Pardon me, if this belongs to a different sub-forum, I've read that this one is only for systems that are being looked at, but I couldn't find the list of systems that are currently being looked into, and since research is set-up for "ongoing improvement" and "stealthy additions" i figured it's not locked.
Closer to topic. I think research system can be improved. Currently it gets progressively longer to reach each breakthrough. (which i understand is a counterweight for territorial growth and acceleration of research effort). I don't know how exactly this is calculated so I can't say much about it. Having more information would help.
What bothers me is that Old research takes progressively longer as well.
As a result of current selection system (I'll refer to compiling list of available research as "pulling" from now on to make it shorter) player has to decide if he wants some crucial "green" research or something rare but optional. I do not know if this is intentional or not, but i find such system lacking.
Since every breakthrough makes it progressively harder to reach next breakthrough, such a choice is very "heavy" in terms of consequences.
I also find red research "lacking" in rarity versus effectiveness.
What I consider a good aspect of games with randomly generated content (D2 and the like excluded), is that player adapts to work with what he gets, overcoming problems and bolstering boons. I expected red research to carry more "value" in those terms.
Let me elaborate on my vision of how research should be spread:
Green research should (imho) contain necessary things.
There are basic things, which you pretty much have to obtain to progress. There's no substitute for it and you need it. For example a "hut" is such necessary thing.
Yellow research should (imho) contain optional things.
These are the things that give you something you can find a substitute for. Hard to name an example. Lets pretend that, contrary to how it is, the market is less efficient than a gold mine. In this case if I can't access gold mines I have to build markets instead. Gold mine would have been such optional thing.
Red research should (imho) contain powerful things.
These are things that give you advantage strong enough to bend your advancement around. Holsten library for example gives 10 R.P. It takes quite a bit of resources to get this much through other means. Holsten Library is such powerful thing.
Grey research should (imho) contain outstanding things.
These are things that give you advantage strong enough to bend your whole strategy around. Hard to comment on this ones, as I've not encountered such a thing in Elemental yet. The closest thing that comes to mind is SupCom's "game enders", except "game enders" are too strong for this category and waaay too strong in general. Considering the low chance of gray research appearing, it should (imho) cause player to radically change his plans and incorporate his new advantage into his strategy.
How I suggest to change research system:
Introduce "Complexity level".
On the turn of starting next research effort player chooses the level of Complexity of that effort.
Current RP cost of breakthrough represents C.Lv. 3. (Growing as it would normally grow with current system).
C.Lv. 2 takes 75% as much RP as C.Lv. 3.
C.Lv. 1 takes 50% as much RP as C.Lv. 3.
Each individual research (NOT entire Rarity type or tree type but the actual knowledge which we chose after breakthroughs) belongs to a certain complexity level.
When breakthrough is achieved only research of same or lover C.Lv. than the one selected at the start of the effort is eligible for pulling.
The chance of pulling is unaffected by C.Lv.. As long as research can be pulled, it gets it's standard pulling chance.
The C.Lv. of research is determined by the following rules:
Any research not previously pulled belongs to C.lv. 3.
Green research gets C.Lv. -1 each time it is pulled (which means every breakthrough) all the way to C.Lv 1.
Yellow research gets C.Lv. -1 every second pull all the way to C.lv 1. (only pulls count here, so it will stay hard if you get repeatedly unlucky with it, but will quickly slide down if you pass it up several times).
Red research drops to C.Lv. 2 after 3 pulls and stays there. So it gets somewhat easier if you get it often enough.
Gray research only ever appears on C.Lv. 3.
What this means for gameplay:
This system allows player a bit better control of his chances by allowing to skip more common research at a bit lower cost to his advancement than the one we currently have. Payer will not be forced to pay full price for crucial research he rapidly skipped in favor of something more appealing. As such, player will be able to take what he wants when the opportunity presents itself without gimping his advancement too much due to crucial research being put off for later. With current system some opportunities have to be skipped in favor of keeping up to pace with ongoing complication of research.
Have you ever been forced to take green research over yellow and red because without it you can not keep up with growth of current RP cost? That's what I'm talking about and that's what I'm trying to suppress here.
What it means for immersion/realism.
As the country advances, progress becomes more and more complicated, however older technology does not suddenly turn complex. Instead of researching unexplored avenues of knowledge scholars can turn their attention to archives and finish research that was not given due attention or effort in the past.
Concerns:
What I perceived as flaw might actually have been a design choice.
I do not know current method of calculating RP costs, as such I do not know what impact this system will have on current progression or even if it will have any at all.
Research interface will have to be modified. To allow selection of complexity level along with selection of research tree. And to show complexity of available research.
Research pulling/acquisition algorithm will have to be changed to reflect changes in C.Lv and to restrict pulling.
More temporary data will have to be kept (and saved) containing either lists of research per level or individual level for each research. Exact implementation depends on current method of storing information about which research has been eh.. researched and which haven't. If it's ID lists then we are dealing with 3 times more lists of smaller size. If its a pack of booleans we're dealing with variable type change.
Shouldn't be more complications than that, unless of cause my code-fu failed me.
