Rethinking the Lab Module and Science Processing

[Specialist290]

So it seems as though some people are wondering whether the new Science Lab is really worth as much trouble as it takes to deploy. I'll admit that I'm not 100 percent enthusiastic about the way it's currently set up myself now that I've had a bit more time to play with it, and I would like to see something being done with it that makes it truly unique and useful and not just a "booster" for probes. Thus, I'm proposing the following:

• Keep the ability to reset "used" experiment modules. This is incredibly handy.
  
• Rework the experiments processing into a full-fledged new feature rather than a transmission cap boost.
  

To elaborate on that second point, I'm basically envisioning the Lab as a module where kerbonaut scientists can run more advanced and more rigorous in situ experiments on materials than they could with the scant equipment available in a non-specialty pod, and can cross-reference the data they get with data from other experiments as well. Thus, I'm going to propose making this change to the science processing system: Instead of "boosting" a single report, my idea has you processing two reports together into a single, advanced report.

To go into a little more detail:

• Rather than being "enhanced," the two reports are consumed to create an all-new "advanced" report with a value independent of its sources. In other words, you can use it to "recycle" reports that you've already exhausted through conventional means and bring in some fresh science.
  
• All advanced reports transmit for 100% value. This allows the Science Lab to more effectively serve as a "science hub" without any need to constantly return vehicles to Kerbin.
  
• Each particular report combination only works once. This will help cut down on the "science spam" while at the same time encouraging players to explore more of the solar system.
  
• Bonuses are awarded for combining two similar reports, but you'll be penalized if they're too similar. Again, this is to encourage players to explore and experiment with different combinations. I'll admit that I don't have a good idea on how to implement this exactly myself.
  

Hope others find this constructive as well. I'll look forward to seeing other people's feedback on this idea, including ways I can streamline it without losing the essence of making the Science Lab much more useful than it currently is.

Edited December 20, 2013 by Specialist290

[Themohawkninja]

Now that sounds like a more useful lab to me.

Although, I don't exactly see the thematic explanation of the advanced report.

[Stargate525]

Although, I don't exactly see the thematic explanation of the advanced report.

Nah, I get it. For instance...

-Your thermometer reading shows slight fluctuations on Minmus' surface. Cross-referencing it with the results from your gravioli detector shows a correlation between temperature and gravity fluctuations.

-On Duna, the specific shade of red that the goo turns can be affected by different items in the materials bay.

-Tylo's gravity is unaffected by its earthquakes.

-On Laythe, you discover that the barometric pressure rises with the earthquakes that cause geysers.

Stuff like that.

[Specialist290]

Nah, I get it. For instance...

-Your thermometer reading shows slight fluctuations on Minmus' surface. Cross-referencing it with the results from your gravioli detector shows a correlation between temperature and gravity fluctuations.

-On Duna, the specific shade of red that the goo turns can be affected by different items in the materials bay.

-Tylo's gravity is unaffected by its earthquakes.

-On Laythe, you discover that the barometric pressure rises with the earthquakes that cause geysers.

Stuff like that.

Precisely. By cross-referencing different reports for correlations, kerbal scientists could see if there are any factors that interact in new and fascinating ways.

[Themohawkninja]

Oh... that's a neat idea.

Instead of just sending arbitrary numbers, you can send back a whole scatter plot showing a strong correlation between X and Y!

[thereaverofdarkness]

I love the new science lab module. I think it'd be neat if it's an early progenitor, and there would be a much larger version you get later in the tech tree which could perform long-term studies while you are away (on the time scale of minutes to hours) and could get a near-perfect return when completed. But bringing it back home should always give the most. Still, if you could get, say, 90% from temperature studies, 85% from goo, and 75% from Science Jr., then it would certainly be worth bringing along on a long mission to a distant planet.

[Moonfrog]

sounds like a good idea in the right direction.

[J.Random]

Cross-referencing is a neat idea, but I doubt it will be implemented. Imagine the amount of text it will require!

Right now we have 7(?) experiments (not counting crew reports, EVA reports, and EVA surface samples). So, that would be C(7,2) combinations = 7!/(5!*2!) = 21 different descriptions. Multiply that by biome count. Adding a single new experiment will bring it to 28 descriptions per biome. Another one will make it 36, and so on. That's A LOT of text, if you're not looking for some low quality copypasta.

Edited December 20, 2013 by J.Random
fixed C(x,y)

[Minarkhaios]

Or they could just ask the community to create the different messages.

Power in number and that sort of things.

[allmhuran]

I've said it before, but I would take it one step further, in a way that makes a hell of a lot of intuitive sense:

Samples really need to be analyzed before you can get any useful information out of them. So right now, being able to take a surface sample and then transmit the data back to Kerbin doesn't make sense. What data?

I would suggest that any sample based experiments (goo, materials bay, surface samples) should not have transmissible data. Either return the samples to the lab at KSC for the science or, and here is where I start aligning with the OP, use the lab module to analyze the samples and THEN transmit the data back. So, you can only transmit sample "data" if you have the lab to get the data out of the sample. Otherwise you have to return the sample. And remove diminishing returns. 100% value first go. Repetition for the sake of it is bad design.

Now, what you could do - and this again aligns with the OP - is make it so that an "unprocessed" sample only returns a tiny amount of data: Just the data you get from the astronauts making an observation. Call this a "sample observation" experiment, and again, allow transmission for 100% of the value of the sample observation. Then use the mechanic in the previous paragraph to do the "deep analysis" (if you have the lab available) to get the really juicy points of of that sample.

Example:

- Jeb gets duna surface sample. Takes it to command module. Looks at it. Transmits his observations. Science value: 20.

- Jeb then EITHER:

1) Returns to kerbin with the sample. KSC performs full analysis, 100 science value, or

2) Takes the sample to a nearby lab module. 100 science value, no transmission loss.

You could *even* do three levels - basic observation, mobile lab analysis, processing at KSC. If you do any of the "more valuable" processing, you should just automatically get the points for the less valuable processing at the same time. So returning any sample to KSC would get you all possible data out of the sample.

Meanwhile, make simple measurements (temperature, gravity, etc) transmissible for 100% value with, again, no repetition or diminishing returns. Because it makes no sense that you would have to return to Kerbin with a temperature reading to get the full "value" of the information.

This removes grinding (no repetition just for the sake of repetition), it makes transmissible data sensible, and it gives the science lab a truly useful purpose.

Edited December 20, 2013 by allmhuran

[thereaverofdarkness]

I personally dislike having a wide variety of scientific tools that simply collect their own unique data. It means that after I unlock them, I have to build a new spacecraft to visit all of the places I've already been to, in order to collect science with all of them. I'd prefer if we could instead have data collection materials early on, and later get various science modules and effects that allow us to do things that can increase our science output. Best of all, they should be able to work remotely. Example: new temperature processing computer, must scan temperatures in each biome over an extended period of time but can just transmit the data from already existing thermometers on-site to KSC where the computer is. Portable computers could be brought along on missions to collect and store the temperature data, if you prefer to return home rather than set up a communications network.

So basically you don't have to send a new mission all the way out to Duna to get new temp readings, instead you hop in your existing Duna settlement and send some missions out to various biomes with thermometers (because you already have them on Duna) to collect new readings. There is still work to be done after you research new tech, but you don't necessarily have to bring tools all the way from home every time.

Edited December 21, 2013 by thereaverofdarkness

[Specialist290]

Cross-referencing is a neat idea, but I doubt it will be implemented. Imagine the amount of text it will require!

Right now we have 7(?) experiments (not counting crew reports, EVA reports, and EVA surface samples). So, that would be C(7,2) combinations = 7!/(5!*2!) = 21 different descriptions. Multiply that by biome count. Adding a single new experiment will bring it to 28 descriptions per biome. Another one will make it 36, and so on. That's A LOT of text, if you're not looking for some low quality copypasta.

I'll grant that that's true; it would take a lot of work. That said, in a sense KSP's development process could work in our favor in this regard; Squad could implement the idea with a few basic placeholders to "demo" the feature among the general fanbase, then continue to elaborate on it and add more "fluff" as the game updates. As Minarkhaios mentions, they could also potentially turn to crowdsourcing, as long as there's someone at the other end reviewing submissions for general quality.

For the record, I wouldn't mind contributing a little bit of writing myself.

I've said it before, but I would take it one step further, in a way that makes a hell of a lot of intuitive sense:

Samples really need to be analyzed before you can get any useful information out of them. So right now, being able to take a surface sample and then transmit the data back to Kerbin doesn't make sense. What data?

I would suggest that any sample based experiments (goo, materials bay, surface samples) should not have transmissible data. Either return the samples to the lab at KSC for the science or, and here is where I start aligning with the OP, use the lab module to analyze the samples and THEN transmit the data back. So, you can only transmit sample "data" if you have the lab to get the data out of the sample. Otherwise you have to return the sample. And remove diminishing returns. 100% value first go. Repetition for the sake of it is bad design.

Now, what you could do - and this again aligns with the OP - is make it so that an "unprocessed" sample only returns a tiny amount of data: Just the data you get from the astronauts making an observation. Call this a "sample observation" experiment, and again, allow transmission for 100% of the value of the sample observation. Then use the mechanic in the previous paragraph to do the "deep analysis" (if you have the lab available) to get the really juicy points of of that sample.

Example:

- Jeb gets duna surface sample. Takes it to command module. Looks at it. Transmits his observations. Science value: 20.

- Jeb then EITHER:

1) Returns to kerbin with the sample. KSC performs full analysis, 100 science value, or

2) Takes the sample to a nearby lab module. 100 science value, no transmission loss.

You could *even* do three levels - basic observation, mobile lab analysis, processing at KSC. If you do any of the "more valuable" processing, you should just automatically get the points for the less valuable processing at the same time. So returning any sample to KSC would get you all possible data out of the sample.

Meanwhile, make simple measurements (temperature, gravity, etc) transmissible for 100% value with, again, no repetition or diminishing returns. Because it makes no sense that you would have to return to Kerbin with a temperature reading to get the full "value" of the information.

This removes grinding (no repetition just for the sake of repetition), it makes transmissible data sensible, and it gives the science lab a truly useful purpose.

This sounds vaguely familiar. I might have read one of your posts before and gotten a bit of "unconscious inspiration," at it were.

I definitely like the idea of transmitting data not actually consuming the sample, in contrast to how it works under the present system. I'm all in favor of there being only one experiment per container, but it makes sense that the experimental package itself wouldn't simply disappear after you've transmitted the data and would still be available for further study if you managed to get it home or to a lab that could effectively process and analyze it.

In light of both your comments and thereaverofdarkness's (which he later further elaborated on here), I'd propose a sort of three-tier system that combines aspects of everything we've mentioned so far, at least specifically for the physical sample-based experiments:

1. Basic collection and analysis, which can be transmitted for a small part of the experiment's value but doesn't consume the experiment itself.

2. Intermediate-tier processing, as you've described above, which can be done either by returning the sample to KSC or by analyzing it in a "basic" Mobile Science Laboratory.

3. Advanced Processing and Analysis, which can be done in an "upgraded" MSL and uses the system I've described in my OP, or a variation of it. (The reason it wouldn't be available at KSC could be justified in-character as the experimental data being more valuable if processed in situ, and out-of-character in terms of game balance by not allowing players to just cross-reference everything they've already returned to Kerbin for essentially "free science.")

As far as "upgrading" the Lab, I'm thinking specifically of having a non-part-based "upgrade kit" that can be added to a command pod, flown over to the Lab's location from KSC, and "installed" by having a kerbal carry it from the pod into the Lab, much like described here. From that point on, that particular Lab module could be used to conduct more advanced experiments as already described. It would also give a good in-character, in-game reason for rotating crews between the Lab and Kerbin without passively penalizing the player in the manner that a mandatory life support system might.

[allmhuran]

I would say that if you have three different experiments per sample, being

1) Basic observation by the astronaut

2) Analysis in a mobile lab

3) Analysis at KSC

then in terms of difficulty, 3 is harder than 2, which is harder than 1. So if points should reflect difficulty, I think return to KSC should give you the largest number of points, and doing a more difficult one should give you any of the less difficult points as well (because, clearly, if you can return the sample to KSC then your astronaut can do the basic observation alogn the way, and KSC must have all of the tools available in the lab module).

However, I see what you're going for with having the lab module provide more value than KSC analysis - I see that you're not basing it on the degree of difficulty, but rather providing a mechanism to increase the value of orbital stations, surface bases, etc. I can get behind that approach.

Edited December 21, 2013 by allmhuran

[thereaverofdarkness]

I was thinking a research project could improve the quality of both mobile labs as well as returned data, so that the improved field research is better than old sample returns, but returning samples is still best. If the top tier field research is almost as good as returning samples (like 80%) then returning samples will not feel like a necessity and people will be able to run a perfectly good space program without returning any, if that's what they want to do. But those who go above and beyond and return everything home will still get benefit for it.

example: soil sample process in mobile lab: 40/100, return 100/100

advanced tech: process sample in mobile lab: 120/150, return 150/150

And I'd also like an option to process quickly for small return or take time to get a bigger return. Example:

process in mobile lab for 5 seconds: 25/100, process in mobile lab for 15 minutes: 50/100

advanced tech: process in mobile lab for 5 seconds: 60/150, process in mobile lab for 15 minutes: 90/150, process in mobile lab for 6 hours: 120/150

Edited December 21, 2013 by thereaverofdarkness

[Atoning Unifex]

There's some good ideas in here. I like the OP's suggestion, plus the idea of a number of tiers depending on what lab, if any, is used. I do think it's important to remember science is more based on specific goals though, and missions would be launched to meet those goals. So for example you are looking for specific life bearing chemicals in the rock, hence your equipment would allow you to do that to 100%, however you could not gain any additional information about the sample. So perhaps having a sample returned to a orbital lab, or KSC could provide bonus research. This does, however, require a specific research mission to have been set, which doesn't currently happen!

Combinin experiments seems a viable alternative to what I suggest above. In answer to the suggestion that combining experiments would make masses of dev work, well this could simply be reduced by not having all combinations viable - not all combinations of scientific devices would produce correlations etc. Hence reducing work and making it more challenging for us to determine useful combinations (maybe!)

I also agree that it is very annoying to have to grind biomes when new science devices become available. There should be a suite of basic devices followed later on by more advanced ones, but in the same research group, so that you get multiple science devices at once. So you DO have to repeat experiments for additional benefit but only once or twice.

[Commander Zoom]

I like and support the OP. That is all.

[Rath]

I've said it before, but I would take it one step further, in a way that makes a hell of a lot of intuitive sense:

Samples really need to be analyzed before you can get any useful information out of them. So right now, being able to take a surface sample and then transmit the data back to Kerbin doesn't make sense. What data?

I would suggest that any sample based experiments (goo, materials bay, surface samples) should not have transmissible data. Either return the samples to the lab at KSC for the science or, and here is where I start aligning with the OP, use the lab module to analyze the samples and THEN transmit the data back. So, you can only transmit sample "data" if you have the lab to get the data out of the sample. Otherwise you have to return the sample. And remove diminishing returns. 100% value first go. Repetition for the sake of it is bad design.

Now, what you could do - and this again aligns with the OP - is make it so that an "unprocessed" sample only returns a tiny amount of data: Just the data you get from the astronauts making an observation. Call this a "sample observation" experiment, and again, allow transmission for 100% of the value of the sample observation. Then use the mechanic in the previous paragraph to do the "deep analysis" (if you have the lab available) to get the really juicy points of of that sample.

Example:

- Jeb gets duna surface sample. Takes it to command module. Looks at it. Transmits his observations. Science value: 20.

- Jeb then EITHER:

1) Returns to kerbin with the sample. KSC performs full analysis, 100 science value, or

2) Takes the sample to a nearby lab module. 100 science value, no transmission loss.

You could *even* do three levels - basic observation, mobile lab analysis, processing at KSC. If you do any of the "more valuable" processing, you should just automatically get the points for the less valuable processing at the same time. So returning any sample to KSC would get you all possible data out of the sample.

Meanwhile, make simple measurements (temperature, gravity, etc) transmissible for 100% value with, again, no repetition or diminishing returns. Because it makes no sense that you would have to return to Kerbin with a temperature reading to get the full "value" of the information.

This removes grinding (no repetition just for the sake of repetition), it makes transmissible data sensible, and it gives the science lab a truly useful purpose.

i think that recovery should still be better than lab then transmit because a five story city block sized lab will be better than a 2.5 meter lab

- - - Updated - - -

I've said it before, but I would take it one step further, in a way that makes a hell of a lot of intuitive sense:

Samples really need to be analyzed before you can get any useful information out of them. So right now, being able to take a surface sample and then transmit the data back to Kerbin doesn't make sense. What data?

I would suggest that any sample based experiments (goo, materials bay, surface samples) should not have transmissible data. Either return the samples to the lab at KSC for the science or, and here is where I start aligning with the OP, use the lab module to analyze the samples and THEN transmit the data back. So, you can only transmit sample "data" if you have the lab to get the data out of the sample. Otherwise you have to return the sample. And remove diminishing returns. 100% value first go. Repetition for the sake of it is bad design.

Now, what you could do - and this again aligns with the OP - is make it so that an "unprocessed" sample only returns a tiny amount of data: Just the data you get from the astronauts making an observation. Call this a "sample observation" experiment, and again, allow transmission for 100% of the value of the sample observation. Then use the mechanic in the previous paragraph to do the "deep analysis" (if you have the lab available) to get the really juicy points of of that sample.

Example:

- Jeb gets duna surface sample. Takes it to command module. Looks at it. Transmits his observations. Science value: 20.

- Jeb then EITHER:

1) Returns to kerbin with the sample. KSC performs full analysis, 100 science value, or

2) Takes the sample to a nearby lab module. 100 science value, no transmission loss.

You could *even* do three levels - basic observation, mobile lab analysis, processing at KSC. If you do any of the "more valuable" processing, you should just automatically get the points for the less valuable processing at the same time. So returning any sample to KSC would get you all possible data out of the sample.

Meanwhile, make simple measurements (temperature, gravity, etc) transmissible for 100% value with, again, no repetition or diminishing returns. Because it makes no sense that you would have to return to Kerbin with a temperature reading to get the full "value" of the information.

This removes grinding (no repetition just for the sake of repetition), it makes transmissible data sensible, and it gives the science lab a truly useful purpose.

i think that recovery should still be better than lab then transmit because a five story city block sized lab will be better than a 2.5 meter lab