Part refinement through R&D, or the scientific endgame

[SorensonPA]

The new science system has considerably changed the way in which one plays KSP, but for all of the impact of said change it is, nevertheless, ultimetily limited: given enough time, one will eventually conduct enough research to unlock the entire parts catalogue, and from there the science system effectively ceases to exist. This is a considerable shame considering the much-needed structuring and prodding it gives players and the potential of such a system. How, then, to keep the player incentivized to continue conducting experiments?

If the purpose of science is to unlock better parts, then, the answer is to let the player to go one step further: create better parts by utilizing and improving upon their predacessors.

To enable such a system, one first needs a metric of the accumulated scientific knowledge in each of the core part fields of command, propulsion, control, structure, aerodynamics, utility and science: the Science Rating. Each individual technology in the tech tree pulls parts from one or more of these fields, and every technology developed increases the Science Rating of these fields by the science point cost of the technology. Thus, the further one goes along the tech tree, the more developed their knowledge of each core part field as reflected by the increased Science Rating. At the absolute end of the research tree would be a series of new technologies devoted to the research and development of new parts. Once unlocked, these technologies enable the player to do two things: invest science points into a given field's Science Rating and develop new parts.

Part development allows you to copy the basic perameters of an existing part and then modify them in accordance with your part's Science Rating: cost, mass, thrust, resource output, isp, capacity, fuel components and ratios, impact tolerance, max temperature and many factors. Each perameter has an effective cost in Science Rating and each perameter can be improved or sacrificed to offset the Part Rating in a desired way. Not all factors can be changed, however: the part model, scale, functionality and drag remain locked, and model and scale have their own impact on the part's rating, larger parts being generally easier to work with compared to performance-identical miniature counterparts. As long as the Part Rating is equal to or less than the part field's Science Rating, the part can be detailed, confirmed, stored in the savegame and then used.

As an example, let's take the Mainsail: thrust is perfectly fine, but isp is somewhat somewhat disappointing. We could bring it to 320/400 without changing any other factors, but that would increase the Part Rating and might push it over the propulsion Science Rating. At the point where you get Mainsales, you've also got access to solar panels and large-scale batteries, so sacrificing the electricity generation doesn't hurt much and balances out the isp change a bit. It's also on the heavy side, so we could reduce its mass to 3 and then bump up the unit cost to 3,000. If we really want to mess around and flaunt our mastery of propulsion, we could introduce ElectricalCharge as a fuel component and turn it into a petrol-fueled z-pinch drive or crank up vacuum isp and cut down thrust and atmosphere isp to make a 2.5m NERVA. And after we use the new not-a-Mainsail to gather more science, we can clone it, improve it and repeat the process again!

There's only one caveat to this system, however: the development of new parts is theoretically infinite, but the gathering of science is very much finite. Once all the experiments have been done in all the qualifying experiment zones, what then? There are a few ways to get around this impasse:

* Allow for passive over-time science income by deploying probes, ships and stations around bodies and in space and drone rovers and manned bases on bodies

* Earn science and cash by conducting generated missions to do certain things (launch a probe out of the system on a certain vector, do a fly-by of a specific body below a certain altitude, land a probe or kerbal on a specific body, plant a flag at a specific location, land and return from a specific body)

* Buy it outright with extra cash

The science system gives some much-appreciated structuring and prodding to motivate players to explore and unlock everything they can, but an open-ended part development system will extend that structuring and prodding into nigh-infinity. Learn to develop, develop to learn, each step greater than the last.