore reducing research costs for all projects.


Some kind of recursive techs would be cool. Though recursive techs tend to be just refinements.

Maybe we could have random techs. So, if a weapon has certain stats, then a random tech could be created, given a random name and walla, the super ultimate laser 3000, with 6000 damage. Basically stats like damage would be chosen based on how far ahead in research you are, ie the further ahead the more damage. Also things like cost and size should be proportional to the damage, range, etc.

Another alternative is to have heaps of premade techs (maybe 100 or 1000, no asking too much), that are randomly placed onto the end of the tech tree.


I think certain techs could be cheaper for certain races.


This would probably be based on the categories you have:
offense, defense, travel, economy, relations, leasure.

If the categories represent different styles of play, well then there you go. Though, some players may have a style all their own, so you cannot guess how someone will play. I think just make the categories like I have suggested: offense, defense, travel, economy, relations, leasure, plus more I didn't think of.

And I think that clear paths should be removed. SO that there is no one clear path or not even a few clear paths. The path you choose should be based on category. So you want weapons, look to offense category.


It is kept, so it can be picked up next time. Think of it is priceless data. It should be filed away and backed up, ha ha. Your scientists should be able to look back on past projects in their research database. They would see that project X was never completed, so an ambitious young scientist will want to prove himself and complete it. Sorry for the story, heh.

How do you plan for psilon to be ahead of research? Or what if some wish to play humans and wish to go with research path? You can't get ahead of research if you capped it, which was one of the main criticism of HOI model.



Nah, I always thought names like fried-chicken is better. Giant Apes is good too. I'm pretty sure I understood the question. ~_~. How about some totally random names like skdiw.




I think research should be flexible enough to do both.



I meant to say either way is good.



I mean I don't want something too complicated.



Something like list of priorties seems complicated. I want something simple like the 3 clicks.

"If you had ANY projects on hold you would not be able to start ANY new ones" I dont think I like this rule.



Dont really have an opinon on this atm.

@untilae: Some ppl complain that in small galaxy sizes, they are stuck at early techs and never really gets a chance to advanced to the uber techs. So they want a scalar factor so the they can get to the later techs. Basically, by making game features relative, the galaxy size basically determine how long the game will last, which isn't a bad idea at all. But only other hand, I agree with you that part of the reason why I play a small map is because the scenario is different so your strategy have to change so the game is more interesting.

_________________

Essentially no. The point of the HOI model of research is that it creates discrete (quantised even ) steps in the rate you can spend your RP. If you have all the available projects running, and still have spare, or have some projects but not quite enough to start another, you are going to have 'unspendable' RP. We can't just let the player put it into the current or next tech, since that would mean going back to the straight MOO model where you just have to collect enough RP points.


Acutally, the idea of scaling tech costs to galaxy size is to make research go at the same pace in small and large galaxies.



I just don't see this. The number of worlds to conquer will be different, so the length of time required to play will be different. The idea is just that generally by the time you've taken half the galaxy, you'll be at roughly the same tech level in both the small and large galaxy.

On the other hand, I don't suppose it's that relevant really, so long as we allow players to set the research pace at game start up, then if we chose not to make it scale with galaxy size, players who want it to can just remember to slow it down a bit as they increase the size to compensate, and if we do make it scale, players who don't want it to can do the opposite. Therefore provided we allow the setting of research speed at game setup over a reasonable range this is a non-issue.

re:cancelled research


If we do go with allowing you to "pick up where you left off" on cancelled projects could I suggest a "dusting off" cost. i.e. the longer you leave the research 'abandoned' the less useful that research becomes. (Technically, I'd suggest an exponential decay of the number of turns you're considered to have done on the project). This might also apply to suspended projects, or maybe you could have to pay a small amount of RP to keep suspended projects ticking over to avoid them decaying, thus pushing players to cancel projects outright rather than leaving them suspended indefinitely.

I envisage it like this, at any one time, a race might have for example four to six choices of projects to research. The RP/turn cost of techs would be pitched so that a typical race could not afford to research more than, say, half (i.e. two or three) of their available options at one time, and typically will only have a small amount of RP left over to refine things.

Research poor races would able to research one at a time, or maybe two occasionally. Strong research races would be able to research most of their available options at once, and if they also had a strong empire would have regularly have a fair amount or RP left over after they'd started all the projects they wanted to refine existing techs with, so they're stuff would be more advanced than their opponents even at the same generla tech level.

The upshot of this is that research strong races would be at the cutting edge of tech, advancing as fast as possible as well as having a broad sweep of tech and having generally better stuff (from more refinements), whereas the less research intensive races would either fall behind slowly, or be forced to concentrate on a narrow path up the tree in order to keep up.

What it would prevent would be the Psilons hiding in a corner of the galaxy,pouring all their research into a narrow path, racing up the tree until they got an unassailable lead, and then coming out and taking over the galaxy with their single unstoppable death ray titan.

Even the most powerful research heavy race would only be able to get ahead slowly, and only a certain distance. In other words the system would be self limiting and self stabilising, not suffer from the incredible runaway feedback that the Moo series tends to suffer from in tech and many other ares.



I'm assuming I'm missing the point of this comment, possibly a result of the lack of expressiveness of text, but I'll take it as a vote for Gameplay/Strategy related categories unless corrected.



The 'three clicks' approach is fine, but it would require us to demand that the player chooses which projects to suspend every single time they fall below the currently required RP/turn,i.e. prevent them finishing the turn until they had, since we would have no hint as to which project should be cancelled. If that seems like a reasonable restriction to impose to avoid having a drag-to-reorder list of projects, then we can do that. Also I'm not sure what we would do about unexpected drops in RP production (e.g. if a major research world gets glassed/taken), how do we decide what projects to suspend, or do we do all research with the RP collected last turn in order to avoid these problems?



If we don't have this rule, and especially if we give players the option to elect to suspend projects even if they actually have enough RP to keep them running, then in effect we have indefinite 100% underspending, and more particularly have negated the planning element that the HOI model gives us, i.e. that you have to plan to commit x RP for y Turns. I'm happy with that if everyone else is, but I'd prefer not.


p.s. I'm coming more and more to the feeling that modifications of particular weapons (e.g. autofire/continuous etc.), if we have them, would work best by being 'refinements' i.e. researched by dumping left over RP into an existing application.

Alright, I thought I'd chime in for once (it's been a while).



I think that these should just go to waste. The micromanagerial user can pull out his hair and try to achieve a perfect number of RPs for the projects he has running, and the macromanagerial player will make sure he uses all the RPs he can, and won't sweat the small amount of waste. We should let the user find his own way, when possible, and reduce the overall number of exceptional rules.



I would prefer it if refinements were just smaller, cheaper, faster applications with proportionately smaller benefits. Ideally, I would like it if it were possible to get roughly the same benefit from a primitive weapon type that has been refined to perfection as from a new, higher-tech, but unproved weapon system.



I agree with this; I think HOI did a good job of categorizing, and we should strive for a similar breakdown in our categories.



Varying this solely based on galaxy size seems arbitrary to me. I understand the reasoning behind it, but to be honest there are as many good reasons for as against, because the desirable rate of research is purely a matter of taste. If this is going to be variable, it should be adjustable for every game, not just those in different sized maps. I say this because the desire to play on a larger- or smaller-sized map is not necessarily the desire to play a faster- or slower-paced research game. We should not link the two because they sometimes affect each other, but sometimes do not.



I think we should avoid this, because it makes things harder for the AI programmers. Scripts become a nightmare if they have to account for too many possiblities; the winningest AI strategy might depend on a tech that is suddenly gone in a particular game. Finding the best strategy can be hard enough, and finding a whole bunch of them and using them based on what techs are available might just be impossible. Also, I think the same kind of replayability can be achieved by making the tech tree broad and rich.



I like a lot of categories, but not too many (8 to 12).



I think this wrecks the whole model, and it certainly violates KISS.



I can see why some people favor this, but I can't stand seeing "Our scientist have discovered {micro|macro|nano} - {nuclear|subatomic|gravitonic|biological} farms." You get the idea.



I think research point progress in a project should decline over time (fairly quickly, like at the same rate it was being researched) if it is not being actively developed, but that you should be able to de-fund anything at any time.

I think it's worth remembering that we may be talking about a relatively large amount of waste. If we usually have about six projects running at once (a number plucked out of thin air), then assuming similar levels of cost for all you optionsyou could typically expect to waste up to a seventh of your RP output, etc. That might be enough waste to annoy even the fairly 'macro' managers.

Unless we have a lot of small projects available, I think we'll find we need to do something to mitigate this loss. (I'm personally in favour of using it for refinement, but that's probably fairly clear,and totally irrelevant).

I'd agree with that. It's simple(a BIG plus) and clear and suitably penalising for players who try to play the system. Plus it feels right.

I actually thought this originally (because that's how HoI handles it and it isn't too much trouble in HoI to manage it). But I've been won over to using it to handle refinements. I wanted refinements to be applications, but instead I think they should be neither theory nor application; rather a separate division of research (perhaps 'engineering') that does work more like a MOO2 model that just eats your spare RP.

Now, even with this, you'll still have RP going to waste, because you won't always have things to refine and maybe you won't want things to refine. So we could pour excess RP back into the economy somehow, except that we haven't yet figured out what to do with money (or even if we want to have money).

That will probably be the next review. (ugh)

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

I favor treating excess RPs as waste, because
1) If you're wasting too much, you should just stop producing RP-generating facilities, spend less on research, etc.
2) It prevents you from "gaming" the game. In MOO2 you could get cash for food (!), and sometimes I would jack up my food output to pay for a bigger fleet! This is just silly.

Having said this, I don't feel too strongly about this particular point, so I won't say anything else about it.

Maybe we could use excess rp to represent those researchers who make accidental discoveries or to represent those researchers who are off on their own doing research the government has not told them to do.

So basically the exzcess rp will go into some kind of bonus tech, which is only revealed once completed. So, as your playing the game a message pops up "an enexpected discovery: explosives".

If excess rp was going into refinements, then that would mean that there would be no control over refinements, it would be automatic (an inflexible system). I think you would need to be able to make a choice, to invest more in normal research (theory, applied) or invest more in refinements.


Also, I think that there should be an option "scale research with galaxy size", instead of just scaling the research with galaxy size.

This is an unfounded assumption. The decision for this question is simply what to do with it. The level of control you have over it would be decided once we determine that this is what we want in the first place.

We could, for example, direct the 'engineering' or refinement RP as specifically or generally as we want.

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!