how would you like research to work in ksp2?

[Guest]

53 minutes ago, boolybooly said:

Would like research to be far more granular, individual parts for some branches of the tech tree like SRBs and a separate branch which is similarly one at a time research progression for very large engines like Mammoth and Mainsail. Then test firings etc could add gimble cabability to prototype engines and improve it for production models as well as improve ISP and tweak atmo/vac  ISP performance to your requirement, but consume resource of some kind, like research and funds with exponential scaling of cost per improvement. Many types of parts could be improved, like landing legs could have better load bearing / heat tolerance / lower weight but if you raise heat tolerance or load the weight goes up. 

I've always been against this because it leads to a fragmentation of what a part mean when sharing a craft, If you develop your Mainsail different than mine my craft could work differently in your savefile, but now I think there could be a pretty simple solution to this, instead of thinking about improving on the stock performance what about building up to them?

You start with a way worse engine and then invest in it to make it better and cheaper (mass production), not some mayor and long grind but something giving unlocking new engines a special feeling, and that unlocking is not a binary thing but you get the basic version of the engine immediately and then build up to its maturity.

This could also work with multiple way of improving an engine for example if the engine has 5 steps you can either pay for a step, do a research program (with science points and time or whatever equivalent), do some test flights with fixed requirements or just do incremental updates through production (every 10 engines used you get a step upgrade).

The last step would be the "stock" performance with maybe only 1 or 2 steps further that only improve cost, time or resources required to buy/craft at a colony

[pandaman]

2 hours ago, Brikoleur said:

It's hard not to timewarp when you're doing interplanetary missions. That will fill up your labs PDQ.

Definitely.  But that is what I meant by 'doing something else',  as opposed timewarping  just to skip the waiting so long for the Lab.  

[Guest]

56 minutes ago, pandaman said:

Definitely.  But that is what I meant by 'doing something else',  as opposed timewarping  just to skip the waiting so long for the Lab.  

The upshot is the same, you'll max out your tech tree by the time your interplanetary mission arrives. So if your primary motivation is progress – and for most of us, it is, Pavlov's insights apply to us more than we'd like – you might as well not even bother sending that mission but rather timewarp the years away. Labs are just wildly unbalanced.

This is a more general problem: the mechanic ties progress in the game to passage of game time, while at the same time there's no cost to passing time and it's really easy, even required to do it. Consequently the MPL gives you a free lunch, and free lunches are not much fun in a game.

(This specific problem could be mitigated by slapping a fixed monthly Funds/Reputation cost on your space program, which would limit your timewarping. I don't think it would be a good solution, however, because it would make a mess of other things -- specifically, it would force you to do busywork while your Eeloo mission is in transit, and I would hate that. Which brings me back to what I said a few posts up – in my opinion the entire Sci mechanic is kind of DOA and needs to be rethought from scratch; the MPL is just the cherry on the cake.)

[swjr-swis]

13 hours ago, jastrone said:

and the moon is dusty

You're either missing the point, repeatedly, or just trolling me. So I'm gonna let that munar dust bury this exchange.

[pandaman]

6 hours ago, Brikoleur said:

Which brings me back to what I said a few posts up – in my opinion the entire Sci mechanic is kind of DOA and needs to be rethought from scratch; the MPL is just the cherry on the cake.

Agreed.  That's why i only pjay sandbox now.  I like the idea of career, and would want to play it, but it just doesn't inspire me to continue playing it after the first few missions.  Hopefully the KSP 2  equivalent will be to my liking.

[Pthigrivi]

I love this topic. 

Let's start with science collection. I think thematically both having physical experiments and using labs is really important, its just that the execution in KSP 1 is pretty finicky and a little lackluster. Breaking Ground introduced some really cool things that helped a lot (surface features, scanning arms, surface experiments, etc.) but it couldn't undo some of the fundamental structural issues. Leaving aside science points for the moment there are two main issues for me.

1) Collecting science should factor into mission planning but shouldn't interfere with the player's focus while flying. In fact the act of collection itself should take up as little player time as possible. Even if all in-flight collection happened with one button or was passive and automatic we shouldn't be clicking through a dozen report windows. 

2) Experiments should provide information valuable to the player besides science points. This brings me to one of KSP1's key deficiencies: mapping. Each body should have a series of overlays that can be viewed seamlessly both in-flight and in map-mode, especially: Altimetry, Slope, Resources, and Scientific value. There might be other overlays showing things like atmospheric Xenon and Radiation. The available experiments should really be designed around gathering this information and should be the main reason to send probes ahead of crewed and colonization missions. There's other non-map information that could be unlocked by conducting experiments like trajectory prediction factoring drag, reentry heating prediction (will I burn up?), and landing site prediction that could be key to aerobraking and precision landing. It could also be key to unlockable autopilot functions if that becomes a thing.

This might leave us with the following experiments:
- Altimetry Scanner
- Resource Scanner
- Anomaly Scanner (Science value)
- Electromagnetometer (Radiation levels)
- Atmospheric Analyzer 
- Surface Sample Collector

There might be more or different levels of each especially if they were passive, but I'd be inclined to keep it relatively focused and simple. My personal feeling is the orbital scanners should work similarly to SCANsat in that they produce the maps over time as they pass over the surface, and all of the experiments except surface sample collection should happen more or less passively with no micromanagement. I might also be convinced we could drop the concept of transmission entirely, that upload would be automatic if there was a radio connection with Kerbin or a local Science lab. 

This post is already long so maybe I'll noodle on science and tech development in part 2.

Edited May 12, 2021 by Pthigrivi

[Kerbart]

On 5/10/2021 at 8:38 AM, jastrone said:

its pretty safe to assume there will be a career or atleast a science mode in ksp so how would you like that to work?

For starters I wouldn't assume that there are "science points." That solution has been there in KSP 1 "for reasons" but I don't think many of us are happy with it It's grindy, you can do any experiment anywhere and use the outcome for anything. I realize that realism for realism's sake isn't making the game necessarily better but doing 18 EVA observations around Kerbin leading to having an atomic engine is asking a lot of suspension of disbelief as well.

I'd rather see a much more integrated game mechanic where specific (classes of) experiments lead to specific "discoveries." A few examples of what I mean:

• Atmospheric research (temperature, pressure) lead to aerodynamics
• Ground samples (durface sampling, some core-drilling devices) lead to better landing gear
• Fundamental research (magnetics, spectroscopy) leads to better engines
• Material experiments lead to better parts
• And so on

From a game-playing perspective I wouldn't mind that you even don't get to choose what exactly gets unlocked; technological progress doesn't work like that in reality either. It would make no two games the same. And to make bigger discoveries you'd need bigger (more complex) experiments. Maybe one team setting up a reflector and another team shooting a laser at it.

You can throw in random artifacts that lead to surprise discoveries, and maybe even "reward" crashes (as opposed to revert to launch or reloading quick saves) as the research of an accident might reveal something from time to time.

I like the idea that more labs in places can speed up a discovery (unlocking something in the tech tree). Perhaps something for artifact research? That could also tie in with unlocking parts further down the tech tree that are not yet accessible, but when an earlier node gets unlocked "suddenly the weird materials found in the Duna expedition make more sense now. You now have access to..."

Doing it this way would make science not merely a game-token collecting exercise, but it would be a story-telling device and an incentive to go exploring.

 

[boolybooly]

19 hours ago, Master39 said:

I've always been against this because it leads to a fragmentation of what a part mean when sharing a craft, If you develop your Mainsail different than mine my craft could work differently in your savefile, but now I think there could be a pretty simple solution to this, instead of thinking about improving on the stock performance what about building up to them?

You start with a way worse engine and then invest in it to make it better and cheaper (mass production), not some mayor and long grind but something giving unlocking new engines a special feeling, and that unlocking is not a binary thing but you get the basic version of the engine immediately and then build up to its maturity.

This could also work with multiple way of improving an engine for example if the engine has 5 steps you can either pay for a step, do a research program (with science points and time or whatever equivalent), do some test flights with fixed requirements or just do incremental updates through production (every 10 engines used you get a step upgrade).

The last step would be the "stock" performance with maybe only 1 or 2 steps further that only improve cost, time or resources required to buy/craft at a colony

I dont see this as a problem at all, you could just import the engine data with the craft file to use in sandpit mode and filter engine data in career mode and allow substitution of an imported engine model/specs with a player owned engine model, maybe add a warning or comparison like with prototype part testing mission designs in KSP1, where you can build a craft with a part you have not unlocked if you have a mission to test it and retain the craft design if you completed the mission but cannot build it until you unlock the part. The programming for that has probably already been done.

Then with an imported design in career you can try to use your own engines or develope engines to meet requirements and match the imported craft spec. If anything it becomes an incentive/challenge to R&D better engines.

Any designs included in the game would have to work at min spec for engine performance obviously.

Personally I would like it if the product of engine R&D started below stock capability i.e. prototypes and then matched it to become a production model and went on to exceed it to a limited extent with trade offs allowing excellence in one capability. There might need to be different models of the same engine type for that reason, player named and player developed. Obviously you would not want it to get too silly but on the other hand it might be something to do with end game resources.

[Guest]

22 minutes ago, boolybooly said:

Personally I would like it if the product of engine R&D started below stock capability i.e. prototypes and then matched it to become a production model and went on to exceed it to a limited extent with trade offs allowing excellence in one capability. There might need to be different models of the same engine type for that reason, player named and player developed. Obviously you would not want it to get too silly but on the other hand it might be something to do with end game resources.

Starting below stock performance is exactly what I was thinking about, with multiple paths you can then follow to improve the engine up to stock (to give the player more choice and avoid repeated grind). 

I'm not at all convinced about personalization, it would have all the problems of a fully procedural engine system without any of the advantages.

If you really want branching and choices leading to different types of engines you could work on families or series instead of singular engines, as an example you could start the game with the LV-T30 downgraded to terrible performances, you upgrade it with a few flights/money/science/tests, then it branches to the LV-T45 and maybe you could redesign a bit the LV909 to make it into a vacuum optimized variant.

With every variant then having their own upgrade path up to stock characteristics (and maybe the LV909 having a branch for the compact version that's the current design).

That way you make the tech tree feel more like  actual technology progression and you could use it to de-clutter the part list, turning engines into variants of other engine.

It could also work if they implement different fuels, I can totally se the LV series being the Kerolox one, the KS-25 split into 3 engines and turned into the 1.25 hydrolox family and so on.

 

I think that is as complex as you can go, more than that and it becomes way to much for a game in which at some point you will switch to completely different engine technologies.

[XLjedi]

For me, it would be nice if the research were a bit less generic.   

I would propose the generic "Science" be replaced with tokens earned in maybe 4-6 different disciplines.  Then have a branching tree where the items on the tree cost a combination of the various tokens.  Maybe no item cost more than 2 or 3 of any discipline token?

[Pthigrivi]

Question for all: How do you feel about relays for probe control and data transmission?

[shdwlrd]

24 minutes ago, Pthigrivi said:

Question for all: How do you feel about relays for probe control and data transmission?

They are necessary for a real life feeling to the game. The only thing I don't think as necessary is the delay associated with the transmissions.

[Dientus]

25 minutes ago, Pthigrivi said:

Question for all: How do you feel about relays for probe control and data transmission?

Relays as in the antenna network designed to bounce and boost various signals across the cosmos?  I like them ingame, adds a level of realism and a reason to built and launch satellites in particular advantageous orbits, if thats what you're asking.

 

[Pthigrivi]

@Dientus Yeah, basically. Do people find it fun? I do but it is work. My main issue is with the UI. I've learned from experience what the ranges mean but it would be nice if there was more in-game visual feedback showing what your probes range was before you sent it all the way out to Jool without realizing it didn't have a big enough dish. 

[Dientus]

2 hours ago, Pthigrivi said:

Do people find it fun?

I myself enjoy it as well   I assume by visual feedback you would mean something like some transparent circle surrounding the dish that shows you it's farthest range. That would be a quick way to go about it, but unless was very advanced wouldn't take into account the planetary bodies blocking it and such.

I just do the math and go from there but having a visual representation wouldn't hurt for those who want it. If it were on all the time and not in a separate map or not being able to be turned off it would definitely make my map cluttered.

[Guest]

3 hours ago, Pthigrivi said:

Question for all: How do you feel about relays for probe control and data transmission?

In my opinion CommNet is one of the best subsystems in the game. It's simple, easy to understand, and creates a ton of varied, emergent gameplay. It's a shame it's optional, lots of players are missing out unnecessarily because they're intimidated by it and switch it off.

[OHara]

4 hours ago, Pthigrivi said:

How do you feel about relays for [. . . ] data transmission?

The idea that you need a line of sight, or chain of such lines, to transmit research seems to fit well with KSP. 

I never use KSP1's CommNet, though, because some details are too complicated for me to remember: only certain antennas are 'relay antennas', the rules about partial control and pilots on craft with a relay, the range of an array of antennae on a craft, . . .

It is disappointing that CommNet did not come with a method to store the research at the KSC, like the maps we gradually build with the ScanSat mod.  In fact, the data-flow direction with KerbNet seems backwards: we need a live connection to KSC in order to get a live display of the planet we are scanning (!) as if KerbNet was a satellite-navigation service drawing on database of maps already compiled at the KSC.

On the main topic of the thread, I would similarly be happy with a simpler set of rules than KSP1's four measures of science / technology / funds / reputation. 
In KSP1 I am happy to cheat science to complete, set the option to have entry costs for each node and part in the tech-tree, fund my career mostly with rescue contracts,  and record my exploration of the Kerbol system in ScanSat maps.

[Jack Mcslay]

 

On 5/10/2021 at 11:59 PM, Brikoleur said:

That could become really tedious upon replays.

You could probably disable it much like the custom difficulty options  in KSP1

[Guest]

14 minutes ago, Jack Mcslay said:

You could probably disable it much like the custom difficulty options  in KSP1

I'd rather they put the effort into game mechanics that stay engaging enough that you wouldn't want to.

[Pthigrivi]

@Brikoleur + @Dientus Yeah I love it too. I wasn't sure how widely shared that was. And yes I definitely imagine it as a toggleable overlay similar to the current sight-lines. I also agree with @OHara that it could be simplified by having all dishes relay-capable with a simple weight/range tradeoff. The main thing for me is that we get the good parts (planning and strategizing around keeping radio contact) without getting frustrated by unexpected behavior or bogged down with pop-up windows. It should all be seamless and in the background. If that were the case I think it could be fun if colonies and crewed vessels served as local hubs for control and research. 

Let's imagine for the moment that the basic principle that experiments and labs are used to upgrade tech remains. There are a couple of changes I'd love to see.

1) Don't create incentives for repeat missions. No more diminishing returns or endless biome hopping. Especially if we're trying to get players to rapidly press out into the solar system and beyond we don't want to encourage multiple stops at the same place or tedious farming around Kerbin. This means science around Kerbin and its moons should be rapidly mined out for scientific value. For this reason I'd suggest replacing biomes with an overlay showing a varied range of scientific interest. Some large features like crater basins might be worth more, and even smaller areas around anomalies might be worth much more to encourage precise landing and surface scouting via rover. I think as you collect surface samples though the values should deplete rapidly for the entire body until only a few unique anomalies are still worth visiting. There shouldn't be much reason to visit a body more than 2-3 times, and it should almost always be more enticing to travel somewhere you haven't been than to farm around where you already have. Speeding up the pacing here will leave more time for exciting things like colonies and interstellar travel.

2) Keep it focused. I think if we were to abide by the principle that many experiments produce useful information then only a few need to be for gathering Science. Again this is about disincentivizing tedious farming. One solution would be to say only surface and atmospheric samples produce Science. Maybe there's a couple other smaller payout methods but not half a dozen parts to click through and upload and farm all over. You can visit and collect samples with crew or you can send probes to collect samples, but they're only worth points if you deliver them back to Kerbin or to a local science lab. This incidentally would become the reason to set up local research stations, so that you wouldn't have to send everything back to Kerbin or wait for return trips. Whether these samples take time to process or not the value should be finite so you aren't time-warping to infinite science. 

3) Clean up the tech tree. It's honestly a mess right now. There may be radically different ways of unlocking parts that others have mentioned, or it could be as simple a single science resource so long as the various nodes and costs were much better organized and balanced. I'd personally like to see it streamlined and based around technology tracks. For instance put the Skipper and all of the 2.5m tanks on the first node in a chain, with smaller downstream nodes for the Mainsail and Poodle. Organize xenon engines into Ion 1, Ion 2, Ion 3, etc. Gather up most of the simple structural parts and things like lights and ladders in really early nodes rather than distributing them randomly down the tree. There might also be non-part technologies that we'd need to develop like autopilot, communication ranges, repair capabilities and science collection bonuses that apply across your population rather than being tracked to individual kerbals' experience. Although maybe this latter category should be unlocked by hitting bigger and bigger off-world population thresholds rather than spending science.

Edited May 13, 2021 by Pthigrivi

[mcwaffles2003]

I've gone over this topic a lot in other threads so I will try to keep it short and highlight what I envision as a better system for how we interact with science.

I hope:

• Science gets catalogued and becomes useful as practical information detailing the characteristics of planetary bodies which helps inform the player to make intelligent decisions with how they will interact with that body (kind of like how we actually use science...)
• Probes will be useful in finding the hazards that we want to avoid  when sending out our kerbals (hopefully some landings will be amazingly hazardous and difficult)
• Telescopes will be useful in cataloging and mapping possible bodies to interact with
• Labs stop being science multipliers and instead just be necessary in completing some forms of research
• Other forms of infrastructure will be necessary to complete experiments incentivizing colonization
• There will be a greater emphasis on sample retrieval instead of simply relaying data back
• Science can imitate SCANSat and be required to find the best locations for getting necessary off world resources
• Science can be gathered passively over time similarly to the manner in which kerbalism approaches science as opposed to clicking a button at a single crucial moment (creating the challenge of holding a condition instead of reaching it for a quarter second and spam clicking)
• If science points still exist that they can not be abused to fill out the tech tree so quickly

[mcwaffles2003]

16 hours ago, Pthigrivi said:

Question for all: How do you feel about relays for probe control and data transmission?

I love setting up relays, as you mention, it is work... but it's fun work that's involving and gives room for creative thinking.

Also, I know it's not everybodys cup of tea, but I would like to see transmission delay so long as there is some form of mission planner that includes conditional statements and automated responses. To go further having conditional statements be dependent on equipment that is on the ship (i.e. the player can't use temperature conditions if they don't include a thermometer on the craft). This would incentivize using kerbals for more precise/in the moment flying while also making probes more effective while also making them challenging.

Edited May 13, 2021 by mcwaffles2003

[Guest]

10 hours ago, Pthigrivi said:

I wasn't sure how widely shared that was.

First things first, I think commnet is one of the best pieces of game design in KSP, now I want a whole progression mode designed as good as that.

 

On the topic of science I think I may have a problem with this:

10 hours ago, Pthigrivi said:

Let's imagine for the moment that the basic principle that experiments and labs are used to upgrade tech remains. There are a couple of changes I'd love to see.

1) Don't create incentives for repeat missions. No more diminishing returns or endless biome hopping. Especially if we're trying to get players to rapidly press out into the solar system and beyond we don't want to encourage multiple stops at the same place or tedious farming around Kerbin. This means science around Kerbin and its moons should be rapidly mined out for scientific value. For this reason I'd suggest replacing biomes with an overlay showing a varied range of scientific interest. Some large features like crater basins might be worth more, and even smaller areas around anomalies might be worth much more to encourage precise landing and surface scouting via rover. I think as you collect surface samples though the values should deplete rapidly for the entire body until only a few unique anomalies are still worth visiting. There shouldn't be much reason to visit a body more than 2-3 times, and it should almost always be more enticing to travel somewhere you haven't been than to farm around where you already have. Speeding up the pacing here will leave more time for exciting things like colonies and interstellar travel.

Specifically this bit:

10 hours ago, Pthigrivi said:

There shouldn't be much reason to visit a body more than 2-3 times

Actually 2 problems, one is a small point on the new ground tech, have you seen how complex Pol terrain has become? That's basically more complexity than the whole KSP1 system together and I want to have an excuse to explore it.

The other problem I have is that I think that 2-3 missions is actually a good target if by that you mean "per mission-type", after 2-3 missions orbital probes no longer bring new science, after 2-3 missions the same happens for probe landers and then the same for "Apollo Style" touch and go missions. The "early exploration phase" can last 5 to 12 mission total and then you move to another body.

But that's just the first phase, while you're doing the same for Minmus you start sending mapping satellites, prospecting probes and maybe and orbital scientific outpost on Mun, and then that orbital station act as a starting point for "the Martian" style expeditions, with crews building temporary habs on the surface and doing longer term exploration and experiments and maybe using early IRSU technology. Again this lasts for 2-3 missions and after that the mining and permanent base building gameplay kicks in, leading to colonies (I think there should be a difference between a temporary hab, a mining/scientific outpost and a colony).

Obviously you won't have to advance that much and continue to the "second phase" for every dusty rock you find along the way and, at some point, you may want to skip some steps and bring all the components of the first phases in one go on a bigger mothership.

 

The problem with KSP1 is not the amount of mission to the same destination, is that the game is built for that "apollo style touch and go" as soon as you try doing something even a little more complex the game's limitations (both technological and game design related) start to show up.

Edited May 13, 2021 by Master39

[mcwaffles2003]

2 minutes ago, Master39 said:

But that's just the first phase, while you're doing the same for Minmus you start sending mapping satellites, prospecting probes and maybe and orbital scientific outpost on Mun, and then that orbital station act as a starting point for "the Martian" style expeditions, with crews building temporary habs on the surface and doing longer term exploration and experiments and maybe using early IRSU technology. Again this lasts for 2-3 missions and after that the mining and permanent base building gameplay kicks in, leading to colonies (I think there should be a difference between a temporary hab, a mining/scientific outpost and a colony).

100% agree. I don't want to be "forced" to go to a planet 5 or so times, but I want a staged approach to be incentivized by being useful to the player.

Lets say I want to make a Eeloo colony to harvest a specific resource there. I feel a good game cycle that a player could fall into would be:

1. Discover Eeloo via telescope
2. Send probe to Eeloo to get a rough idea of its conditions and learn about it
3. Send relay probes to establish a communications system with near Eeloo orbiting satallites
4. Send a mapping probe to learn of Eeloo's terrain and rough composition (this is where the player discovers that Eeloo may have a resource that could be helpful and isn't found in any useful abundance nearby)
5. Send prospecting probes to possible resource hotspots to confirm the resource is there in useful abundance (maybe even include some false signatures to incentivize prospecting so players don't skip this)
6. Send in the kerbals and begin a colony near the useful resource
7. Establish a mining post at the hotspot while setting up the logistics network to supply the new Eeloo colony

That's 7 missions to a single body and the player is creatively involved the entire time. Each step is useful to achieving the next step and the player isn't being dragged along by some obscurity. They wanted the resource and had to come up with a plan to procure it on their own.

[mcwaffles2003]

31 minutes ago, Master39 said:

the game is built for that "apollo style touch and go" as soon as you try doing something even a little more complex the game's limitations (both technological and game design related) start to show up.

KSP 1's real downfall, in my opinion. People get mad at the tech tree and contracts system, but to me, this is the thing that hurts it the most.

It was just made too early by an ameatuer (no offense intended, the game is still a masterpiece for what it has done and how original it is)