0.22 My view on how to improve science.

[hebdomad]

When 0.22 landed it was like a breath of fresh air. Even for a KSP veteran like myself, going back to basics was challenging again, but there was a few things that need improving in my opinion. Now this thread is not about discussing the tech tree or what parts go where, but the manner in which you collect science.

Science as you might know is collected through crew reports, surface samples, and running experiments. However it's the limited potential of the science bays that have me worried about the long term, more specificity the 'mid game' and 'end game' scenarios in which these parts are used.

While starting out, you only have the option to get crew related science done. EVA's, surface samples, crew reports and the like. This section is mostly fine as it promotes exploration. The real issues come about when individual experiments roll out. After running the experiment, you have the option to send the data and re-run the experiment, or recover the experiment. With my limited time with 0.22, it seems there is no advantage in returning experiments, even with the loss of data factor, as you can just re-run the experiment and send it again.

There is also the amount of different experiments that you can run. There are seven experiments you can run. Total. Compare that to the International Space Station that has hundreds of experiments running at this very moment. With the limited about of experiments available, it makes space stations inefficient science wise. One single rocket launch can take care of most if not all the science available in low Kerbin orbit. At this stage, gaining science is like mining. You go to one location, squeeze all the science out you can and then never return. This is bad for replayability.

On the subject of science drying up, all flights of significant duration should be rewarded with some science. I can see a lot of new players getting stuck not being able to unlock new parts while having depleted all the science in areas in which they can reach. Players should be rewarded even if their rockets explode mid flight.

I believe that what needs to be done is that...

Science units need to have a limited if not single use only.

> A soil sample experiment might be able to run five times with a bonus for different biomes and distance apart, as well as a value based on how much of that soil you've collected. A crystal or mystery goo experiment should only be able to be run once. Experiments should never really return zero science unless they have been over done to a ridiculous level.

Generic science bays.

> Surly the Kerbals in the labs have more than seven experiments. A one time experiment that takes a few days to complete would make space stations viable. The same return or transmit mechanics would work fine here also.

Science needs to be transferable between ships via docking.

> People want to live out their Apollo dreams. Being able to transfer science documents while docked would do wonders.

Now I've listed some problems and some solutions. What I've not covered is the end game section of KSP. I am still waiting to see how Squad implement the missions section of Kerbal Space Program. Will there be missions on every body and biome? Will there be generic missions like launching weather satellites, colonising planets or building space stations? I guess we will have to wait until 0.23 ( or 1.0? ) to find out.

(note this is a rushed first draft will come back later to fix spelling and grammar )

[Proply]

How I would do science!

Some nice ideas, but first off let me say that I really like the latest update and am greatly enjoying the new science system! Also it should be noted that I am aware that the development of the science system is an iterative process and the current state may only represent part of the final vision the devs wish to implement. Anyway, as I see it at the moment here are the three main problems with science (ignoring tech tree balance which is a massive topic in its own right), and here are the solutions that I would use [Disclaimer: all examples are for illustration purposes only!]:

TL;DR VERSION:

-Rework antennas so higher tiers are more efficient at longer distances

-Make partial science recovery from transmission non-recursive

-Rework Goo/Materials bay to enable science generation on space stations

Antenna mechanics

At the moment there is no point using the high-tier antennas, as they use MORE power to transmit the same amount of data (albiet a little faster, but that is hardly a major balance factor). My solution to this would be as follows:

-Change the power requirements from charge/Mit (Mit = unit of data) to a charge/second of transmission time, and scale this charge/second cost so that it increases with antenna size (eg: 10 c/s Tier 1, 20 c/s T2, 30 c/s T3).

-Either make Mits/second transmission rates the same across all dishes, or scale them so that the higher tier dishes are faster, but still less energy efficient (eg: 10 Mit/s T1, 15 Mit/s T2, 20 Mit/s T3).

-Give each dish an "attenuation" value that defines the amount the the signal transmission rate decreases as a function of distance from KSC, with higher tier antennas having lower values (and therefore less signal attenuation at longer distances).

The end result of this would be that the T1 antenna would be the most efficient for short range transmissions (say within the Kerbin system), but T2 and T3 antennas would be more efficient at longer distances, letting each fill a distinct niche in the science transmission mechanics. This would also allow for the potential creation of relay mechanics at later date (eg: relay satellites that, when placed between the transmission source and KSC, would reduce the time required for transmission, and therefore the power required).

Transmission mechanics

Currently from a clinical gameplay perspective there is no point in doing return missions (after getting solar panels to replenish energy supplies at least), as the transmission system allows you to generate roughly the same amount of science by simply transmitting the same information many times, making it more efficient of your time to send one-way missions. To be fair this has been somewhat addressed by the vessel recovery bonus, and may be somewhat remedied in the future by part recovery with currency and/or addition of penalties for not returning kerbals; however I still feel that this could use some tweaking. At a basic level I would:

-Change the transmission mechanics so that a 20%/40%/60%/whatever return on transmitted data is not recursive; once the data is transmitted, the value is removed from the total science pool for those sample conditions, and the remainder can only ever be recovered by a return mission.

For a more comprehensive overhaul I would do the following:

-Change the science recovery rate from transmission of experiments such as thermometers, seismometers, barometers, etc, to 100%.

-Remove the Goo and Materials bay experiments (these parts will be repurposed in my third section), and replace them with a camera (100% transmission return but high data packet size).

-Change the science transmission recovery rate of EVA reports to 100%.

-Change the science transmission recovery rate of surface samples to 0% and increase yield.

-Add a rock drill to take surface samples on unmanned probes, still 0% transmission yield and only 25-50% science value compared to surface samples taken by a kerbal.

-Add modifier parts that do not provide science on their own, but instead increase the transmission yield of certain other experiments (for example a spectrometer part that increases rock drill sample transmission values from 0% to 40%, could also work with atmospheric samples or liquid samples).

These changes would enable useful low-science missions for probes with passive sensors, med-science missions for robotic landers/rovers with sampling tools and modifier parts, high-science missions for robotic sample return missions (that don't require the added weight of modifier parts, but lose everything if they don't make it back to Kerbin), and very high science missions with manned return landers, with each successive mission type obviously costing more to execute (higher cost/risk, higher reward!).

Lack of science from space stations

Currently a space station is not really any use for generating science, and I have some ideas to change this based around repurposing the Goo Canister and Material Bay parts:

-Goo canister and materials bay become science experiments that slowly accumulate science capped at a max value as long as they are in space and attached to a manned vessel, these would not be resettable.

-Differentiate between the two by having a 100% transmission value for the Goo, but low total yield, and 0% transmission value for the Materials Bay, but a higher yield.

-Make each kerbal on the attached vessel be able to run only a small number of experiments at once; if there are less kerbals on the vessel than the amount required by the number of experiments attached, the rate of science accumulation is proportionally reduced across these experiments.

This would enable science research on space stations; experiments would be flown up, docked to the space station, managed by the resident kebals, then either transmitted + discarded (in the case of the Goo Canister), or returned to Kerbin to harvest that sweet science in the case of the Materials Bay. This would be intended as a slow but steady way to accumulate science that requires assembly of a space station, and transfer of the science modules to/from the station, but would not require constant oversight while the science is accumulating (so you can go off and manage your other missions!). An attenuation value could also be added to each kerbal based on time spent in space that reduces the research accumulation rate the longer the kerbal stays in space to give a mechanism encouraging crew rotation.

-Thanks for your time! Proply.

Edited October 24, 2013 by Proply

[boolybooly]

Personally I just think the total cost should be higher but spread out, so instead of researching 5 components in one go, you research a component node with one main component and then individual components sprouting from that node.

So say you unlock a satellite core node for 300, you can then unlock a solar panel for an additional 100 and a mini fuel tank for 50 and a small rocket for 40 and a circular battery for 60 etc. So in the end the entire node costs you more like 600. So you can then either research the parts you want or you can move up to the next node.