Galactic Armory 2.0 Research explained
One of the most extensive changes we are planning for the next major release of Galactic Armory is a complete overhaul of the existing research. This includes the technology tree of the game as well as the technology curve, better known as level gain curve.
In its current state, research in Star Ruler (and Galactic Armory) is neither really exciting nor very rewarding. The player basically queues his research (or lets the automated research feature do its thing) and simply waits until everything is unlocked. The flaws of this system become even more obvious when general science is queued repeatedly in combination with turning most of your planets into research worlds. What follows is usually sending out unstoppable fleets or even single ships that can clear out most of the galaxy from hostile forces without much resistance.
Why is that so? There are several reasons. First, to improve your existing technology you have to research one or multiple technologies. Repeatedly. However, this usually also unlocks a large number of more advanced technologies. In most cases this more complex technology is, despite drawbacks and increased cost, often almost a straight upgrade from existing technology. Long story short: There is enough reason to make players WANT to upgrade everything because it is easy AND rewarding. This is made even worse by the fact that general science has a huge impact on research rate as well.
Another reason for the aforementioned issues is the exponential levelling curve. Not only does research easily unlock dozens of new, better subsystems, it also increases everything by a whopping 35% over the previous level (default setting). The differences in overall fleet and empire strength resulting from even very small differences in research rate (and focus) often make intelligent ship design and efficient empire management irrelevant. Whoever has the most research wins more often than not. Even economical strength is mostly based on reserach, so no amount of good leadership will save you against a technologically superior foe. Is this fair? And is it fun? We believe this is not necessarily the case and have decided to change both the nature of the tech tree as well as the overall impact research has on the game.
Brace yourselves. Change is coming.
There are three main goals we want to achieve by redesigning the research in this game:
1. Research must allow for and reward specialization. Unlocking each and every last thing should not be expected in an average game.
2. The structure of the technology tree must be easy to understand.
3. Research should be important, but never be the single factor deciding over victory or defeat.
"Research must allow for and reward specialization." What does it mean? Simply put, subsystems will be spread out more over a larger tech tree. Not necessarily by level - we aim for a rather flat unlocking progress curve - but by research node. This makes it less likely that every player will end up with everything unlocked over the course of a game. Research takes time, and we want to set up research in such a way that it may be desirable for players to focus on few fields of research instead of researching a little of everything.
In addition, though it will still take some levels of research to unlock, we want more advanced subsystems to be more than just straight upgrades of previous ones. This will also require an extensive rebalancing and overhaul of all subsystems. One aspect of this approach is giving players more choices about what they actually want to upgrade instead of a simple "I'll take everything at once" approach. You want more precise weapons instead of only going for increased damage? No problem. Need more range? Or rate of fire? The choice is yours.
"The structure of the technology tree must be easy to understand." This one should be pretty self-explanatory, but it deserves a few words since it is so easily gotten wrong or forgotten about altogether.
When creating a mod or game, it is surprisingly easy for a developer to get lost in complex mechanics and formulae. In fact, this can make one lose track of seemingly obvious things - like making things understandable and accessible for someone outside the development team. While our research tree is still considered work in progress, we have already gone through several different layouts and versions during the concept phase until we managed to come up with one that isn't overkill, i.e. consists of several hundred nodes.
What are nodes? Basically each research node is either one of two: A root node or a branch/sub node. A root node is the main field of research for a whole set of fields. Currently each set consists of one root node and three sub nodes. For most subsystems, root nodes and sub nodes don't differ that much; research root nodes unlocks new subsystems, which also holds true for many sub nodes. Weapons, however, are a different matter. Weapon root nodes handle all of the unlocking for a set of weapons (i.e. beam energy weapons). They also improve some of the most basic features, which includes damage at the very least. Weapon sub nodes on the other hand do not unlock new subsystems but mainly serve to improve specific properties of a weapon system. The actual number and type of features improved by each weapon research node may vary between weapon groups, giving each weapon type more character.
"Research should be important, but never be the single factor deciding over victory or defeat." Like we examined before, research currently plays a major role in the game, often make designing clever ship layouts and countering enemy layouts completely moot. Sadly this does not do Star Rulers combat mechanics and blueprint design justice. Because we feel there is so much more to this game that is often neglected or simply not seen due to very few but far-reaching issues, we have decided to change this aspect of the game to something that favors the strengths of the game mentioned before. This is achieved by changing the exponential level gain curve (the formula that determines how much stuff improves per level) to a flat increase modifier. Though we haven't decided on specific values yet, here is a quick example to show the difference:
Level: Exponential @1.35 Linear @1.2
1 1 1
2 1.35 1.2
3 1.82 1.4
4 2.46 1.6
5 3.32 1.8
6 4.48 2.0
7 6.05 2.2
8 8.17 2.4
9 11.03 2.6
10 14.89 2.8
So, in a nutshell, the difference compared to level 1 becomes huge with an exponential system - especially at later levels. In this example, a 35% exponential increase would result in a 14.89 times higher value at level 10 than at level 1 - compared to just 2.8 times the original value with a flat 20% increase per level.
All of the above should be considered work in progress for all intents and purposes. All of these mechanics is still subject to change, but we felt this was a good time to give you a general outlook of things to come. We hope you enjoyed this update and look forward to the next one!