Let's talk about how Science can be improved

[Pthigrivi]

So we were on a bit of a tangent in the LS thread last week and began talking about the broader sections of the game that don't feel fully ironed out yet. There were some, Regex and Klgraham included who felt career mode was disjointed and grindy enough to prevent them from using it. While I disagreed with their broader assessment, I do have to agree that at the moment the fun process of building rockets and gaining rewards and earning new parts is missing something. I've played a good deal in stock career and found that at least for the amount of time I can reasonably set aside for KSP (probably 6-8h/wk) I generally have to set reward settings rather generously to progress without feeling like Im bogging down. While there are many aspects of the game that are obviously incomplete, experience and building tiers for instance, I've grown to feel the most established but lagging component to the game is the Science system. Squad made a nice pass on the Tech tree, which was great and helped, but I think didn't address the fundamental issue. I don't personally have a problem with the general concept of experiments producing points which then can be spent on a tech tree. It's simple enough to understand and broad enough to accommodate most play-styles. One problem, I think, is that the actual act of planning and conducting experiments isn't actually very exciting. Another deeper problem is that for many players the Tech Tree is the objective of the game, so once its complete the reward structure ends and they feel like the game is over.

There are a few reasons for this, which I thought we could consider as we offer ideas on how to improve the system.

1) KSP is a rather unique game, wherein not only craft but play-styles are near infinite. This means that any science system has to be flexible enough to let people play the way they want to play and still set clear rules and incentives for progression. Additionally players start careers with varying levels of experience, so any successful science system needs to not only serve as a tool for introducing new parts and concepts to first time players, but remain fun and challenging for veterans.

2) KSP isn't a game with a "win" condition. Its a creative building game with costs and incentives, much like Simcity. This means that the scaling of the tech-tree and the meaning of science points once its complete is more difficult to gauge than in most games. When stretched thin the game becomes a grind, and when condensed things become too easy and players max out quickly and no longer know what to do. Strategies were implemented in part to deal with this, allowing players to exchange mountains of excess science for funds for instance, but once a player has a maxed out tech tree and a million bazzillion funds they are essentially playing sandbox. That seems an odd end-game reward.

3) And this is important, KSP is at its heart an indie game that bats way above its weight. Its incredible what they've achieved, but Squad is a small company and development time is desperately precious. What I think would is most helpful is sussing out the precise issues causing problems and finding succinct corrections that would address them. I laud people's ambitious ideas, but let's try to think realistically about what changes really could be made to improve the system and the game.

I have my own ideas, but I'm interested to hear what others think.

[Alshain]

I think we need another game mode that has funds without science. My game doesn't begin until the tech tree is finished. Sometimes I just 'cheat' to get the science.

I just want sandbox with funds and contracts.

[Superfluous J]

I think we need another game mode that has funds without science. My game doesn't begin until the tech tree is finished. Sometimes I just 'cheat' to get the science.

I just want sandbox with funds and contracts.

Agreed. I actually want checkboxes for stuff.

[ ] Funds

[ ] Science

[ ] Kerbal Experience

[ ] ... uh... other stuff too.

All checked would be full career. All unchecked would be sandbox. Various in-between alternatives would match things like Science Mode and "Funds Mode" that you're talking about, and anything else.

[sumghai]

As I've often said in the past, I personally think that rather than turning surface samples into generic Science points, the (repeated) collection of (large numbers) of samples could be used to gradually uncover more useful information pertaining to a planet's composition:

As may have been hinted by Squad in certain places, a resource system featuring sample collection / analysis / return as well as mining / processing / storage has been suggested for an as-yet undetermined future release of KSP. This would be an additional justification / goal for players to send spacecraft to other planets/moons (in addition to flying Jeb and his flag there, of course ) - the ability to do real science.

While the actual implementation is still quite far off in the horizon, I figured I'd start a thread anyway detailing some of my ideas and musings, which could be potentially applicable to both the stock game and mods such as Kethane, Extraplanetary Space Centers and tek Industries Science.

What Squad has WIP'ed so far...

(Note that these may not necessarily represent the what Squad will ultimately implement - I've simply relinked them from the last known reddit thread for your viewing convenience)

• Spacecraft-mounted scanners (Long range directional, Short range omnidirectional)
  
• Drills (for acquiring Solids and Liquids)
  
• Storage containers
  
• Various refiners (1, 2) / chemical processors (Small, Medium and Large)
  
• A tentative list of resources, (prototype) resources map for each planet and a (now-outdated) resources flow chart describing how said resources would be turned into spaceflight consumables such as fuel and life support - it has been speculated by the reddit thread author that Squad is going to simplify the processors to be able to handle a variety of resources (rather than overspecializing by having one dedicated processor per resource).
  

My own musings...

Initial Surveying / Sampling

• At the very beginning of a new savegame, players would immediately be able to view Kerbin's resource map for its surface, oceans and atmosphere, without needing to do any additional prospecting / science (I presume Kerbals would already know their home planet's composition well enough). Conversely, the resource map for other planets would initially be blank.
  
• The scanner array(s) would be used to provide very rough surveys of other planets and moons from orbit - they would probably detect the presence of certain resources and locations where they are generally found, but little about their actual quantity / proportion / purity.
  
• Ground-based robotic or crewed missions would be required to obtain better data through sample collection / statistical analysis missions
  
  • Units of rock, liquid and gas samples would be collected by Kerbals on EVA or dedicated robotic arms
    
    • Rock samples provide data for surface makeup
      
    • Liquid samples provide data for ocean makeup (or additionally, underground crude oil-like substances such as the namesake of the Kethane mod)
      
    • Gas samples provide data for atmospheric makeup
      

    [*]Each sample obtained would contain differing proportions of selected substances found on that planet/moon

    • The rationale for this is that one does not simply have samples purely made up of whatever
      
    • Sample #005 may contain 45% Metaxium / 10% Blutonium / 30% Alium / 14% Kerbon / 1% UNKNOWN SUBSTANCE 06
      
    • Sample #021 may contain 67% Metaxium / 30% Zanotite / 3% UNKNOWN SUBSTANCE 02
      

    [*]The more samples one gathers, the more accurate the final resource map for that planet would be

    [*]"Known" resources are those that are already also found on Kerbin - it is enough to simply identify them and the proportion present in each sample

    [*]"Unknown" or "Rare" resources are those that are not found on Kerbin - a sample return mission is required so that Kerbal scientists back home can properly study it

    • Once sample(s) containing unknown substance(s) is/are successfully returned, new resource names are automatically added to the resource map list
      
    • e.g. UNKNOWN SUBSTANCE 06 will automatically be renamed to Jebedynium, for instance
      

Prospecting / Mining / Processing / Storing Resources

• A combination of orbital scanning and ground sampling would be used to determine which sites give the best purity for whatever resource the player is interested in mining
  
• Resources would be mined via various parts
  
  • Certain part types would limit where resources could be obtained
    
    • Drills acquire solid resources from the surface
      
    • Drills with pump attachments acquire liquids from below the surface (again, think Kethane)
      
    • Pumps acquire resources from oceans
      
    • Intakes acquire resources from the atmosphere
      

    [*]Parts of the same type would be able to mine all types resources found in their specific region, and differ only in extraction rate and efficiency

    • i.e. Pumps PX-100 and PX-200 are able to acquire the same resources, but the PX-200 works twice as fast
      

  [*]Refiners would be able to purify solids into any mineral-based resource desired

  [*]Chemical processors would be able to purify any liquids or gases desired

  [*]Bulk and fluid storage tanks are able to store any proportion / type of refined solids and liquids respectively (e.g. any combo of minerals can go into one bulk storage tank)

Using Resources

• Liquids and gases can be used to fuel rockets or supply life support systems (similar to Kethane)
  
• Minerals or metals can be used to fabricate basic spacecraft components such as fuselages and fuel tanks (like the "ore" used in Extraplanetary Space Centers)
  
• Rare resources would most likely be semi-conductor like minerals used for making more hi-tech components such as SAS, command pod instruments
  
  • May also reduce the cost / weight & increase the performance of basic components
    

  [*]Refined resources can be flown back to Kerbin - if resources are involved in career mode, this could be a source of financial revenue / means of replenishing part stocks

...And how to potentially implement them (from a code perspective)

Initial Surveying / Sampling

• The resources map proportions for each planet and moon are actually predetermined and hard-coded into KSP itself.
  
• A proposed resources.cfg for each savegame would contain:
  
  • An initial seed for the random deposits location/quantity generation (sharing seeds results in identical deposit locations)
    
  • Resource names and proportions available on each planet / moon (initially blank - populated through gameplay)
    
  • Deposit locations / purity (initially blank - populated through gameplay)
    

  [*]When players start with simple orbital scans of various planets and moons, the persistence file is updated to store the names of the resources found on said planet/moon, along with a quantity of 0%

  • This would be displayed in-game as "present" i.e. Kerbon - present
    

  [*]When players collect physical rock/liquid/gas samples:

  • The presence and proportions of resources are NOT stored on a per-sample basis (massive data overhead!)
    
  • Rather, Kerbals on EVA and robot samplers would simply have a incremented "SampleCount" up to their "MaxSampleCapacity"
    

  [*]When these samples are offloaded into analysis parts:

  • The code simply loops through the the number of samples deposited
    
  • For each sample/loop iteration, dummy variables are used to temporary hold resource names and quantities
    
  • The composition of each sample is randomly generated, but strongly determined by the hard-coded proportions for each planet
    
    • e.g. Assume Duna is hard coded to have 75% Metaxium on its surface
      
    • Therefore any given rock sample is more likely to mostly contain Metaxium
      
    • The actual quantity of Metaxium returned as outpost from any particular sample may actually be anywhere between, say 25% to 50%
      

    [*]The results for each sample are averaged out and written back to resources.cfg map, as the "actual" resources map proportion for each planet/moon

    • i.e. only the final running average for the whole planet/moon is stored
      

    [*]The more samples one collects, the closer the savegame's data tends towards the hard-coded ideal values

  [*]Example:

  • At the start of the savegame, nothing is known about the Mun's composition
    
  • Internally in KSP, the Mun may have hardcoded resource proportions of 10% A, 20% B, 30% C and 40% D
    
  • When a player scans the Mun, resources A, B, C and D will be detected but their quantities will be unknown
    
    • i.e. A - present, B - present, C - present, D - present
      

    [*]When a player analyses their first rock sample, the random number generator may assign said sample as having 70% A, 20% B and 10% D

    • Mun Resources map is now updated to A - 70%, B - 20%, C - present, D - 10%
      
    • Obviously this is grossly inaccurate compared to the overall hardcoded figure, but remember that we've only taken one sample
      

    [*]By rock sample #025, the figures might get averaged out to:

    • A - 12%, B - 18%, C - 33%, D - 37%
      
    • Basically, the more samples we take, the closer we tend towards the "ideal" hardcoded proportions
      

Prospecting / Mining / Processing / Storing Resources

• Improved orbital scans would result in deposit locations/quantities/purities being written to the resources.cfg savefile (much like kethane's own kethane.cfg)
  
• Deposit quantities would limit how much resources can be extracted before the deposit runs out
  
• Deposit purities would determine the extraction rate for any extractor part
  
• Processing efficiencies would further influence the amount of purified resource actually obtained
  
• Storage is already implemented in-game
  

[Pthigrivi]

Agreed. I actually want checkboxes for stuff.

[ ] Funds

[ ] Science

[ ] Kerbal Experience

[ ] ... uh... other stuff too.

All checked would be full career. All unchecked would be sandbox. Various in-between alternatives would match things like Science Mode and "Funds Mode" that you're talking about, and anything else.

Right, but even the ask contains a veiled sort of indictment of the system as it stands. The question is what specifically would you change to make science fun and challenging for you?

[Superfluous J]

Right, but even the ask contains a veiled sort of indictment of the system as it stands. The question is what specifically would you change to make science fun and challenging for you?

...I don't know.

I don't want to have to play mini-games. I don't want to have to do something over and over again. I don't even like right clicking parts to run experiments and collect results.

Honestly for me the current system, if all the science gathering happened automatically, is fine. For me KSP isn't a game about simulating science experiments. It's about going to places where the experiments need to be run.

[Kaos]

I also would like very much to have the possibility to fine tune which concepts are used in a game, like Alshain and 5thHorseman suggested; I have that suggestion even linked in my signature ( http://forum.kerbalspaceprogram.com/threads/124711-Unify-game-modes-for-greater-configurability ).

For how to improve science itself, I would suggest some generic science parts, which do not directly produce science, but can be used in contracts to build up (and extend, via the new upcoming system to interact with existing stations) a science station.

Example:

* Contract1: Bring a gravioli-twister (20t) into orbit of Dres and operate it with 3 scientists of which at least one was started from Kerbin after accepting this contract; gains xx science points

* Contract2 can only show up, after Contract1 was finished: Bring a helium-cooling unit to station yy (the one with the gravioli-twister) and let it work for 100 days (needs 1t helium per day) to enhance the experiments. We need 3 engineers there to connect the parts

These science parts may (or may not) only show up, when a contract needs them. Perhaps only one part is generated per contract, hence the experiment fails if one does not manage to start it properly. Some science support parts (as the helium-cooling unit) may be regular parts. Other parts may be unique science parts. Perhaps there are some generic science looking parts models for some different sizes and weights and the parts are generated with this fixed model, but contract-specific, partly auto generated name.

[Joonatan1998]

I think they should make most science experiments give the same results in all biomes, but give a lot more science for each test than they do now. Just have 1 experiment biome specific on low orbit, (gravioli?) 1 flying low (the analyzer? Or maybe thermometer?) and 1 landed(acellometer?), to keep polar satellites/planes/rovers usefull.

Would make science far less grindy, and the multiplier would be far better at changing difficulty.

-Make the lab into something else, i just doesn't fit the other game mechanics.

-Maybe make one of the science parts require you to return the whole part, instead of just taking the science.

-All experiments should have either 100% transmission or 100% recovery.

-crew/EVA reports should give less/no science, but instead give large amounts of reputation. (At least for first landing on each object.)

Edited October 29, 2015 by Joonatan1998

[jwbrase]

A couple things I'd like to see are science experiments with secondary benefits and a few of the more common real life experiment types.

For example, one big thing that real life probes do is imaging. So you could have camera parts, and for a given planet you'd get so much science per square kilometer imaged at a given resolution (with resolution being a function of the imaging abilities of the camera part and the distance to the imaging target). But in addition to science, youcould also get other benefits from imaging: Pretty pictures are good PR, so imaging could contribute to reputation. Also, as things currently stand, you get full-featured maps of each planet from the get-go. With imaging science the player would start out with a high detail map of Kerbin, a mid-high detail map of the near side of the Mun, a mid detail map of Minmus, and low detail maps of Duna and Eve, with everything else being completely unmapped. You would need to map a planet to locate biomes and potential landing sites, and surface contracts for a body would only be given for mapped regions.

[Pthigrivi]

Yeah this is pretty close to where Ive come down: the fundamental lag isn't with science points or the tech tree, but with the act of conducting science. Ironically, I think its that the conditions for successfully conducting an experiment are too easy to meet, meaning they must be repeated many, many times to keep pace with funds and rep. We have a number of parts that work in varying numbers of biomes but are essentially indistinguishable as devices, so you end up strapping on as many as you can and then endlessly checking them as you fly. This is what causes the "clicky" feeling. Adding more devices without addressing this central problem in many ways makes matters worse.

There's a fine balance though between 5thHorsemans instinct to make science automatic Joonatan's instinct to further restrict a part's use. If parts are too selective opportunities for dovetailing science collection with fulfilling existing contracts are lost. If science is everywhere and automatic there's no real need to think about it and no real challenge.

My thinking has been that each experiment really aught to require something different from the mission. Minigames could become very tedious; the challenge of executing the experiment should really be in-game. For instance, instead of a point reading the barometer could be a part that can be started and stays active, requiring a small amount of charge while running. The more of the air column it passes through while active the more it pays out. This means it could be run on a launch or during a landing, but could also be attached to drop probes and parachuted into different biomes. The seismometer could be run as part of an impactor experiment: once on the surface it detects impacts within a certain radius that would be visible from orbit in the UI. If another object is slammed into or near the circle the seismometer pays science depending upon how much the sensing radius overlaps with the impact radius. Experiments like these, while more difficult to execute, should pay out much higher than they currently do, requiring less repetition. Very often I hear people comment that the worlds themselves feel empty, that you can go to any old location on a body and once you get there its kind of "now what?" If the "what" was to execute something more challenging perhaps time spent on-surface would feel more rewarding.

[Joonatan1998]

Yeah this is pretty close to where Ive come down: the fundamental lag isn't with science points or the tech tree, but with the act of conducting science. Ironically, I think its that the conditions for successfully conducting an experiment are too easy to meet, meaning they must be repeated many, many times to keep pace with funds and rep. We have a number of parts that work in varying numbers of biomes but are essentially indistinguishable as devices, so you end up strapping on as many as you can and then endlessly checking them as you fly. This is what causes the "clicky" feeling. Adding more devices without addressing this central problem in many ways makes matters worse.

There's a fine balance though between 5thHorsemans instinct to make science automatic Joonatan's instinct to further restrict a part's use. If parts are too selective opportunities for dovetailing science collection with fulfilling existing contracts are lost. If science is everywhere and automatic there's no real need to think about it and no real challenge.

The main reason I think most experiments should not care about biomes is to reduce the "clickyness" by reducing the amount clicks for each situation the craft can be in.

I don't think it would really make the science parts significantly less useful if they gave a lot of science once for each planet instead of a bit of science for each biome, it would just make each experiment you do more valuable.

There would still be a reason to take all the experiments to each planet, just wouldn't need to do as much click-grinding with Planes/rovers/polar satellites.

Making science automatic would also reduce clicking, and it could work well for some experiments, like the gravioli detector.

But I think we need more diversity with the experiments, and don't think they all should be automatic.

I think the first landing on an object being much more valuable than visiting different biomes on the same object would make the game better.

Edited October 29, 2015 by Joonatan1998

[Torgo]

I would love to have the option to have science seem more 'sciency', where the knowledge gained from research impacts your abilities or efficiency later in the game.

An analog would be the mining system as is. You can send a drill and tank to the Mun, land it and mine ore. If you do the orbital and more localized scans, you can find the optimal spots to mine.

As noted in a post upthread, imaging could mean that you start with really low-res images being available (as seen through telescope) and applied to bodies in the tracking station. As you send mapping probes, you can develop a higher resolution map and imagery.

I'd love to see more practical experiments, like a laser measuring system requiring the setup of a reflector on the Mun that has to be set up within certain tolerances so the beam can hit it and reflect correctly.

A lot of this is already served by mods, and maybe that's the better way to address it.

[sumghai]

Any thoughts on the "take more samples, get more accurate planetary composition pie chart" idea I proposed?

[Waxing_Kibbous]

Science is certainly an area that has a ton of potential still. I like the idea of discovery/obtaining data about planets via science as opposed to just having the planetary info at hand. This could open up the possibility of a mode where the planetary parameters (and/or the entire Kerbin system) are randomized so that one would need to find out about them before exploring, adding a ton of replay value. A space telescope could give the visuals, the various other units would let you know about the gravity/atmosphere, etc. I think that part of the game is still pretty young though and look forward to the next versions.

As a side note I cheat and use the For Science mod/start with a bit extra because it is a bit grindy for my taste. Somewhere I stumbled across a mod that has the options for starting a game as stated earlier, like infinite money, all science unlocked, etc. but I can't seem to find it again.

[Shark4558]

Any thoughts on the "take more samples, get more accurate planetary composition pie chart" idea I proposed?

Well, I wish the implementation of the analysis of samples, because I find it odd that a surface scanning apparatus "magically" shows the exact concentrations of minerals without previous samples or something, I would leave in the last place the scanner surface. It would be something like:

the sample surface reveals the chemical components of the place.

the orbital scan shows abnormalities of the normal composition of the place (eg iron concentrations in the lunar soil)

and the surface scanner displays the actual value of iron in a specific place.

and as for the system of science in general, it could be improved

KSP has great potential, to my knowledge, KSP is the only game that lets you explore planets the "modern" way, and if it were possible to analyze a planet in many ways, plus doing real science,not bring apparatus and click "log data" science.it would be an awesome game !.

Edited October 29, 2015 by Shark4558
the translator mistranslated ... oops!

[regex]

Honestly for me the current system, if all the science gathering happened automatically, is fine. For me KSP isn't a game about simulating science experiments. It's about going to places where the experiments need to be run.

This. This right here. KSP is about going places. It's about launching rockets, building neat things, exploring. Right-clicking for points is terrible so I play with mods like ForScience and ShipManifest so I don't have to worry about doing it. Science should just be automatic, like the world's first contracts. You go places with the instruments and you get to unlock new parts. Simple.

[sal_vager]

I'd still like the science parts to spawn minigames of some kind when you use them myself.

Maybe have a large version with buttons and dials so you can tune it in, or be able to grab and shake them, the better you do the more science you get at that location.

And each should reveal a hidden layer, just as you can see craters on the Mun, you should be able to find hot areas with the thermometer, shapes in the gravioli fields, areas of higher tectonic activity.

That'd be fun.

[Pthigrivi]

Yeah Sumghai I read it and its cool. Its a bit dense though, Im worried it doesn't really alleviate the grind many players are experiencing. One of the big drawbacks at the moment is that science demands many repetitious, unrisky, urewarding activities. I do agree though that in some way surface samples should be linked to prospecting. What I would rather see is Roverdude's surface scanner redesigned as a surface sample collector for probes, and all samples indicate ore levels once analyzed.

Which is another general ask Ive seen from many people: science data aught to be useful in the game beyond just unlocking parts in the tech tree. I think this is important not just as means of basic information discovery, but also in extending the need for science after the the tech tree is complete. For instance completing a barometric scan of a planets air column could unlock trajectory and aerobreak predictions, and gravoli detectors could unlock topographic overlays and accurate distance to surface information.

In most instances its not that things are overcomplicated, its that they're complicated in the wrong ways. Keeping track of what sensor will work at what altitude above what biome and clicking it seems pretty fussy. To be honest the whole transmission value mechanic feels unintuitive and overwrought as well. To simplify things I feel like most experiments aught really just to have a 100% transmission value. The only thing that really needs to be physically returned are samples. To compensate for how much more difficult it is to return a sample the science value should just be much higher.

- - - Updated - - -

This. This right here. KSP is about going places. It's about launching rockets, building neat things, exploring. Right-clicking for points is terrible so I play with mods like ForScience and ShipManifest so I don't have to worry about doing it. Science should just be automatic, like the world's first contracts. You go places with the instruments and you get to unlock new parts. Simple.

I hear you, but in what way is a game in which all science is automatic distinguishable from a game that has no science at all? You're essentially describing a game in which you take off, fly to arbitrary location on planet x, don't bother with eva, lift off, and go home. Thats pretty much KSP .18.

[regex]

You're essentially describing a game in which you take off, fly to arbitrary location on planet x, don't bother with eva, lift off, and go home. Thats pretty much KSP .18.

Because I can right-click and then click five times instead of once on EVA, that's what makes going someplace more fun than just going somewhere? Or would we maybe want some contrived and forced mini-games to get our points? Even making science more "sciencey" pretty much entails putting a craft what you built in a certain place. KSP is about the journey, from designing to building to flying.

[Superfluous J]

I hear you, but in what way is a game in which all science is automatic distinguishable from a game that has no science at all? You're essentially describing a game in which you take off, fly to arbitrary location on planet x, don't bother with eva, lift off, and go home. Thats pretty much KSP .18.

Because you don't see the benefits the science brought you until you go to that place and do it.

In $FPS_OF_CHOICE you don't have to right click each bullet and select "Load into clip" and then right click each clip and select "Reload gun with clip." You also don't have to do a mini game where you hold your gun up, do some sort of click or shake or whatnot to unload the current clip and then load the new one. No, you hit "R" or - in some games - the gun just reloads. In what way then is reloading important? It puts a constraint on the player and gives that player something extra to think about. Do I reload now and risk someone coming out of that door, or do I kick down the door and risk running out of bullets in the next firefight? The important thing isn't that you accurately mimic reloading. It's that the player must make a choice and do the FUN part of the game with interesting challenges and side effects.

Likewise, the important part about getting a surface sample on Laythe is not right clicking your Kerbal. It's getting that frickin' returnable vessel to Laythe.

[Pthigrivi]

Oh don't get me wrong, Im in total agreement that clicking through each experiment part is not the best set up. I also agree that minigames aren't probably the best solution either. However there is a difference between hitting R to reload and a game auto-firing for you. At some point there needs to be player involvement beyond 'put-all-the-science-parts-on-every-vessel.' You should have to think about what experiments are on board and what you need to do to make them work. The question Im posing is what is that balance that keeps things fun and challenging without becoming a thoughtless checklist?

Edited October 30, 2015 by Pthigrivi

[Superfluous J]

Actually when 1.0.5 comes out I'm thinking of trying my hand at a Contract Configurator config that will set up a bunch of auto-complete contracts that just give you science points for reaching certain milestones with science gear on board, and coming back if that's a possibility (I think it is). Note: This will be the 2nd config I've started and the first is in limbo because I got buried in minutiae...

I agree that there's a line between "mini game to reload" and "the game shoots for you" but I think we disagree on where that line is, and where (comparatively) it is in the science mini-game vs just-play-sandbox spectrum. I personally think that, by choosing to bring those science doodads and a scientist, I've fulfilled my role and Bob should get on with what I'm paying him for, once I get him to wherever I'm sending him.

I guess for you, doing science is like shooting the gun, while for me it's like reloading the bullets.

[Pthigrivi]

And I actually don't blame you for being turned off from it, given the way experiments now work. I guess my instinct is to make experiments require slightly more from the player, thats all. There needs to be some risk/reward structure there. They needn't be minigames so much as in-engine specialized maneuvers, the way docking, precision landing, and polar injections are.

[regex]

There needs to be some risk/reward structure there.

huhwut? How does that pan out for new players? How does that work for repeated play-throughs?

They needn't be minigames so much as in-engine specialized maneuvers, the way docking, precision landing, and polar injections are.

You mean like flying over and landing in biomes?

E: Think about how science is done IRL. You put instruments on a craft, put the craft on top of a launch vehicle, and then you send it places where it uses the instruments. It really isn't any different in KSP except that you have to right-click excessively and the results take far less time to process. Even Cassini flying through Enceladus' plume is just like flying over a biome and recording the observations (and if we had plumes in KSP that still wouldn't enhance the science bit, it'd just be adding to the go places and explore bit, only you might miss the experience because you'd be frantically trying to click at the right moment).

Edited October 30, 2015 by regex

[Joonatan1998]

E: Think about how science is done IRL. You put instruments on a craft, put the craft on top of a launch vehicle, and then you send it places where it uses the instruments. It really isn't any different in KSP except that you have to right-click excessively and the results take far less time to process. Even Cassini flying through Enceladus' plume is just like flying over a biome and recording the observations (and if we had plumes in KSP that still wouldn't enhance the science bit, it'd just be adding to the go places and explore bit, only you might miss the experience because you'd be frantically trying to click at the right moment).

There was that one comet probe with the explosive warhead.

I think that would be a specialized action that isn't just going somewhere.

Edited October 30, 2015 by Joonatan1998