My Science Lab Manifesto

[Aegolius13]

I know there have been quite a few discussions about the game balance on the Science Lab / MPL, but I wanted to get all my thoughts down in one place.  In my mind, the part is possibly the biggest gameplay problem in the game today.  I'll start with a few opinions about the current state of affairs:

-The current lab's science processing function is hideously overpowered.  5x science output, without the need to even return the sample to Kerbin (which would get you even more points), is crazy.  It's way too easy to max out on science just by doing this.  Among other things, this devalues going interplanterary.  I think it's already too easy to max out on science just by strip-mining the Mun and Minmus, but the lab makes this problem even worse. 

-Time is not a meaningful limit on the lab's effectiveness, since for the most part you can warp ahead with impunity.  By the same token, scientist experience levels, which only effect research speed, don't matter.  (This is the same problem with engineers and mining, but that's well beyond the scope of this rant.) 

-The current processing function is counterintuitive, and generally not much fun to use.  The science vs. data distinction is confusing, especially to new players.  It does not make sense that all samples are destroyed when you feed them into the lab (I can buy this for physical experiments like the goo or Science Jr., but not for temperature readouts).  Having to bring up two samples (one to feed to the lab, the other take home), is fussy, but is at least is a little easier now with the science container.  And since you can't tell what experiments you've already fed to the lab, it's hard to remember what you need to collect if you do want to keep feeding the thing.  Finally, having to go back and transmit every time stored science gets to 500 is annoying, especially if you're doing large warps for an interplanetary trip.

-Allowing repeat processing of the same experiment is ridiculous and makes the MPL even more game-breaking.  Limiting it to one processing per experiment per MPL is not a significant limitation, since it's easy to just launch another one (and you may do that incidentally, by upgrading your main space stations).  

-While realism is not always necessary, the lab just does not make all that much sense.  Why can two Kerbals floating in space get 5x more knowledge from a sample than all the assembled (but admittedly, fairly dumb) scientists back on Kerbin?  I do get that there would be some additional knowledge from analyzing a surface sample from the environment it came from.  But not 5x more.  And again, this rationale does not hold up for something like temperature readings, which are just data.

-The MPL's other features (resetting science experiments, enabling KerbNet, and leveling up Kerbals) are fine, and serve to make the MPL kind of a remote hub for your activities, which I think is a good thing.  But these features aren't all that useful by themselves, and they don't have all that much to do with the part being, you know, a science lab.

-I don't want to sound like I'm opposed to using the lab to get science points.  I very much want to have a science lab up in space, doing science stuff.  That just seems like one of the things you should be doing in a space simulator, and it's one of the main reasons to have a space station or surface base.  The game should reward you for doing this, by providing some (reasonable) output of science or other resources.   Moreover, I want my scientist Kerbals to have a function (resetting experiments is fine, but again, not enough by itself).  And I want leveling up your scientists to help you in some meaningful way. 

 

All that said, what would be a good method to fix the MPL into something that's balanced, still useful, and more fun/intuitive to use?  Frankly, I don't claim to have a complete answer.  But here are some suggestions, some or all of which might be involved in a given fix.  

-Scrap the current "MPL eats experiments, generates a new resource called data, eats the data and then poops out massive science points" mechanic.  It's bad.  Just bad.  But at least, cut the output back to something more reasonable, and remove the repeatability in different labs.

-Use the lab as a means to get most or all  the "recover" value of experiments without having to send them back to Kerbin manually (I think the lab did something like this in the early versions?).  Maybe scale the recovery value to the experience level of the scientists operating it.   This goes with the idea that the MPL is not "better" at science than the folks back home; it's just more convenient.  To make this function meaningful, they might have to decrease radio transmission values relative to fully recovered values.  That is fine by me, again especially for goo and the other physical data samples.  

-Perhaps provide a small science bonus for analyzing the physical samples in the SOI or biome in which they were found.  

-Offer contracts to go out and get a particular sample from a particular biome, and take it back to an MPL for testing.  This makes a lot of sense - suppose the Kerbin State University just got a grant to test the soil for the Mun's East Crater for the presence of isoexplodium or something, and they want you to help out.  These contracts could reward you with both cash and fairly significant amounts of science.  

-When "launch a space station" contracts require a science lab, require it to be manned by scientists.  This would be a more significant issue for stations outside Kerbin's SOI.

-It's probably fine to leave the other auxiliary functions of the MPL as is.  They're just not that big of a deal.  It might make more sense to limit the level-up function to scientists, but on the other hand it might be odd if there were no comparable level-up methods for the other classes.  

 

I still feel like there's something missing, though.  I would like to have a reason to put my Kerbals up there and have them sit around sciencing, like on the ISS.  And none of my suggestions get to that all that well.  But maybe that's just an inherent problem when time is basically unlimited unless you're using a life support mod of whatnot.  

Sorry for the lengthy rant.  Feel free to add any other ideas for rehabilitating this poor misunderstood part!

[Allocthonous]

You make some good points. On my last career game in 1.13, I avoided using the MPL entirely because I felt it was too easy.

So, what uses do space labs have in real life? For the most part, we haul or transmit experiments from the earth up to the lab, have the astronauts perform the required experimental procedures and observations, and then haul them or transmit them back down to earth.

This is kind of what the Materials Observation Bay and the Goo capsule are, but those can be done entirely with probe cores, so they don't really fill the same niche. I'd like to see an 'experiment package' part that has to be hauled up to an MPL to allow it to generate science. Heck, make 'experiments' a consumable resource. Pump them out of the experiment package to the MPL, and pump 'finished experiments' back. Give a science bonus for recovery of finished experiments on Kerbin, in addition to some quantity of transmittable science from the MPL.

I would keep the in-situ analysis function you're describing above, as well. That would give some incentive for hauling an MPL out to Jool or wherever instead of just keeping them in LKO.

[Aegolius13]

4 minutes ago, Allocthonous said:

So, what uses do space labs have in real life? For the most part, we haul or transmit experiments from the earth up to the lab, have the astronauts perform the required experimental procedures and observations, and then haul them or transmit them back down to earth.

This is kind of what the Materials Observation Bay and the Goo capsule are, but those can be done entirely with probe cores, so they don't really fill the same niche. I'd like to see an 'experiment package' part that has to be hauled up to an MPL to allow it to generate science. Heck, make 'experiments' a consumable resource. Pump them out of the experiment package to the MPL, and pump 'finished experiments' back. Give a science bonus for recovery of finished experiments on Kerbin, in addition to some quantity of transmittable science from the MPL.

That makes sense.  I think that's kind of what I was getting at with my last thought, but was not sure how to capture that ISS-type thing in a gameplay mechanic.  Also seems to tie into the idea of getting a bonus for local processing of physical experiments.  Besides the home SOI/biome, I guess there could be more generic bonuses for analyzing a sample in zero-gee, or in high-gee environments (which I guess means Eve?).  

On the other hand, that's also kind of like a version of what we have now.  E.g., take a Kerbin surface sample, haul it up to the lab to process it (meaning see what happens to it in zero-gee), send results back for more science.  That's a reasonable way for a lab to work, but in the present gameplay leads to the absurdities mentioned above.  Maybe the discrete "experiment package" idea would prevent running ridiculous amounts of science through the lab as happens now.  The experiment package could also part of a contract, kind of like tourists: take X package to Y lab, process it, bring it back, get science and money (or no money, either way could work).  If you only get them once in a while, less prone to abuse.  

One final thought - there might be a conceptual difference between a science lab (providing space-specific science value, like the ISS) and a mobile processing lab (doing regular, Kerbin-style science on experiment results so they don't have to get sent home).  Which way the part evolves might depend on which function was prioritized.  

[Blaarkies]

1 hour ago, Aegolius13 said:

-Offer contracts to go out and get a particular sample from a particular biome, and take it back to an MPL for testing.  This makes a lot of sense - suppose the Kerbin State University just got a grant to test the soil for the Mun's East Crater for the presence of isoexplodium or something, and they want you to help out.  These contracts could reward you with both cash and fairly significant amounts of science.  

Completely agree, the MPL destroys any game where the player presses that "Transmit science" button.

But be careful with science rewards in contracts, they follow the same path to evil that the MPL does. When i first started KSP, back in the "First Contract" update, i quickly got hooked on the part testing contracts because they gave a significant amount of science("...just need a bit more to get that engine").
This unlocked the entire techtree way too soon, but now in 1.2 it feels like contracts give way less science(1-5sci).

Regardless, all science points are sacred. They should not be clone-able(current MPL) or infinite(old contracts) or too easily obtainable(Minmus farming)...at least asteroid farming takes a lot of effort, and forces you to go halfway interplanetary(quite literally)

Your idea is better than the one i had about making the MPL a special kind of experiment on it's own

[Allocthonous]

1 hour ago, Aegolius13 said:

One final thought - there might be a conceptual difference between a science lab (providing space-specific science value, like the ISS) and a mobile processing lab (doing regular, Kerbin-style science on experiment results so they don't have to get sent home).  Which way the part evolves might depend on which function was prioritized.  

This had occurred to me, and I'd love to see a processing lab part that's entirely separate from an experimental science lab. I would say that it shouldn't process the experiments that can be reset without a scientist, though--crew and eva reports, temp, pressure, seismic, and gravity scans. Only the ones that would actually involve a physical experiment needing processing. Samples, materials bay and goo, maybe the atmosphere sampler.

I think it comes down to the MPL being too abstract. It's sort of both types of lab at the moment, and it's terrible at both of them.

[Wjolcz]

Science system, ho! Yep. The whole thing is a mess. It started as a mess ans then got only worse because of this avalanche effect when devs tried to fix it.

I really like the idea of MPL being a part of a contract. That's way better idea than limiting the output, or whatever. What I would also like to have is an option that let's me choose which currency (science, funds or rep) science experiments generate. That thing alone would fix a lot of problems I'm having with the current science gathering system.

[Kerbonaut905]

The science as whole needs a revamp. It's too simple and undiverse IMO.