Poll: What's Wrong with Stock Science

[Snark]

On 10/22/2018 at 3:02 PM, natsirt721 said:

This poll was created based on my own experience with career KSP and about two hours of googling.  If you think I have left out a grievance leave a comment, and I'll add it to the poll.

Unfortunately it won't let me vote in the poll, because none of the items listed in the "what do you dislike" question apply to me, and if I leave it unchecked, I get an error message saying that the item is required. 

Would be helpful to add a couple more options to that question:

"I think the science system is fine as it is"

...and...

"Other issues not listed here (please explain)"

[sturmhauke]

17 minutes ago, AlamoVampire said:

To me it is poorly balanced. So much so that a simple command pod with the basic starter experiments can gather enough points on the pad to let you unlock enough parts to vaccuum thee KSC deeply enough to fill the tree if not fully then nearly so. At least this was the case in mid/late 2016 when last I tried career.

Uh... I agree with the sentiment, but it's not that bad these days. You can get enough points to unlock a few nodes, but nowhere near the entire tree. You can do it just with Kerbin/Mun/Minmus though.

[Superfluous J]

1 hour ago, sturmhauke said:

I agree that it represents something more complex, but in real life we have all kinds of probes sending all kinds of data remotely. Data from the Webb telescope or Parker probe wouldn't be worth any more if we brought the sensor packages back to Earth.

I agree that there is a place for an experiment that does not require a Kerbal and can transmit for full. I'd hate to see another one added though without at least some thought put in to balance of how much science you get. You can already unlock the entire tree from the science in Kerbin's SOI. I'd hate it if you didn't even need to leave Kerbin's atmosphere

I wouldn't mind converting the atmospheric pressure doodad into something, or maybe giving the telescope a science return that can just be transmitted. Then you wouldn't need new parts at least.

EDIT Or what if the Science Jr was repurposed so that you have a choice when transmitting, either 100% value with no reset ever (it becomes "broken") or the normal value with the option to reset with a scientist/lab like it is now. It is the heaviest part so it makes sense that it should have more capabilities. And then you can kind of RP that it's a special use part that was built just for this mission.

Edited October 25, 2018 by 5thHorseman

[Fraktal]

On 10/24/2018 at 5:15 AM, Francois424 said:

Restricting science gains thru the game difficulty menu is flawed.  As it reduces science from Kerbin as well, meaning it's an absolute grind to get off the ground, so to speak.

Agreed on that part. I'm currently running a 10% Science game and got stumped where I got literally all science I could out of Kerbin, including the KSC mini-biomes (which alone took hours due to the extremely glitchy biome detection), as well as Mun orbit, and I'm currently stuck not being able to build a lander I want. That is, I don't want to include instruments that require a reset without bringing a scientist along, but I'm 41.5 points short of getting inline lander cockpits or the OKTO. I could go to Minmus and get the science from there, but I want to do everything in order and it feels sequence-breaking to go to Minmus orbit before Mun orbit. Another option I'm considering is using a manned pod to haul a Stayputnik to munar orbit, decouple (since I don't have docking rings or stack separators yet, the probe's engine will have to blast the decoupler away) and attempt a highly inefficient manual landing (the last time I tried a fully unmanned Stayputnik launch to the Mun, I ran out of fuel before even completing the transfer burn despite having packed nearly 6000 dV).

The KSC mini-biomes are the only reason I even got this far. Because if I hadn't gone grinding, I'd be two or three nodes short of where I am right now and would've had to choose between developing aircraft and sweating getting to the Mun, or developing Mun travel and trying to hit Kerbin biomes with suborbital launches.

In any case, what I think would definitely help in reducing the grinding of the early game of a low-Science game is zeroing out the KSC mini-biomes but increasing Kerbin's multiplier so that Kerbin as a whole still gives the same amount of science.

10 hours ago, LorenLuke said:

Sort the tech tree into part classes and types with each low level available- solid propellants, fuel tech, structural reinforcements, probe cores, capsules, space planes, reentry tech, EVA tech, science processing, and power systems, so the result ends up being more nodes, but less restrictions on the paths you can take. Want to not do planes? Great. Want to focus probes? Great. It expands the play style options to suit player decisions rather than force them into a box.

Yes. This.

What we need is not a longer tech tree with more nodes per branch, but a wider tech tree with more branches. Some things are not very intuitive: fuel and rocket engines get their own nodes, but science instruments, solar panels and probe cores are all jammed together? What does a barometer have to do with safe landings? Why is the Stayputnik only unlocked at a point where the player no longer has any use for it? A related problem is nodes that unlock too many parts, including stuff the player is never going to use. A good example of this are the FAT airplane wing parts, which are unlocked alongside regular wings but nobody uses them due to their incompatibility with elevons and being too heavy for the landing gears that get unlocked alongside them. By the time you have gears that can actually hold their weight for more than short hops, you have better wing parts.

When a part is only practical at the point where you no longer need it, there's something wrong there. I mean, I could put the Stayputnik to the Start node, but it won't do any good without a reaction wheel and a battery, now would it?

56 minutes ago, 5thHorseman said:

I agree that there is a place for an experiment that does not require a Kerbal and can transmit for full. I'd hate to see another one added though without at least some thought put in to balance of how much science you get. You can already unlock the entire tree from the science in Kerbin's SOI. I'd hate it if you didn't even need to leave Kerbin's atmosphere

I wouldn't mind converting the atmospheric pressure doodad into something, or maybe giving the telescope a science return that can just be transmitted. Then you wouldn't need new parts at least.

EDIT Or what if the Science Jr was repurposed so that you have a choice when transmitting, either 100% value with no reset ever (it becomes "broken") or the normal value with the option to reset with a scientist/lab like it is now. It is the heaviest part so it makes sense that it should have more capabilities. And then you can kind of RP that it's a special use part that was built just for this mission.

I agree with this as well. Been thinking about something in this regard a while ago: simple experiments (temperature, barometer, crew report, EVA report) should transmit for 100%, the rest should remain lossy BUT the loss rate should decrease as the Tracking Station and the R&D Facility get upgraded (due to increased reception gain and better data encoding/compression, respectively).

Additionally, for repeatable experiments (ie. those that must be run multiple times per biome to get all science out of it) except surface samples, I'd make it so that even if you return the data instead of transmitting it, you only get the full result if the instrument comes back as well. That is, if you run a goo experiment, take the data out of the goo canister, put it in a command pod or experiment storage unit and only return the data, you don't get all the science from it. You don't lose as much as if you had transmitted it, but you still lose something. But if you bring the goo canister back to Kerbin and recover it, you get full science from it, no deduction. This way the player is encouraged to try and bring those expensive instruments back home, which becomes tricky with materials bays due to their reentry characteristics. Again, this does not apply to surface samples which of course have no instruments to bring home. And this loss rate too would decrease as the R&D Facility gets upgraded.

[Pthigrivi]

31 minutes ago, Fraktal said:

A good example of this are the FAT airplane wing parts, which are unlocked alongside regular wings but nobody uses them due to their incompatibility with elevons and being too heavy for the landing gears that get unlocked alongside them. By the time you have gears that can actually hold their weight for more than short hops, you have better wing parts.

I quite like the FAT wings (I just think they look better), but I take your point on sequencing. Another example is the TD-25 decoupler and rockomax adapter which come before you have any 2.5m tanks to put them on. 

[Fraktal]

Two FAT wings, one 1.25 tank and two 0.75 tanks caused landing gear bouncing from overstress before 1.5. If you want to decrease weight by ditching the 1.25 tank for a Structural Fuselage and putting the fuel in multiple 0.75 tanks, you have to research another node which means another several hours of grinding in a low-Science game and would cost you enough points to put you halfway towards a node on the next tier. It's simply not worth it.

[mistergoblin]

I'm fine with the mechanics of it, but would love to see more and more varied science experiments.

Like a radial-mount Science Jr like SSTU has. Maybe some lower-profile goo experiment part also.

Add radiation sensors to the list of basic rerunnable experiments.

Or some later-in-the-tech-tree part that contains all the basic rerunnable experiments.

Or something kerbal-deployable you can bring along and leave on the surface, which accrues science over time like a mini lab.

Some smaller atmospheric sensors that you can mount on smaller probes, and get smaller amounts of data to transmit from smaller antennae.

Maybe an option for a mission-long experiment running in the background, like a kerbal wearing a med sensor and you get science points at the end for the med data gathered.

Maybe even some imaging parts where you can frame, zoom, and shoot amazing in-game photos, transmit them back, and get science points, and also sell the images to the press or something for money and reputation points. Their value can decrease, so the first close-up image of Duna will be super valuable, then decrease with each subsequent image, maybe per biome. 

Maybe have the planets in tracking station be basic sphere geo without the proper planet model, which is only revealed when you actually get a craft to fly by, orbit, or land.

I could go on, but this is just a wish list. I've been enjoying the science system as-is personally.

[Fraktal]

Forgot to mention this in my above rant. Why are the big, high-payoff experiments given to the player before the low-payoff ones? It reduces the incentive of even bothering with the latter. It should be the other way around: simple stuff like thermometer, barometer and seismometer coming first, then the goo canister and materials bay later on so that the player's science payoffs increase alongside the costs of later nodes. Small cost, small reward; bigger cost, bigger reward.

Edited October 25, 2018 by Fraktal

[FinalFan]

I like some of the ideas in this thread but others are almost terrifying.  Science per day?  Seems like it punishes you for not timewarping.  

I wholeheartedly agree that the running and collection and inspecting of already collected experiments is way too clunky, but I'm not sure what could be done to fix the latter while staying simple.  

I disagree with the proposal to essentially do away with manual-reset experiments.  I think it is sort of neat to have a couple special ones that need a hands-on technician (or an advanced lab) to prepare for reuse.  But I like the idea of those being given a "one and done" full-value science transmission option.  It would add options to no-return missions and make me feel a lot less silly when accepting contracts for goo-equipped satellites.  

However, I do think there is at least one thing that could be done to help the current state of science activity that seems like it might be relatively easy to do.  Add a feature to the storage container to "run all experiments".  It is already capable of collecting them, so why not let it run them? 

[DerekL1963]

6 hours ago, AlamoVampire said:

So much so that a simple command pod with the basic starter experiments can gather enough points on the pad to let you unlock enough parts to vaccuum thee KSC deeply enough to fill the tree if not fully then nearly so.

That is, if you choose to vacuum up all the science - there's nothing intrinsic to the game that forces your hand.  Or, to put it another way, your playstyle choices are not indicators that the system is broken.

You also have to consider that the availability of that much science at KSC gives not-rocket folks who (for whatever reason) want to play Science (or Career) mode a head start without having to build rockets.

 

6 hours ago, AlamoVampire said:

Science needs to be tweaked to be made more meaningful and to give us a reason to push beyond LKO.

I already have a reason - I hate the KSC biome grind and take the option (make the choice) to avoid it by going to orbit and beyond.

[DerekL1963]

3 hours ago, Fraktal said:

Agreed on that part. I'm currently running a 10% Science game and got stumped where I got literally all science I could out of Kerbin, including the KSC mini-biomes (which alone took hours due to the extremely glitchy biome detection), as well as Mun orbit, and I'm currently stuck not being able to build a lander I want.

Honestly, reading your post I'm not hearing a broken system.  I'm hearing a player that made a choice starting the game (setting science returns to 10%) and is now unhappy with the consequences of that choice.

I've been there too, and I just deal and do whatever I can to work through the problem... Or I restart after backing the difficulty down a notch.  That's just the nature of playing computer games.  Player choices matter, and sometimes players make the wrong choice.

[natsirt721]

15 hours ago, Pthigrivi said:

From this I think we can glean a few things:

• Science quantities and distribution are polarizing, but mostly balanced. Similarly the delegation between contracts and science appeals to and offends a roughly equal number of players.
   
• The transmission loss and science points mechanics are in principle broadly approved of. 
   
• The process of gathering science, the tech tree, and the mobile processing lab balance have more detractors than supporters. I will include clickiness and data management in this category as I imagine there would be very few votes for "Make science more clicky and cumbersome." 

I was going to wait for a few more people to respond before drawing conclusions, but this sums up both what I expected coming into this, and what I interpret from the results so-far.  If that makes this poll somewhat self-indulgent, well, so be it :p.  As the results have been coming in I've been formulating what a rework would entail and how to do it right - this started out mostly as a comm-net rework in my head, but my thoughts on that would involve rebalancing science gathering, so here we are.

I especially agree with your distinction between transmissible and processable experiments - the current transmission efficiency model, while somewhat effective, doesn't really make sense to me.  Instruments produce ones and zeros, and lossless digital data transmission is a thing, as long as you are willing to wait long enough.  As you note, this simplifies a 'run all - send all' interface to mitigate the clickfest, and emphasizes multiple missions for unmanned landers, sample return missions, and crewed missions.

As far as time-based mechanics are concerned, my conceptual rework will hopefully address this.  I recall reading a post that said that HarvesteR was opposed to wait based mechanics, but not time based mechanics.  Currently, my solution involves exponential curves, data rates, and significant EC costs, but I'll elaborate on this later once I sort out the big picture.

On the surface, I don't seem to like the time-based milestone system - however it is definitely food for thought. Perhaps it could be integrated with the persistent milestone system like someone (the name escapes me) suggested previously.  

The tech tree could use some work.  I agree with the 'wider, not deeper' attitude, and will probably sort something out based on current examples.

[Fraktal]

1 hour ago, DerekL1963 said:

Honestly, reading your post I'm not hearing a broken system.  I'm hearing a player that made a choice starting the game (setting science returns to 10%) and is now unhappy with the consequences of that choice.

I've been there too, and I just deal and do whatever I can to work through the problem... Or I restart after backing the difficulty down a notch.  That's just the nature of playing computer games.  Player choices matter, and sometimes players make the wrong choice.

It's not the choice that's the problem here. I made the choice specifically because I didn't like how easy it is to get literally hundreds of points of science with minimal effort. I wanted the game to be long, where progress is a matter of effort, not choice. I do not regret 10% science whatsoever because the game feels meaningless if there is no effort behind it. The problem I'm having is that I've "fallen into a gap", so to speak, where I'm able to do what I want, just not how I want. I don't need an OKTO or an inline cockpit to land on the Mun, hell, I've pulled off Munar landings with even less tech before (a 100+ part monster of a rocket built entirely out of tech 3 or lower parts as an experiment at getting to the Mun without Terriers). I just want to do it that way but I can't because I ran out of science to collect halfway to the node I need.

The problem isn't with the mission, it's with the mission profile:

1. Kerbin Explorer 1/A: manned orbiter, no landing.
2. Kerbin Explorer 1/B: manned orbiter, unmanned one-way lander.
3. Kerbin Explorer 2/A: manned lander with no orbiter component.
4. Kerbin Explorer 2/B: manned lander, unmanned orbiter with relay antenna.

Phase 1 is complete, phase 2 is where I'm stumped at. I can do phase 3 with the tech I have right now and phase 1 has enough range to reach Minmus and gather science there for phase 2, but I'd like to do phase 2 at the Mun first if at all possible.

Edited October 25, 2018 by Fraktal

[DerekL1963]

20 minutes ago, Fraktal said:

It's not the choice that's the problem here. I made the choice specifically because I didn't like how easy it is to get literally hundreds of points of science with minimal effort. I wanted the game to be long, where progress is a matter of effort, not choice. I do not regret 10% science whatsoever because the game feels meaningless if there is no effort behind it. The problem I'm having is that I've "fallen into a gap", so to speak, where I'm able to do what I want, just not how I want.

If you don't think your choice is a problem, then...  I honestly don't see what your complaint is as every one of your complaints are a direct consequence of that choice.  You can't have it both ways.  You can't claim that you're happy with your choice to make the game harder, and also complain when the game turns out to be harder.  These two things are mutually incompatible.

Having to make an unpleasant tactical choices because of your previous strategic or tactical choices isn't indicative of a broken system.  That's how games are supposed to work...  For a player's choices to be meaningful, there must be the possibility of undesirable scenarios and unintended consequences.

[sturmhauke]

I haven't been playing KSP as long as some of you. I played a little bit back when the Mün was the new hotness, but then put it down and didn't pick it up again until 1.3.

So I don't know how career mode or science mode used to work, only how they work now. I get the impression that the tech tree progression maybe used to make more sense, but new parts were added over time and just kind of stuck in wherever. I'm in favor of things like reorganizing the tree into more closely related parts and themed branches.

[natsirt721]

10 minutes ago, sturmhauke said:

I get the impression that the tech tree progression maybe used to make more sense, but new parts were added over time and just kind of stuck in wherever.

No, its always been kind of a cluster...

@DerekL1963 @Fraktal this discussion is about stock science settings - DerekL is correct in saying that your personal gripes with your choice are not relevant to the discussion at hand.  That being said, Fraktal's comments about being 'stuck in a gap' are indicative of imbalanced science gain in the early-mid game, a sentiment I have seen expressed in other places as well.

Edited October 25, 2018 by natsirt721
added second paragraph

[DerekL1963]

1 hour ago, natsirt721 said:

That being said, Fraktal's comments about being 'stuck in a gap' are indicative of imbalanced science gain in the early-mid game, a sentiment I have seen expressed in other places as well.

"Stuck" implies not being able to proceed, and even he admits that's not true.  He can proceed, just not in the way he wants to.  That's a playstyle choice, not a fault in the tree.

On the larger topic of imbalanced gains, you still have to be careful to distinguish between problems caused by player choices of how to climb the tree and actual deep problems in the tree.  I have yet to see a true case of the latter in the wild.  There's some minor fixes needed here and there to make a good general purpose tree... (Particularly in the lower tree and the CommNet/solar electric branches.)  But I am not sold on the idea that a wholesale overhaul of the system is required.  I am convinced there's a lot of people who've never really thought through a strategy for climbing the tree (you really have to think two steps ahead to do well) and/or approach it with the "gotta catch 'em all" mentality and thus forces themselves into needing too much science for their actual game play needs.

[MajorTomtom]

Just to come back on the science per day revenue, a work around would be a fixed science per space mission (and eventually a science per aircraft mission) revenue, that would be proportional to your R&D center level and your reputation. That way warping your way around is not an option, but you still have the same results, and its closer to real life.

The experiments gathering could also be made less clicky by providing either a shortcut "run all" (for the moment you have to set it up manually using actions shortcuts and its not really beginner friendly or interesting) or run science experiment automatically once entering a biome with a simple indicator in one corner of the screen and a "science window" where you can manage which experiments to store/send/process etc (with a send all button of course).

A rework for the tech tree could be the following:

-You don't spend science for nodes but just money, you just accumulate it

-Each node has a required overall science threshold

-A lot more nodes that each contain only a few linked parts (for example : ungimballed basic engines, upper stage engines, fixed solar panels, deployables, basic science experiments, batteries etc...)

That way you have the analog of real life R&D where some parts are developed because we have the technology matures (~enough science data since the beginning of space program to make it work) and the money to fund it.

[mikegarrison]

KSP is basically a sandbox game with "career mode" grafted onto it. It is an orbital mechanics and takeoff/landing game, not really a space program simulator.

The lab is really how all science in the game should work -- you gather data and over time your labs turn that data into usable science. It would be kind of cool if you actually had to do something once you landed somewhere. Like you had to collect temperature readings from 5 locations nearby your lander, get samples from three locations, etc. Then bring it back to home or to a mobile lab and watch it turn into science as your scientists work on it.

But like many other aspects of playing KSP, this would only work if the core mechanics were very different. Playing with life support mods makes anything that has to do with the actual passage of time quite a bit more interesting. But time warp gets messy.

If you were designing KSP2, this is the kind of gameplay you might want to design into it right from the start. But it's very difficult to really integrate into KSP as it is today.

[helaeon]

I've been playing some No Man's Sky recently... I think what is missing is discovery. To me the best part about NMS is landing on a world and seeing what weirdness is going to come out. Maybe a little procedural generation would be a good thing. I'm into KSP something like 5k hours now, I've seen everything, I've been everywhere. I think the only way to deal with folks like me is have the game come up with subtle interesting things so each landing has that bit of the unknown.

I'm not a fan of load up ship with every  experiment you've unlocked and then hop biome to biome and run them all for the boat of science. Labs are better. Still kind of weird that you unlock technology with science points though, maybe game would make more sense if science -> money and you use cash to unlock parts (I saw that above). Or nodes unlock with cash and meeting some other condition. I think I saw that up above.

Have science experiments chain into other ones. Discovered possible microbial life on Duna or Laythe? Better send a base to do on the spot science, find more samples, and maybe grow that stuff. I think the science itself maybe needs to be a game of its own and on top of that getting to new things drives the engineering of new vehicles. I think I'm kind of talking about combining the contracts and science system not unlike what some folks have done with contract configurator. This would also be a way to inject Making History type missions into Career mode.

[LorenLuke]

16 hours ago, DerekL1963 said:

"Stuck" implies not being able to proceed, and even he admits that's not true.  He can proceed, just not in the way he wants to.  That's a playstyle choice, not a fault in the tree.

Is the fact I think that probes should come first for space exploration a fault of my thinking rather than that of the tech tree, because my ideas don't just happen to align inside this box of possibilities that can be done?

Could not that be the fault of the tree for not allowing such a playstyle, instead of expanding the available, feasible styles of play?

 

You do get quite the science glut at the start of the game if you choose to thoroughly explore Kerbin and the KSC (ironic since, not only do you live on a supposedly explored and colonised planet, apparently the Kerbals are learning things they didn't know about the very place they work at). But then you get this strange effect where the expenditure of effort to available science takes a nose dive. You can attempt to biome hop, hoping it doesn't foul your mission and your Kerbals' chances of getting back safely... But the cost requirement per science of a Mün lander seems to jump dramatically over LKO exploration. 

However, science does yet another odd thing as we branch out of Kerbin's SOI, it increases again, where a probe telling you that, predictably, the space around Dres is cold, has no pressure, and make liquid goo freeze is so much more groundbreaking a discovery than saying so about Minmus. 

The problem comes from the gameplay loop of exponential gains which supports exponential exploration. The whole of gameplay is 'spend points to get places, get points from getting to places'. There needs to be an alternative or more balanced, linear path to collecting science, rather than this sort of 'sideways hourglass' that bottlenecks us into cheesing the KSC on harder difficulties to get anywhere.

[Tyko]

1 hour ago, LorenLuke said:

Is the fact I think that probes should come first for space exploration a fault of my thinking rather than that of the tech tree, because my ideas don't just happen to align inside this box of possibilities that can be done?

Could not that be the fault of the tree for not allowing such a playstyle, instead of expanding the available, feasible styles of play?

The game does provide way for alternative play styles by being very mod-friendly. There are several mods out there that convert the tech tree and you can even tweak them further yourself. I agree that starting unmanned is more interesting for me, so I use a modded tech tree.

Have you considered that?

6 hours ago, helaeon said:

I'm into KSP something like 5k hours now, I've seen everything, I've been everywhere. I think the only way to deal with folks like me is have the game come up with subtle interesting things so each landing has that bit of the unknown.

Have you considered a planet pack? GPP or RSS both give all kinds of new places to visit

6 hours ago, helaeon said:

Have science experiments chain into other ones. Discovered possible microbial life on Duna or Laythe? Better send a base to do on the spot science, find more samples, and maybe grow that stuff

This would be so COOL!!

[Pthigrivi]

22 hours ago, DerekL1963 said:

On the larger topic of imbalanced gains, you still have to be careful to distinguish between problems caused by player choices of how to climb the tree and actual deep problems in the tree.  I have yet to see a true case of the latter in the wild.  There's some minor fixes needed here and there to make a good general purpose tree... (Particularly in the lower tree and the CommNet/solar electric branches.)  But I am not sold on the idea that a wholesale overhaul of the system is required.  I am convinced there's a lot of people who've never really thought through a strategy for climbing the tree (you really have to think two steps ahead to do well) and/or approach it with the "gotta catch 'em all" mentality and thus forces themselves into needing too much science for their actual game play needs.

I also don't think it needs a completely new concept, but it could use a thorough rebalancing. There are a few glaring mistakes and lots of minor ones. The one example I would give of a truly deep problem with the tech tree is that it doesn't cater to a huge number of players who prefer to start with planes or with probes. Im not one of them, but I respect that gamestyle and think they're underserved. The split for these playstyles just comes a bit too late. I get that Squad wants brand new players to open up the VAB and have their first mission be a crewed flea mission to get used to the basic controls and staging, but immediately after that there should be the option to branch if a player wishes. 
 

On 10/25/2018 at 2:30 PM, natsirt721 said:

I was going to wait for a few more people to respond before drawing conclusions, but this sums up both what I expected coming into this, and what I interpret from the results so-far.  If that makes this poll somewhat self-indulgent, well, so be it :p.  As the results have been coming in I've been formulating what a rework would entail and how to do it right - this started out mostly as a comm-net rework in my head, but my thoughts on that would involve rebalancing science gathering, so here we are.

I especially agree with your distinction between transmissible and processable experiments - the current transmission efficiency model, while somewhat effective, doesn't really make sense to me.  Instruments produce ones and zeros, and lossless digital data transmission is a thing, as long as you are willing to wait long enough.  As you note, this simplifies a 'run all - send all' interface to mitigate the clickfest, and emphasizes multiple missions for unmanned landers, sample return missions, and crewed missions.

As far as time-based mechanics are concerned, my conceptual rework will hopefully address this.  I recall reading a post that said that HarvesteR was opposed to wait based mechanics, but not time based mechanics.  Currently, my solution involves exponential curves, data rates, and significant EC costs, but I'll elaborate on this later once I sort out the big picture.

On the surface, I don't seem to like the time-based milestone system - however it is definitely food for thought. Perhaps it could be integrated with the persistent milestone system like someone (the name escapes me) suggested previously.  

The tech tree could use some work.  I agree with the 'wider, not deeper' attitude, and will probably sort something out based on current examples.

Sorry to step on your toes! But I think we agree on the early conclusions. I actually think Roverdude is a brilliant designer of underlying mechanics. Where some of the breakdown is occurring is in the UI, making it really visually clear how transmission works, predicting transmission ranges and especially reducing all of the repetitive clicking and windows. Thats especially why dividing things up between transmissions and samples makes sense I think. I take @FinalFan's point about having some manual interface with the Goo and Materials Bays, but it gets real old climbing out and reseting them every time you enter a new biome. What might work better is to put Goo and Materials bays into the "processable" category, meaning you could take as many readings as you wanted without having to reset (perhaps with the aid of an experiment storage container), but in order to capitalize on the data you'd have to return it home or manually transfer it to an MPL for processing. 

I go more into the whole time-based World First bonus idea here. mainly its about moving Milestones out of the UI darkness and making them up-front and center in Mission control. Just like now you'll always get paid to complete them, but if you complete each one by a specific date you get a World First bonus on top. Whats nice about this is its clearly understandable to the player (Get to Duna in 300 days!), and its also scaled individually to the relevant body. In other words you aren't bleeding away rewards at a bunch of difficult to predict rates over time, you just get bonus rewards for meeting specific deadlines and are therefore disincentivized to warp through other time-based game elements (construction time, research, MPL processing, etc) or risk missing them.
 

Edited October 26, 2018 by Pthigrivi

[tater]

The linkage of planetary science to tech is bizarre. The entire career/science/tech paradigm needs to be thrown away and done well, instead.

[tater]

There is no altering science without rethinking what career is supposedly FOR.

Barring the ability to have an actual space race (which is implied in career since there are "firsts," which implies the possibility of not being first), what is the goal?

It's not some sort of space program management, since there is no possible way to fail unless you crank all the rewards down, and pretend like that makes sense. It's not a race. It's not about exploration (if it was, that would presumably be what "science" mode should be, right?)

I think it should be one or all of those things.

Program management? This doesn't mean spreadsheets, it means that you have a couple program types, a NASA analog, or private space venture (SpaceX/BO) analog. In the former, you'd have a budget, and you future budget would depend on meeting some milestones. You'd not get rewarded with funds, you'd design a mission (land on body X and return, for example), and you'd bid a budget to accomplish the goal. You'd need something like KCT, too. Private space would be different, you might get a grant for X, otherwise you launch stuff (or kerbal tourists, etc) for profit.

Exploration? In the NASA-analog model, some of this would be what drives future budgets. Otherwise, in general it should have "fog of war" where the player knows what they can know from Kerbin, and some stuff you learn by sending craft. Mars was the old "canal" drawings until Mariner took pictures. Jupiter's moons were dots. This would require some form of randomized solar system that the player learns about via doing science missions (abstracted----fly certain instruments, and you unlock useful data for actually visiting that world).