Science in KSP2

[Pthigrivi]

One mostly overlooked detail from the Ep. 4, we got a little bit of news about Intercepts' thinking on the science system. From Shana's explanation about topology and game design (around min 6:30) it sounds like the basic structure of biomes, experiments, and gathering science to unlock parts will probably be carried into KSP2, with added emphasis on tying these biomes to interesting and unique geological spaces, which sounds great. This is echoed a bit by Tom's mention of 'goodies' to be found in Dres' equatorial mountain range (could be ISRU goodies as well). I know it's been a somewhat controversial topic in KSP but Im personally pretty happy to see science continue to be important to exploration and development. At the same time I think a lot of people would agree science in KSP1 had some big issues, primarily clickiness and grind, that would need to be resolved. There were also some issues with the structure of the tech tree itself. There's also the factor of KSP2's vastly expanded scope both in terms of planets and new technologies that would need to be taken into account. 

For feedback I thought we could talk about the pros and cons of science in KSP1 and ideas for how KSP2 could make something great out of it. In all likelihood they've already got a pretty strong feeling about it but its fun to speculate. First off I think there are subtle but important pros to the biomes + point-buy system: it's an open way to reward players for visiting new places and gives them the flexibility to develop as they choose. You're not tied down to a linear set of tasks to get the parts you need. Whether you go land on Minmus first or orbit the Mun or do a sub-orbital mission to Kerbin's poles you're still making progress. I also just like it conceptually, that science and doing experiments and discovering new places are critical to actual gameplay. A lot of people have proposed radically different solutions to the science system but few that I've seen seem to achieve these pros while making the game more fun. 

Then, sadly, there are the cons. The most common complaint I see is a general sense of grind, but there are a lot of sources for this feeling. Right off the bat there's gameplay and UI issues. The way the experiments and biomes are structured players are begged to right-click several very small parts, sometimes hidden inside service bays, and then individually dismiss half a dozen readout screens, often in mid-flight. It's a huge distraction and even when hot-keyed and becomes tedious really quickly. The second problem is a pacing problem caused by the shear volume of science and potential experiment repeats. Besides asking players to manually conduct science while landing and taking off it encourages a huge number of repeat missions in KSOI leading to a near complete tech-tree before their first interplanetary mission. Then there's the tech tree itself, which is a mess, and a host of other minor issues with science labs, balance, lack of mapping, etc. 

Given all that I do think it's salvageable and with some work could be really fun and great. This post is already getting long so I'll let others chime in before breaking down some possible solutions.

Edited November 2, 2021 by Pthigrivi

[Tw1]

Personally, I would rather science to be more passive in a lot of cases.  You shouldn't need to trigger a temperature sensor at a specific moment, just remember to have the thing turned on. My major turn of for the KSP1 science system was how much it felt like fetch quests, rather than actually doing science.
My dream for that game was to have more things which you'd actually get data from, values which change per location, stuff you could plot, etc.

The other problem was how much it was linked to scoring simple points, rather than science for the sake of science. Not against it helping you get somewhere, but I hope it has more emphasis on the building-up-an-understating-of-the-planet side of things this time. I want to see more graphs and charts to complete.

Hopefully this time, using your science to get somewhere might mean learning what areas are easier to travel through in rovers, or what hazards you might encounter, or what resources there might be on the surface. More of that please, less of "This arbitrary amount of points lets you gain one unrelated set of parts".

The issue for me with both science and tech tree was how unauthentic they felt compared to the space travel side of things. They broke the illusion, and made it clear I was just playing a game. I hope it appeals more to the imagination this time.

Edited November 3, 2021 by Tw1

[shdwlrd]

I remember science being discussed at length in the forum in 2019, 2020 and earlier this year. @Pthigrivi you have listed your preferred fixes in each one of them. Because the devs mentioned biomes and science system in the same sentence doesn't open any new insights into how the science system will work. They basically confirmed that biomes will be a thing in KSP2 and you will be able to get science from them. That's it, nothing else was confirmed. Do we need to have another science gripe and fixes thread?

[Pthigrivi]

1 hour ago, shdwlrd said:

I remember science being discussed at length in the forum in 2019, 2020 and earlier this year. @Pthigrivi you have listed your preferred fixes in each one of them. Because the devs mentioned biomes and science system in the same sentence doesn't open any new insights into how the science system will work. They basically confirmed that biomes will be a thing in KSP2 and you will be able to get science from them. That's it, nothing else was confirmed. Do we need to have another science gripe and fixes thread?

Wow, brutal. But yeah to say that there will be biomes, experiments, and science to unlock the tech tree does confirm there won't be any radical rethinking of the process, so that's interesting at least. It's one of the small bits of info about gameplay that they've let slip, and cool to hear about how the art and geology could be related. I know we've talked about science before but it's been a while and I don't actually recall what my suggestions were? Im also curious if folks have shifted in their own thinking about this? Like I said they've probably made a lot of progress already but its still fun to speculate.

Edited November 2, 2021 by Pthigrivi

[TLTay]

I do hope they make science more interactive. I'm sure they've already set a path on science by now and won't be changing much at this point. I'm sure it will be modifiable if it's not ideal. 

[Zhetaan]

Aside from a general hope that they go with region over biome (which I think is rather a silly descriptor for a portion of a lifeless, airless moon), I do like @Tw1's general thinking that science should feel more like a process than a game of fetch.  I can't speak to ease of implementation, but I think that there's something to be said for splitting out science from a general collection of points to something a bit more subject-based.  I'm not looking for something so involved as to be confused for the real thing, but I don't see why taking the temperature of the ocean (or, for that matter, recovering a rocket for the first time) should help me build a more efficient rocket engine.  Perhaps have a few broad categories, and having running experiments (or recording, in the case of sensors) in a given region will help to add to the collective body of knowledge in that subject.  Linking the tech tree to interrelated fields of science, rather than an arbitrary point total, would probably help the thing feel a little less slap-dash, too.

However, though I cannot speak to ease of implementation, I can say that this would probably become unwieldy in a hurry, which is unfortunate.

[Ahres]

I haven't given it much thought while I'm sure some of the other KSP2 science speculation threads have, and I'm not sure @Pthigriviwas wanting to go this direction, but I'd love a science system where scientific observation is useful for future missions. An example would be something as simple as atmospheric height. I hope when I go to my first extrakerbolar planet I don't know a thing about it (except what can be gleaned by telescope), even how high the atmosphere reaches above the planet's surface. I want my science gathering to give me that information - I want exploration and discovery and reconnaissance to be way more involved.

And then along with that I agree with the OP, let's get rid of the grindiness and give the planetary science some meaning. In KSP1 we can actually turn on the display of a thermometer and see what temperature the vessel is at, but who bothers to look? It doesn't do anything, it's not recorded anywhere so we can compare against other biomes (or regions, @Zhetaan), or other planets. 

So what's the solution? I don't know. Let's see what Pthigrivi has in mind.

[t_v]

Well, I think that this discussion is different enough from the others in that there will be no radical redesign of the science collection system. To begin, I agree that some form of automatic collection is vital to avoid a game-breaking grind. So most of this discussion will be about how the tech tree is organized. 
 

I love how versatile the point-buy system is when deciding what to specialize in, but it becomes very hard to balance science with progression, especially with labs. KSP 1 play throughs with a lab unlocked will end as soon as a trip to duna brings back its first haul of data, and even sooner if players revisit the mun or minmus. Ideally in a play through, I would like to have to visit the jool system at least once before finishing science progression. With the expanded KSP 2 tech tree I’m sure that players will have to go to more places to finish the tree, but it could potentially be completed by the end of the first interstellar voyage, depending on scaling. And there is no good way to balance this with a simple point-buy system. 
 

My solution: have different planets unlock different places on the tech tree. Essentially, each planet can offer science but the tech tree is broken into sections so that you can’t rush to endgame content prematurely. For example, gas giant science would be good for advancing ion/plasma propulsion. Kerbin would allow you to unlock initial technologies in every field, and gas giant planets further in progression allow more advanced technology in ion propulsion. To avoid the chore of going back to an easy destination to get the science that you didn’t collect during that phase of progression, going to a late-game planet will give you the science to fill that branch of science in. This makes choosing where to go more important depending on what you are trying to research next, and forces you to visit types of planets you might not consider otherwise, thereby getting the full experience. 
 

for a list of prospective planet-subject pairs, here is what we could do:

kerbin, the mun and minmus all provide basic science up to a certain level

Duna allows for the next early level and starts going into xenon propulsion, nuclear propulsion, and space labs. 

Eve-like planets and high gravity planets all provide more powerful propulsion tech, with more and more powerful lifters as you go on.

Very small planets and planets with just rock and no atmosphere help develop bigger structural parts and containers, allowing for mega stations and bases. 

Gas giants will help with ion propulsion (we saw a Hall effect thruster in the latest dev log so they definitely are adding more ion stuff)

Hot planets will help with heat dissipation, starting with mojo and probably ending with charr 

Kerbin-like planets will provide general science. 
 

once a “branch” of the tech tree has been completed, additional collected science can spill over into general science

the philosophy behind this particular solution to the science problem is that it feels too gamified and with not enough real depth. Breaking up the tech tree is just one way to add depth and I’d love to receive a better idea because this one is flawed.  

[coyotesfrontier]

This might be an unpopular opinion, but I love the science system in KSP. It gives you a reason to explore all areas of a planet, build surface bases, and build orbital stations all in a really simple and elegant way. It can definitely be improved in a few ways, but it's still great already.

4 hours ago, t_v said:

for a list of prospective planet-subject pairs, here is what we could do:

kerbin, the mun and minmus all provide basic science up to a certain level

Duna allows for the next early level and starts going into xenon propulsion, nuclear propulsion, and space labs. 

Eve-like planets and high gravity planets all provide more powerful propulsion tech, with more and more powerful lifters as you go on.

Very small planets and planets with just rock and no atmosphere help develop bigger structural parts and containers, allowing for mega stations and bases. 

Gas giants will help with ion propulsion (we saw a Hall effect thruster in the latest dev log so they definitely are adding more ion stuff)

Hot planets will help with heat dissipation, starting with mojo and probably ending with charr 

Kerbin-like planets will provide general science. 

I like this idea, but I feel it should only provide a small bonus as to not restrict player freedom.

[Staticalliam7]

I think honestly we should, 'ya know, learn stuff from the science. Not just points, but actually science.

[shdwlrd]

5 hours ago, Pthigrivi said:

Wow, brutal. But yeah to say that there will be biomes, experiments, and science to unlock the tech tree does confirm there won't be any radical rethinking of the process, so that's interesting at least. It's one of the small bits of info about gameplay that they've let slip, and cool to hear about how the art and geology could be related. I know we've talked about science before but it's been a while and I don't actually recall what my suggestions were? Im also curious if folks have shifted in their own thinking about this? Like I said they've probably made a lot of progress already but its still fun to speculate.

LOL If being honest and annoyed is considered being brutal, then you may not like the rest of my thought process. The line you are quoting is being used as an example. (See the spoiler below to see the full (corrected) transcript from YouTube for that section.) Shana says "That plays actually into our science system as well. in ksp, you can go to various biomes and perform experiments get science and use that to find new parts." In that moment, she confirms that Intercept has made there own science system and is using KSP1 as an example to how the science is tied to the different biomes. I'm taking that as proof that Intercept isn't following Squad's example and is correcting Squads mistakes. Unlike most people who are assuming that Intercept is just going to copy verbatim what Squad did with the career and science modes. (Which has already been said in this thread. ) I do get annoyed that these threads turn into a gripe fest about the errors that Squad did with these systems and by proxy, that Intercept will do the same thing. Any new ideas or possibilities get drowned out by the gripes, bickering, corrections of the assumptions.

There have been many interviews that have Nate saying that they are scrapping Squads career and science systems and are creating their own. In this one aspect of KSP2's development, I actually do trust that Intercept will redo everything (hopefully) for the better. (There are ways Intercept could screw it up, but it won't be the way Squad did.)

You have to remember that Intercept is very good at not saying anything specific about what they are working on. Like in the statement above, they will confirm that they made a change, but nothing else will be said at the moment. But they are more than willing to use/look at KSP1 for examples.

Spoiler

it feels like (from) concept to gameplay (there) is a very high level of iteration between design and art. like in some ways there's… there's some story right. you're like this is a star system (and,) this is the story of this particular star. these are the kinds of planets that would be around this star and so you kind of get, like, here... here's the basics of it. and then art gets in there. and they're like "I’m gonna make this cool as hell", and kind of add all these nice little bits to go find. and... and you're like okay this is neat. and so like we may be responsible early on as design for like here. here are the orbital challenges just trying to intercept (this thing, verses like...) the saying the person's like here are some of the physical characteristics of this planet that may make it a challenge to navigate through. and then art goes in and says like, "here are all the cool places that you can go to." that plays actually, into our science system as well. in ksp, you can go to various biomes and perform experiments get science and use that to find new parts. and we started working with art to make sure that we were tying biomes (in)to, like, these really distinct locations that they've been putting their heart soul into.

(made some corrections to the transcript. added some missing words and punctuations since it's just a few giant run-on sentences. there seems to be a couple bad edits too.)

Now with that out of the way... I do like speculating about how Intercept may do the different systems in KSP. I do like to see other peoples ideas/thoughts on how these system might be executed. But man, people need to stop assuming that Intercept is Squad 2.0. 

6 hours ago, Zhetaan said:

Aside from a general hope that they go with region over biome (which I think is rather a silly descriptor for a portion of a lifeless, airless moon),

You have a point there. The scary thing is that it would make sense for a planet with life too. Who knows, they may make that change.

[Pthigrivi]

1 hour ago, shdwlrd said:

In that moment, she confirms that Intercept has made there own science system and is using KSP1 as an example to how the science is tied to the different biomes. I'm taking that as proof that Intercept isn't following Squad's example and is correcting Squads mistakes.

 I surely hope they do. But when she says “that ties into our science system” “that” refers to “cool places you can go” and then she  goes on to describe the science system as being very much like KSP1s. I happen to think thats a good thing, and said so several times in my opening post. My gripes about KSP1s science system are about execution, not its fundamental structure. They’re also widely shared and I think well founded caution notes. I simply tried to be a little more specific and thoughtful about the nature of the problem than to say “its grindy”, which means different things to different people and deserves some unpacking. So Im not sure what the issue actually is?

[Tw1]

12 hours ago, Zhetaan said:

 hope that they go with region over biome (which I think is rather a silly descriptor for a portion of a lifeless, airless moon)

This. It's like they learnt the word from minecraft. Even back in the day we thought it was silly, but they never changed it.

12 hours ago, Ahres said:

 I'd love a science system where scientific observation is useful for future missions.  ...... - I want exploration and discovery and reconnaissance to be way more involved.

This too.  This would be better than the tech tree approach, especially 

 

I think in my ideal science system, there would be three levels.

1. Passive sensors - scanners, instruments, etc,
Stuff which you just turn on as part of your ship, and return data based on the situation they're in. Novel situations bring you more interesting (or maybe valuable) data.

2. Active experiments - stuff you have to interact with.
Drilling stuff, or chemical tests, stuff you actually have to move around and click to make work.

3. Long term projects. Stuff you set up to run then leave for a while.
Research in stations and bases, or something which involves signals between multiple probes, or that you deploy on the ground for long term data collection.

We kind of got these all in KSP1 by the end, but it would be nice for them to be designed in a more comprehensive way from the start. Something that makes them interact in interesting ways.

I hope KSP2  embraces (cost / time x depreciation) mechanics. There's potential there which was wasted in KSP1, fast forwarding years was too easy a choice.

Edited November 3, 2021 by Tw1

[shdwlrd]

59 minutes ago, Pthigrivi said:

 I surely hope they do. But when she says “that ties into our science system” “that” refers to “cool places you can go” and then she  goes on to describe the science system as being very much like KSP1s.

Nope, nope, nope. She says "plays into" and not "ties into."

Her meaning is that the artists are doing such a good job creating kick ass areas to visit, it's making it easy to place/write elements of the science system within/for them.

Whereas your meaning is that the artists are creating these specific areas and in a specific manner because that is where the placement/written elements of the science system will be at. 

I know we're getting into semantics, but trust me. She means that the artists are making the developers jobs easier placing the science, not the artists are creating these places because that is where the science will be at.

When she transitions to the statement about KSP's science, that is a standard lead in for a quick example/explanation here in the US. I use it all the time when I need a quick example when explaining something.

She is saying that there will be different science in the biomes like in KSP1, not likening the science system to KSP1. I don't think Nate or the PR department would of left that statement in if was the true method of how the science system works without it being ready to release that information.

2 hours ago, Pthigrivi said:

My gripes about KSP1s science system are about execution, not its fundamental structure. They’re also widely shared and I think well founded caution notes.

My gripes about KSP1 science system is that it's completely flawed and not well thought out. How many other games have you played where a lunar study can lead to unlocking better aerodynamics? It doesn't make sense.

In Homeworld, you have specific areas of research. You research a corvette upgrade, you get an upgraded corvette, not a better fighter. You don't get cruisers until you max out the destroyers research tree.

The lack of fore thought killed the science system for me. But I do agree, let the science system in KSP1 be a lesson in what NOT to do if you are going to have specific research tasks.

2 hours ago, Pthigrivi said:

So Im not sure what the issue actually is?

It's two fold, you're claiming that Intercept is going to follow Squads formula for science and tech unlocking when the statement referenced was never intended to be nothing more than an example.

The title is too general. It invoked a "not another aimless thread about science" reaction from me. Every time I've seen one as of late, it's claiming that Intercept is going to follow KSP1 style of science, which needs to be corrected. Or it's griping about how KSP1 science system is broken and Intercept shouldn't follow it.

[Pthigrivi]

@shdwlrd Sorry, man. I think we’re talking past each other. Im saying it seems likely that KSP2’s science system will be similar to KSP1 in that biomes will exist and you will use experiments to gather data from it and unlock parts. Nothing more. That doesn’t mean the experiments will work the same way or that the tech tree will be structured the same way or anything else. I think we both agree those are the areas that could use the most improvement. 

[Pthigrivi]

The truth is my thinking about all this is pretty open at this point. Lets look at some common points:

On 11/2/2021 at 12:32 PM, Tw1 said:

The other problem was how much it was linked to scoring simple points, rather than science for the sake of science. Not against it helping you get somewhere, but I hope it has more emphasis on the building-up-an-understating-of-the-planet side of things this time. I want to see more graphs and charts to complete.

21 hours ago, Ahres said:

I haven't given it much thought while I'm sure some of the other KSP2 science speculation threads have, and I'm not sure @Pthigriviwas wanting to go this direction, but I'd love a science system where scientific observation is useful for future missions. An example would be something as simple as atmospheric height. I hope when I go to my first extrakerbolar planet I don't know a thing about it (except what can be gleaned by telescope), even how high the atmosphere reaches above the planet's surface. I want my science gathering to give me that information - I want exploration and discovery and reconnaissance to be way more involved.

18 hours ago, Staticalliam7 said:

I think honestly we should, 'ya know, learn stuff from the science. Not just points, but actually science.

I think this is would be really great for a few reasons. KSP is currently lacking a lot of useful information players would want like altitude and slope maps, biome maps, etc. and using orbital scanners would be an excellent way to deliver those. It also seems like the prospecting process will be more complex so using experiments to locate resources and determine concentrations could be really seamless. Determining atmospheric height and density and radiation levels could also be important, along with predicting whether you're going to overheat on a given re-entry profile or optimize radiators for local heat dissipation. The important thing is that the information isn't just flavor text, it's actually useful for navigating space and playing the game. I would personally tailor the slate of experiments to answer these specific questions so each experiment felt vital. The question remains though--are these experiments also providing science points that can be applied to research? Like are there places with pure scientific data? Unique geological features or signs of life? Im thinking about how surface samples could be something of a unique case.
 

21 hours ago, Zhetaan said:

Aside from a general hope that they go with region over biome (which I think is rather a silly descriptor for a portion of a lifeless, airless moon), I do like @Tw1's general thinking that science should feel more like a process than a game of fetch.  I can't speak to ease of implementation, but I think that there's something to be said for splitting out science from a general collection of points to something a bit more subject-based. 

9 hours ago, Tw1 said:

I think in my ideal science system, there would be three levels.

1. Passive sensors - scanners, instruments, etc,
Stuff which you just turn on as part of your ship, and return data based on the situation they're in. Novel situations bring you more interesting (or maybe valuable) data.

2. Active experiments - stuff you have to interact with.
Drilling stuff, or chemical tests, stuff you actually have to move around and click to make work.

3. Long term projects. Stuff you set up to run then leave for a while.
Research in stations and bases, or something which involves signals between multiple probes, or that you deploy on the ground for long term data collection.

21 hours ago, t_v said:

I love how versatile the point-buy system is when deciding what to specialize in, but it becomes very hard to balance science with progression, especially with labs. KSP 1 play throughs with a lab unlocked will end as soon as a trip to duna brings back its first haul of data, and even sooner if players revisit the mun or minmus. Ideally in a play through, I would like to have to visit the jool system at least once before finishing science progression. With the expanded KSP 2 tech tree I’m sure that players will have to go to more places to finish the tree, but it could potentially be completed by the end of the first interstellar voyage, depending on scaling. And there is no good way to balance this with a simple point-buy system. 
 

My solution: have different planets unlock different places on the tech tree. 

9 hours ago, shdwlrd said:

The lack of fore thought killed the science system for me. But I do agree, let the science system in KSP1 be a lesson in what NOT to do if you are going to have specific research tasks.

This all goes to some general problems with implementation and what the mechanism should be for getting from raw gathered data to unlocked parts. Im pretty sympathetic to @t_v's view that while versatility is nice, it can lead to other problems like players wringing KSOI for every drop of science before sending their first probe to Duna. I think they're right that you want some way of creating either different categories of science or development gates in the tech tree that only allow certain groups of parts to unlock after you've delved deeper into the Kerbolar system. But like @Zhetaan says, this could get overcomplicated quickly and detract from core gameplay. So it's a balance. I definitely agree with @Tw1 that some experiments should be passive and others should be active and durational. So the question is what are those and how do they answer question 1 above about useful data?


 

Edited November 3, 2021 by Pthigrivi

[SciMan]

Personally my problem was that I can't think small. I can't design a "minimalist" mission to ANYWHERE, because that means if it's a lander I'll have to send it far too many times to cover all the biomes of a given planet or moon. Additionally, because I can't design minimalist missions, I need "most" of the parts to be able to be accessible to me, in order to get the simplest things done. I'm not saying that's a failing of the game, no. What I'm saying is that most of the time the entire reason I need a given tech tree node is because I need maybe 10% of the total parts in it, with the exception of the fuel tank and rocket engine tech tree nodes, which I do tend to use most of.

My solution is that you should be able to research new parts on a PER PART basis.
Additionally, to FORCE you to explore beyond the gravitational influence of Kerbin, there should NOT be enough science points available to be gained on Kerbin, the Mun, and Minmus combined to unlock the entire tech tree (even with science lab use, which also needs to be re-worked). To counter-balance this, you would be rewarded MORE than currently for exploring planets beyond Kerbin's SOI.

Of course, what this ends up doing is splitting the tech tree into many tech trees, each centered on a specific type of part.
For example, you'd have something like the following, with each individual part being a node in the relevant tech tree (aside from maybe the first 2-3 tech tree nodes (not tiers of nodes, literally 2-3 nodes and that's it) containing multiple parts before it branches out into the following tech trees):

• Propellant containment (fuel tanks of all shapes, sizes, and maybe propellant types if that's not readily switchable in the editor or maybe even in-flight (empty tanks can be switched to another kind of fuel within limits, say you'd be able to switch a fuel tank from holding Methalox to holding Hydrogen, or Monopropellant, or that kind of thing, but not Orion drive charges or Fusion fuels))
• Cargo Containers (for holding things used in either orbital shipyards or building colonies on other worlds)
• Adapters (When you want your rocket to change cross-section, use a part from this category, offerings would include both Fueled and Hollow adapters, and for certain things you'd even be able to select a Skeletal type for even more mass savings. Multi-couplers would also be here, but engines should have a surface-attachment ability for custom clustering anyways)
• Propulsion (Get your engines here! Perhaps one tech tree per which fuel combination those engines are using, as well as another tech tree for the air-breathing or atmosphere-using engines, or they could be branched off of each other in one large tree I guess but picking where they each branch off would be something for a proper game designer to do)
• Aerodynamics (Separate tech trees for "airplane parts", "control surfaces", "payload fairings", and "parachutes" would probably be a good idea because not everyone wants to build a spaceplane or a rocket or an airplane)
• Structural parts (Girders, Light-weight Truss segments suitable for the core of an interplanetary or interstellar Mothership, Tubes, etc.)
• Electrical supply (tech trees for Solar panels, batteries, RTGs, and other methods of generating power for your vessel would go here)
• Attitude Control (RCS and reaction wheels and maybe autopilot parts if they're included and not just part of the command pods would go here)
• Command (Pods and probe cores of all shapes and sizes, and you'd start off with both a pod and a probe core (a USEFUL probe core, so ALL probe cores and pods would have at least the basic "SAS" mode available (with the pods it would no longer matter if you have a pilot on-board or not, you'd have at least "basic SAS" so you don't have the "disappointing spinning out of control rocket" that happens with some probe cores and some crew options on crewed pods, because I'm trying to minimize the amount of times people revert to VAB because they forgot something critical).
• Docking and Staging (Decouplers and stack separators and docking ports and even the Claw would go here, also maybe you'd be able to unlock some kinds of parachute here along with in their dedicated tree under Aerodynamics)
• Robotics (should be self-explanatory, all the "servo" and "piston" parts would be here, and so would the "motor" parts but you would be able to unlock the "motor" type parts either here or via research into rover wheels).
• Ground Mobility (Landing legs, aircraft landing gear, rover wheels (and maybe even crawler treads!), grip pads, and yes those "motor" parts I mentioned in the "Robotics" section)
• Thermal Control (Heat shields, radiators (please fix the stock heat bug!), and perhaps even some "open-cycle cooling" parts like a heat shield that uses Methane or Hydrogen or Oxidizer instead of
  "Ablator" to resist the heating of reentry)
• ISRU (Mining drills and converters of all shapes and sizes, including the ones built by a colony. Also change it so that a totally unmanned mining outpost is something that is not just possible, but smart (ie re-work the Crewed/Engineer bonus for mining drills and converters))
• Habitation (parts that can hold crew but can't command a vessel, so passenger cabins of all kinds, hopefully including some Centrifuge and Inflatable parts for building spaceships (and colonies in orbit or on the surface) with)
• Colonization (Find all the unlocks for colony parts (or upgraded colony parts) right here, if you haven't already unlocked them somewhere else, but this too would likely have its own set of branches for each type of thing a colony can do)
• Communication (Antennas of all kinds found right here, a probe's no good if it can't phone home! We'll obviously need new antenna parts to handle the longer communication distances between the different star systems).
• Utility parts (Lights, Ladders, flags, that kind of thing)

I tried to be as complete as possible, but the idea is simple: You shouldn't have to unlock a Cubic Octagonal Strut or a Monopropellant fuel tank part if you want the LV-1 Ant engine.

Hopefully splitting the tech tree up in this way along with reducing the science gain from the Kerbin SOI planet and moons (and re-working the way the science lab works) would do a lot to improve how Science is handled on the "Spending It" side.

 

As far as the "Earning Science points" side of things, I have an idea or two there too.

Science experiments should be something the player is capable of turning on or off, but they should NOT provide instant science points like they do currently. You should have to have it sampling whatever it monitors for a while before it returns some amount of science points. Running the experiment should require both some (small) amount of power, as well as either an Experiment Storage part on the craft, or a communications link TO a craft that has an experiment storage part (to act as a "store and forward" type relay satellite), or ideally a communications link directly back to Kerbin (potentially thru more conventional "direct radio relay" satellites).

The science lab would no longer operate with samples collected directly from the environment. Instead, a science lab would merely be another kind of experiment, that uses data from the samples you've sent back to Kerbin either as data or as a physical sample (there would not be a difference in science gain between those two methods) to create a sample of its own to send up to the science lab.
The more different the conditions are to "landed on Kerbin", the more science would be gained. This already makes it so that science labs landed anywhere on Kerbin will not result in any gain for the player.

Orbital and landed experiments would take weeks to months to complete, whereas experiments that were conducted while the science lab is flying in an atmosphere would take much much less time (say around an hour of flight time). Either way, you'd need to then transmit the results back to Kerbin to get the actual science points out of it.
Furthermore, as you send more non-lab science experiment data back to Kerbin, you'd automatically unlock more advanced experiments for the science lab that would have greater science point returns.
Hopefully this would both keep the science lab useful throughout the progression of the game, as well as not make it "too good for a tiny expenditure of effort" like it is currently.
Also, turning the science lab back into a different kind of experiment makes it no longer worthwhile for a vessel to have more than one science lab unless it's running more than one kind of experiment at a time.

[shdwlrd]

12 hours ago, Pthigrivi said:

@shdwlrd Sorry, man. I think we’re talking past each other. Im saying it seems likely that KSP2’s science system will be similar to KSP1 in that biomes will exist and you will use experiments to gather data from it and unlock parts. Nothing more. That doesn’t mean the experiments will work the same way or that the tech tree will be structured the same way or anything else. I think we both agree those are the areas that could use the most improvement. 

Yeah, I'm thinking that might be the case. I was a little feisty yesterday.

[Pthigrivi]

47 minutes ago, shdwlrd said:

Yeah, I'm thinking that might be the case. I was a little feisty yesterday.

Haha its alright. I don’t disagree these conversations can get a little rehashy. Trying to observe the no-necrobump ethos and let new minds here chew it over. 

[shdwlrd]

10 hours ago, Pthigrivi said:

The truth is my thinking about all this is pretty open at this point. Lets look at some common points:

I think this is would be really great for a few reasons. KSP is currently lacking a lot of useful information players would want like altitude and slope maps, biome maps, etc. and using orbital scanners would be an excellent way to deliver those. It also seems like the prospecting process will be more complex so using experiments to locate resources and determine concentrations could be really seamless. Determining atmospheric height and density and radiation levels could also be important, along with predicting whether you're going to overheat on a given re-entry profile or optimize radiators for local heat dissipation. The important thing is that the information isn't just flavor text, it's actually useful for navigating space and playing the game. I would personally tailor the slate of experiments to answer these specific questions so each experiment felt vital. The question remains though--are these experiments also providing science points that can be applied to research? Like are there places with pure scientific data? Unique geological features or signs of life? Im thinking about how surface samples could be something of a unique case.
 

This all goes to some general problems with implementation and what the mechanism should be for getting from raw gathered data to unlocked parts. Im pretty sympathetic to @t_v's view that while versatility is nice, it can lead to other problems like players wringing KSOI for every drop of science before sending their first probe to Duna. I think they're right that you want some way of creating either different categories of science or development gates in the tech tree that only allow certain groups of parts to unlock after you've delved deeper into the Kerbolar system. But like @Zhetaan says, this could get overcomplicated quickly and detract from core gameplay. So it's a balance. I definitely agree with @Tw1 that some experiments should be passive and others should be active and durational. So the question is what are those and how do they answer question 1 above about useful data?


 

Maybe science can be broken up into a few different categories. Lets say these categories are observational, physical/practical, and theoretical sciences. Observational science is like using ScanSat to start filling in the blanks with new planets. Physical/practical science you have to go to some place and get a sample, or use a new experiential part for some time to help unlock for future use. (The Kerbals don't seem like the type to test a new piece of equipment for decades before actually using it.) Theoretical science is used to gatekeep advanced technologies and to help fully unlock experimental parts for common use. (Much like how aircraft manufacturers are continuously looking at and testing new designs to increase efficiency.) Both observational and physical science requires some level of involvement of the player. Theoretical science can only be influenced by the number of labs and finding the brightest staff for them. 

This is the best I have that can appease the different gripes with science. There's no apparent path to exploit the system and jump ahead immediately. (Except for time warp. I deem time warp an acceptable exploit since player interaction is required for unlocking new parts for common use.) You can use science to actually learn something and to help with planning. It shouldn't feel like a complete waste of time but feel like an accomplishment, or at least feel like it's useful and not a grind.

[KerikBalm]

On 11/2/2021 at 8:37 PM, Ahres said:

In KSP1 we can actually turn on the display of a thermometer and see what temperature the vessel is at, but who bothers to look? It doesn't do anything, it's not recorded anywhere so we can compare against other biomes (or regions, @Zhetaan), or other planets. 

I sometimes look, out of curiosity.

The most useful instrument is the barometer... but I *think* that temperature and pressure combine to affect atmospheric density - Pressure matters for rocket Isp, but density is what matters for drag and lift - which means a molecular weight readout would also be useful (although that doesn't change within an atmosphere, but does vary between bodies) .

I would have this https://wiki.kerbalspaceprogram.com/wiki/Atmospheric_Fluid_Spectro-Variometer display atmospheric MW and density... which are also useful to know.

One other thing: make all science experiments available from the start. Using science experiments to unlock science experiments is grindy. It leads to repeating the same missions over again, just carrying a different instrument to get the science that one missed previously.

 

Edited November 4, 2021 by KerikBalm

[shdwlrd]

20 hours ago, KerikBalm said:

One other thing: make all science experiments available from the start.

Agreed. If not, at least have the basic versions available and allow upgrades to more precise versions.

[Pthigrivi]

I really like this idea that the experiments are not just providing science points but also vital gameplay information. Im not so sure we need ALL experiments to be available from the start, just for them to appear shortly before we need them. You'll notice I've dispensed with a lot of existing reports that gobble up player time and don't produce valuable information, while adding some for advanced prospecting, altimetry, and aerobrake prediction. Some of these could also unlock predictive modeling on new bodies  in the simulator mode they've shown previously. The main thing here is to disincentivize endless biome-hopping and repeat missions, and to push players to reach beyond KSOI earlier. 

Some key ideas to start off: 

1) Make most experiments passive (on by default with low ec draw) always pay out 100%, and be applicable in fewer situations. For instance the thermometer provides a steady service (revealing heat bars) and automatically produces data once per biome. No manual clicking. Others would pay out as a percentage of a total scan like the barometer and orbital scanners. Orbital scanners should also work like SCANsat by passing over the surface ideally in polar orbits at optimum altitudes.

2) Take experiments out of the tech tree and tie them to Boom event rewards, possibly along with building upgrades. In the example below we'd say Tier 1 is available from the start, Tier 2 unlocks on your first Munar/Minmar landing, Tier 3 on your first interplanetary landing, etc. along with lots of other parts held behind Boom event gates. That way there would be no reason to tediously farm science around KSOI, and you'd be highly encouraged to get out and land on Duna or Moho as early as you could. 

3) Let science deplete across an entire body based on collection. Once you have complete orbital scans and 3-4 surface samples it should be effectively mined out. The way to game this would be to make some biomes more scientifically valuable than others, so you'd want to land in 3 or 4 diverse locations to get your max value but not feel inclined to biome hop to every one. This could also be increased by discovering anomalies with their own inherent science value.

4) Clean up the UI/UX as it relates to data vs science and the science lab mechanic. Most experiments should transmit automatically, but they should still take time to analyze at the R+D facility on Kerbin. The best way to increase this research rate would be to employ off-world science labs that could convert data into usable Science that could be spent on parts. As you expand and build more colonies with more and bigger science labs this processing rate could be further increased.

5) Make anomalies into their own little Boom events with special rewards and tailor them towards interesting geological formations and the search for life. There could be lots of associated perks like better resource scanning, increased ISRU extraction and processing, better radiation shielding or LS efficiency, etc.

Here's a stab at a slate of experiments to make all that happen: 

Tier 1: First few launches to first Munar/Minmar landing and return

Thermometer (passive): Determines temperature in each biome and enables heat-bars on parts.
Barometer (passive): After passing through the air column shows atmospheric height and density. Necessary for aerobrake and reentry heat prediction. 
Surface Sample (instant): Performed by Kerbals on the surface. Shows local basic resource concentrations once analyzed. 

Tier 2: First Munar/Minmar landings to first interplanetary landing

Infrared Imager (passive): Determines surface temperature and geothermal locations from orbit (SCANsat style)
Terrain Scanner (passive): Produces altimetry, slope, + debris maps from orbit (SCANsat style)
Biome Scanner (passive): Produces Biome maps from orbit (SCANsat style)
Basic Resource Scanner (passive): Produces resource maps for basic resources (not uranium or He3) (SCANsat style)
Dosimeter (passive): Determines surface radiation and shows radiation exposure on vessels
Core Sample (instant): Performed by Kerbals. Shows Uranium and He3 concentrations once analyzed
Basic Science Lab (durational): Analyzes surface samples and processes raw data into science 

Tier 3: First interplanetary landings to first interstellar mission

Radiation Scanner (passive): Produces radiation maps and can locate uranium and He3 deposits (SCANsat style)
Anomaly Detector (passive): Locates surface anomalies from orbit (SCANsat style)
Atmospheric Scanner (passive): Maps atmosphere and local wind-speeds from orbit (SCANsat style)
Atmospheric Analyzer (instant): Can collect and analyze atmospheric samples for Xenon and O2 concentrations.
Robot Arm (instant): Can collect and analyze core and surface samples.
Deep Resource Scanner (durational): Can precisely map subsurface resource deposits within a local area
Colony Science Lab (durational): Analyzes surfaces samples, increases LS efficiency, + processes lots of data into science

Tier 4: Interstellar and beyond:

Multi-spectrum scanner (passive): Can map surface temp, altimetry, wind, biomes + anomalies from orbit.
Multimodal spectrometer (passive): Can precisely map all resource concentrations from orbit.
Interstellar Science Lab (durational): Analyzes surface samples, big increases to LS efficiency and data processing


 

Edited November 7, 2021 by Pthigrivi

[KerikBalm]

20 hours ago, Pthigrivi said:

Core Sample (instant): Performed by Kerbals. Shows Uranium and He3 concentrations 

This is sort of an aside, but I really hope that He3 will need to be mined from the upper atmosphere of gas giants.

I know people often talk about the moon as an He3 source because of higher concentration there than on Earth, but 100x higher than basically nothing is still close to nothing.

One may be able to get enough to power a reactor to run a vasmir/electric thruster using xenon or argon propellent, or H2 as a thermal rocket, but not to fill up tanks for a pure fusion rocket where the exhaust is just fusion products.

In the former case, you're probably better off using a fission powerplant anyway, so what is the point?

[Bej Kerman]

1 minute ago, KerikBalm said:

but 100x higher than basically nothing is still close to nothing.

It's still something and will probably be quite a bit in KSP.