Long-Term Experiments/Science overhaul

[AlexTheNotSoGreat]

I'm sure we've all considered at some point that the way science works in KSP can be a bit...   dull

In the game, you can send a probe to a planet in orbit or on the ground, use your instruments once, and your done, aside from occasional contracts.

This idea is meant to allow more long term and in depth science experiments, both from already existing parts, and perhaps adding a few new ones.

The idea is that you could do short or long term reading in a planet's orbit or surface. Let's say, we want a map on the variances on Duna's gravity, or the surface temperature at a location around Laythe. You could set up a probe to do measurements of either the whole planet from orbit, or from an area on the surface. You could then send back science depending on what area you measured and what you wanted to measure. Back to the examples, you could, for instance, place a satellite to measure gravity around the equatorial regions of Duna for n days, and send it back for x science for mapping y amount of Duna's gravity (and the science value would vary depending on it). You would also get an actual map (pre-made in the game, of course) of the gravity fluctuations when done. Same for Laythe, send a probe to land on it and measure temperature fluctuations for a decent time time, then get science based on the data gained. Yes, there are contracts that do this, but this wouldn't be to, again, avoid just clicking and using an experiment once, to be more in depth, and to make science more interesting. And of course, these are only a couple examples to explain the concept. On top of that, some new parts, like advanced magnetosphere instruments, cameras and alike could be added for similar purposes.

Putting it simply, this could make science in KSP much more in depth for some many players, and has been talked about for as long as science gathering has been a thing in the game. I hope this is put into some consideration.

-Alex

Edited September 19, 2016 by AlexTheNotSoGreat

[Alshain]

The problem with this type of system is that it is time based.  Why bother doing it for n days when I can timewarp that and more.  You might as well just give me the science in an instant, which is what the game does.

I've considered the possibility of this type of system before, but in the end, time based mechanics do not work in KSP.

Edited September 19, 2016 by Alshain

[AlexTheNotSoGreat]

Now, it wouldn't be purely time based. If you've heard of Orbital Survey Plus, you have to actually orbit over a given area to get information, as opposed to getting it all in one transmission. It's the same case here, you would have to map a certian area to get the science. Yes, you can still time warp, but the point is not to spend time taking measurements. The point is to make you think about how you place satellites and landers a bit more in order to get science rather than click "detect gravity high above _______" and never do it again. It's to make you think just a bit more, as I said in the main post.

[The_Rocketeer]

My tuppence-worth:

4 minutes ago, Alshain said:

I've considered the possibility of this type of system before, but in the end, time based mechanics do not work in KSP.

I disagree with this statement, I think that crucially time is the missing element in all KSP science. RL experiments take time to plan, to conduct, to repeat, to analyse, to encode, to transmit, to decode, to interpret, to prepare, peer-review and publish findings. Now, ok KSP has timewarp so that's all time u can skip if u want to, but for those who like the idea of running a space program timewarping should be something to be avoided as much as possible. Every experiment should have a series of interconnected stages of development, from planning a science mission to publishing findings and receiving rewards, each of which has to be greenlighted, have resources committed and take time to process.

I think the answer here is to mathematically tie funding to 'science points per time interval', with a formula to modify your budget gradually based on trends in your science delivery. I'm no mathematician, so forgive my layman's terms, but for example one interval of no science shouldn't immediately cripple your budget, but perhaps 3-5 intervals would see the budget seriously handicapped. A suitable interval length might be a quarter or half of Kerbin's orbital period. Incidentally, I see no reason to scrap contracts funding, but they should be the cream off the top, something with some panache (like tourism), while the science work is the staple diet of any space program.

[Laie]

On 9/19/2016 at 2:37 AM, Alshain said:

Why bother doing it for n days when I can timewarp that and more.  You might as well just give me the science in an instant, which is what the game does.

Radio Yerevan: in principle you're right, but I think science-over-time can be palatable and even a good gameplay element if you frame it well. Have you ever used Scansat? Mapping a planet takes time, and yes, you can timewarp right through it. Yet it feels right, while the stock method of just handing over the map as soon as you entered polar orbit is jarring.

The better you can give the impression (1) that something is actually happening while you wait, the better it will work. Scansat is hard to beat in that regard, I can't think of anything that would come even close. But generally, setting a timer when deploying an experiment and providing the science points when $INTERVAL has passed would be bad. A solar observation satellite that yields a few science points every couple of weeks might go over better. Or a seismometer left behind on the Mun.

Bonus point if you can frame it so that people don't feel like they have to timewarp-milk it for all it's worth, but can just accept it as something that keeps going while they do whatever they want to do next.

--

(1) or illusion, as the case may be. Because fundamentally you are right about time-based mechanics in KSP. The point I'm trying to make is that an illusion can still be enjoyable.

[Racescort666]

@Laie SCANsat is awesome for science but it does something more than just give you science points, it gives you science that's useful. There was a thread a while back about revamping science and while there was a bunch of suggestions on how to change the science system the best suggestion I saw was making science useful beyond science points. 

SCANsat is easily the best example of this but I think it could be creatively applied to the other science tools as well. EG: Mach number readout for doing temperature scans in Kerbin's Atmosphere or something. It adds to the UI, which a lot of people don't like, but it's useful information that's collected by doing an experiment.

Anyway, the point is, if science points are just another form of currency, it will be probably be a grid to collect them regardless of how it's done.

[Jonfliesgoats]

I love the idea of overhauling the science system.  I do like science points to spend on development, however.

I think long term experiments, more sensors and more dynamics to measure would all greatly increase the variety of missions you can do in career mode.  Earlier I mentioned adding speed and acceleration dynamics for testing and science.  

Long term measurements are a cool paradigm and we already have a model for doing similar things.  We scan for more by putting a probe in a polar orbit and it magically gets the data we need.  We don't have to wait for a thousand or so orbits to build our data.  We could do something similar for long term orbital surveys.

For long term surface data, what about sensors that have to maintain stability for a given amount of time (so many days).  These would yield more science, but they have to be oriented just so (vertically, horizontally, etc. so a tipped over lander wouldn't cut it).

Anyway, many more science options and much more expensive tech progression would create a more varied game experience.

[Laie]

3 hours ago, Racescort666 said:

SCANsat [...] does something more than just give you science points, it gives you science that's useful.

Yes, that's hard to beat or replicate (building an atmosphere model from temperature and pressure data? puh-lease). But I don't think it's necessary or a requisite for science-over-time.

[vasya pupkin]

Besides the obvious part, that IRL a scientific research is even more repetitive and boring, there is another consideration IMHO.

Rephrasing what authors of the game said more than once - this is not a game about research of science (in this case, and there are others), but about launching a rockets.

Whatever authors/developers decide to incorporate into KSP, IMHO, they always keep this in mind.

The more complicated some additional content, the more this draws attention away from the main theme. So authors took a decision to enhance a gaming experience by adding a science concept, and made this change as simple basic ability. Just to an extent of keeping it from overshadowing the main theme.

Here is an additional example of similar suggestion, which was brought to attention previously: Repetition isn't necessarily tedious, but science is. Why, and what should we do?

[Jonfliesgoats]

Vasya makes a good point.  Unchecked people like me will have Kerbals writing grant proposals, running repetitive titrations, etc.  Kerbal grant proposals would be adorable though.

"Rockets.  Give me money.  Rockets give me money."

[AlexTheNotSoGreat]

11 minutes ago, vasya pupkin said:

Rephrasing what authors of the game said more than once - this is not a game about research of science (in this case, and there are others), but about launching a rockets...

People most likely said the same thing with 1.2 and communications networks, making the game too hard. Like the system made in 1.2, this revamped science system wouldn't be hyper-realistic, with analyzing and properly making in-depth observations, but would still display at least some form of realism. The whole reason for my concept is to eradicate the very simple "click and send" science system in KSP currently, and make something at least a little bit more "enriched," if you will.

7 hours ago, Jonfliesgoats said:

I love the idea of overhauling the science system.  I do like science points to spend on development, however.

I think long term experiments, more sensors and more dynamics to measure would all greatly increase the variety of missions you can do in career mode.  Earlier I mentioned adding speed and acceleration dynamics for testing and science.  

Long term measurements are a cool paradigm and we already have a model for doing similar things.  We scan for more by putting a probe in a polar orbit and it magically gets the data we need.  We don't have to wait for a thousand or so orbits to build our data.  We could do something similar for long term orbital surveys.

For long term surface data, what about sensors that have to maintain stability for a given amount of time (so many days).  These would yield more science, but they have to be oriented just so (vertically, horizontally, etc. so a tipped over lander wouldn't cut it).

Anyway, many more science options and much more expensive tech progression would create a more varied game experience.

You seem to get the idea I'm wanting. A more diverse and dynamic science system is what I'm meaning, yet still able to "Keep It Simple Stupid."

 

9 hours ago, Laie said:

Radio Yerevan: in principle you're right, but I think science-over-time can be palatable and even a good gameplay element if you frame it well. Have you ever used Scansat? Mapping a planet takes time, and yes, you can timewarp right through it. Yet it feels right, while the stock method of just handing over the map as soon as you entered polar orbit is jarring.

The better you can give the impression (1) that something is actually happening while you wait, the better it will work. Scansat is hard to beat in that regard, I can't think of anything that would come even close. But generally, setting a timer when deploying an experiment and providing the science points when $INTERVAL has passed would be bad. A solar observation satellite that yields a few science points every couple of weeks might go over better. Or a seismometer left behind on the Mun.

Bonus point if you can frame it so that people don't feel like they have to timewarp-milk it for all it's worth, but can just accept it as something that keeps going while they do whatever they want to do next.

--

(1) or illusion, as the case may be. Because fundamentally you are right about time-based mechanics in KSP. The point I'm trying to make is that an illusion can still be enjoyable.

Same here. You get a better feeling out of when you get your science with mapping and long term scanning.

[The_Rocketeer]

@AlexTheNotSoGreat

I love the idea of making outposts, bases and stations that are essentially 'science farms', with some sort of production synergy depending on which experiments are attached to the craft and how many and which astronauts are aboard. The MPL sort of does this, but in a really dissatisfying way. What I'd like is less of the 'run the experiment and send the results' approach, and more of the 'place an experiment and return later to see what you've learned'.

Another idea I really like is 'modular experiments' instead of 'experiment modules'. This would sort of marry contracts and science experiments into a sort of puzzle that uses the actual data from various experiments in various situations to solve. The closer you get to the precise answer, the better your science reward. That way you reward the player for repeating experiments without insisting on it - pursuit of a perfect score is a choice, but if you just want to rush it and finish fast you just fulfil the basic criteria for a less-substantial reward, without worrying too much about (for example) flying at the perfect speed at the perfect altitude, or being exactly 500m from a biome boundary.

[Jonfliesgoats]

A cool long term experiment: greenhouses!

[Wjolcz]

I just want the experiments to gather data as long as they are active. They consume EC anyway.

[AlexTheNotSoGreat]

On 9/21/2016 at 7:01 PM, Jonfliesgoats said:

A cool long term experiment: greenhouses!

That would be nice if when life support is implemented.

 

On 9/20/2016 at 7:01 PM, The_Rocketeer said:

@AlexTheNotSoGreat

I love the idea of making outposts, bases and stations that are essentially 'science farms', with some sort of production synergy depending on which experiments are attached to the craft and how many and which astronauts are aboard. The MPL sort of does this, but in a really dissatisfying way. What I'd like is less of the 'run the experiment and send the results' approach, and more of the 'place an experiment and return later to see what you've learned'.

Another idea I really like is 'modular experiments' instead of 'experiment modules'. This would sort of marry contracts and science experiments into a sort of puzzle that uses the actual data from various experiments in various situations to solve. The closer you get to the precise answer, the better your science reward. That way you reward the player for repeating experiments without insisting on it - pursuit of a perfect score is a choice, but if you just want to rush it and finish fast you just fulfil the basic criteria for a less-substantial reward, without worrying too much about (for example) flying at the perfect speed at the perfect altitude, or being exactly 500m from a biome boundary.

This all sounds great. I did have the MPL's way of science in mind with this, but with general experiments at lesser degrees. So a crew report could give small amounts of science per day, but for, say, a year, and you get decent science (0.05sci/day LKO = ~21.3sci/year). And the concept of experiments working to answer a large question is brilliant! Geological activity on Mun with seismic/gravity scans, weather patterns on Duna with a few separate landers in different areas with barometers sensing for periods of time, and so on.

On 9/22/2016 at 2:01 AM, Veeltch said:

I just want the experiments to gather data as long as they are active. They consume EC anyway.

I feel like this is already done in a way with the EC to transmit the info, but still a nice detail piece.

[AlexTheNotSoGreat]

*Buffer post*

[smotheredrun]

DMagic Orbital Science contracts, and ScanSat

[AlexTheNotSoGreat]

On 10/6/2016 at 8:32 PM, smotheredrun said:

DMagic Orbital Science contracts, and ScanSat

For more instruments, this is what I was considering

[Jonfliesgoats]

More instruments and more variety in imaging gives us more reasons to launch rockets and spacellanes in various configurations.  

[endomen]

A system where you could take pictures of things and then pick out interesting things in them would be good. A system where in career mode, returned surface samples were kept and could be analysed further as you upgrade R&D could also be useful as a longer term experiment, as the samples from Apollo are still being studied today. And ScanSat should be made stock.

[Jonfliesgoats]

Agreed.  Kerbnet offers the ability to find anomalies, but I have not experimented with it to see how it "feels" to ,e as a player.  

Imaging along with some basic GIS tools in the tracking station would be neat too.  It would be nice to have slopes beyond a certain gradient highlight in red, for example.  Relief shading provides this data in other formats, but it is still difficult to identify locally dangerous slopes.