Discovery / doing actual science

[tater]

46 minutes ago, Veeltch said:

@monamipierrot your post started off good, but then I saw the part about discovering the sun and stopped reading.

If we want discovery we need telescopes. And not only to discover the bodies around the solar system. I want to track actual sources of strange signals, like pulsars, look at different types of stars, hunt for exoplanets and take sweet pics of galaxies and nebulas with the actual scientific facts and data about them.

I think that all the astronomy should be assumed, this is kerbal space program, not kerbal "Enlightenment Science."

Fog of war hides what is only known by space probes, but all other data is available. We knew orbits, masses, even atmospheric values for many places before probes. All that would be available.

I have no desire to have to "launch" a telescope on kerbin EVA and click to observe some planet, that would be mind-numbing. 

what the game would also need is more planning tools (which it needs anyway). There is he mod that predicts atmospheric flight paths, for example, so you can more accurately plot landing zones. What if something like that was included, but the values it uses for the atmospheres are not the actual values, but the values the player currently has unlocked. Your first Eve probe would use the value determined astronomically from kerbin, which would be wrong. Having transmitted some science from sending the Eve probe, your next probe gets to use the newly determined values. (I used Eve as an example, but this could be a new world we have not seen before with an atmosphere that is hard to nail down (like the RL Venus)).

Regarding the notion that the "shared experience" is important to the community... I don't think that is a thing, frankly. I personally don't care even a little about it.

[monamipierrot]

6 minutes ago, Veeltch said:

@monamipierrot your post started off good, but then I saw the part about discovering the sun and stopped reading.

 

Well Veeltch, the idea of "discovering Kerbol" is just based on dozens of different videogames I played with, wether easy or hard, on PC or on smartpohones, similar or not to KSP type of game.

I'll explain: often games, instead of giving stand-alone tutorials (or boring manuals), just take your hands at the beginning of the game. They teach you how to turn left or right, going forward or back, use a inventory, aiming at a target, spot objects, make actions, talk to character, start building something... They often do it in a "task"/"mission" environment where of course the 1st task/mission is obvious and with difficulty=zero. The purpouse of these tasks/missions is to have the noob player being familiarized with basic controls/actions/options in gameplay. This could be done with just a control/option list, but that is tedious and unsexy. So they almost always give those 1st missions a 2nd parallel purpouse of starting the "story" (or the conflict, or the discovery, or the race, or whatever it is the game about).

Sometimes it looks very silly, because the game is about something serious (e.g. some deadly special op corp trying to save the world from some nuclear terrorists) but the 1st "mission" looks like a joke (e.g. shoot some isolated rookie enemy).  Still, developers choose to put it there anyway.

Thanks to the magical mix of sillyness and seriousness in KSP, developers could present the game to noobs and teach them how to "do science" with the excuse of doing something both obvious but also mandatory. In Facebook KSP groups I often see people who started careers and are desperetely grinding on science. Chatting a few minutes with them reveals they DIDN'T KNOW one can do proper science directly (with reports, science instruments etc.), while they thought there was only the tiny science reward from contracts. Why? Because nobody told them! Nobody told them they could right click, "crew report" or right click "start experiment" and so on... I don't think they were idiots. After all they almost got to the moon without science! It's just that there's literally NOTHING that shows this options to noob players. I thought the Kerbol thing could be both functional for noobs and silly in a perfect Kerbal-style (Kerbals were too much focused in building their KSC and dreaming about flying so they didn't notice there was a Sun just on their head).

The concept (guide noobs to "obvious" milestones in order to learn some gameplay control/mechanics/whatever) could be extended to many other aspects, but here I wanted to focus on the EXPLORATION/DISCOVERY (and "science") thing.

The discovery of other "secrets" of Kerbal universe (e.g. that Kerbin is round, or that the Mun orbits Kerbin and that the latter itself orbits Kerbol!) will be discovered few minutes later, almost mechanically, and thanks to "improved" science techniques (crew reports, instruments, unlocking a new building), to which the game will deliberately guide the player. 

And yes you are right, when we start climbing the tech/science ladder we need telescopes, sensors, probe missions etc. and the thing starts to grow serious: the hand-by-hand mode will not be required any more. But a silly start would be good both for the funny nature of the game and both for training noobs.

If you think this makes things more complex for developers, well, IMHO it is not true: implementing a real "discovery" aspect of the game already needs to completely redesign the game UI: not only the map mode (you can't see some bodies, or you can see them blurry, and/or you are still not sure about their orbits...) but also they may want to tweak the normal view mode (to prevent it being too much "spoilerish") If they managed to do this, then to implement the "discovery of Kerbol" or the "discovery of the roundness of Kerbin" would be a joke in terms of coding. And all of this would be CONSISTENT. Also, it would not affect in any way the seasoned player: he would perform those routine "science" actions anyway, at the very beginning of his career, and in the meantime he could still have a laughter at the sillyness of our little green friends.

If you liked this one, please read the rest of my original comment, as well as the linked old post of mine.

 

P.S. I am glad this thread have been revitalized. We should redirect here other threads which are related to this one.

P.S. 2 If Veeltch or somebody else doesn't like that Kerbals don't know Kerbol (or roundness of Kerbin) while already able to build a Space Center, here's (my) explanation: Kerbals always have lived in underground tunnels where they has been exposed to high levels of Kerbonium (which gives them both their characteristic green color, and their absolute sillyness), and developed a tiny but advanced underground civilization under the region which we now know as the KSC plains. When they accidentally one of them dig a tunnel to the surface after drinking too much Kerbonium spirit, they decided to go further up, but they couldn't because after many efforts they didn't manage to dig air. Thus the idea of flying. Here's when it all started.

 

17 minutes ago, tater said:

Your first Eve probe would use the value determined astronomically from kerbin, which would be wrong. Having transmitted some science from sending the Eve probe, your next probe gets to use the newly determined values.

The Discovery overhaul permits this and a thousand more intriguing details.

Anyway, I firmly believe it is mandatory not to show something till you "accidentally" discover it. You will not look for some new planet if you don't know it exists. You'll just put a telescope in orbit and it will do the job for you. You'll just spend money investigating some fluctuations of a body orbit. You'll just build a new tracking building. You'll just send a probe out there... Discoveries will come, somtimes randomly, and sometimes desperately pursued. In any case, very rewarding, at least if you didn't know ANYTHING at the beginning!

[tater]

The idea of discovering planets except perhaps one out in the kuiper belt is absurd. I think it is outside the scope of KSP to have to "discover" Saturn (I'm going to use real world examples, but assume that any I use are really made-up, randomly added planets). 

The starting point for KSP is roughly the Earth in the 1950s to 1960s. The planets would all be known, except, again, possible worlds that are very far away. What would NOT be known is:

The specifics of planetary atmospheres.

What they actually look like on the surface (no one had a clue what Mars actually looked like until the Mariner probe flyby).

What their moons look like, or indeed most of the moons. Until we sent probes to Jupiter, we had no clue how many moons it had, dozens of them were brand new discoveries---everything about them. This is ripe for "exploration" in KSP.

Large asteroids and dwarf planets. Again, some would have the orbital elements known, and nothing else about them. 

There is no reason at all to ask for the added boredom of a terrestrial telescope that you have to click on, or whatever. It is needless. KSP is about spacecraft. Give the player in such a randomized "discovery" mode exactly what data they would have, and nothing more, but certainly don't give them less than they would know. 

The proposal needs to be: 

Something Squad could actually do.

Something people actually want to play.

Terrestrial (whatever the word should be, kerbestrial?) astronomy should not be a thing, it should be assumed.

 

[Wjolcz]

1 hour ago, blorgon said:

 

How about a new, procedurally generated planetary system every time you start a new Science Mode save. What they could do, however, is make a pre-determined set of seeds the game could use, where each one is permanently tied to the progression of save games (Save Game #1 = Planetary System Seed A; Save Game #2 = Planetary System Seed B; et cetera). That way, there'd still be the shared experience that is so important to the community, but would allow a player to start a new save and go on to start exploring system #2 once they have discovered all they can about the current system. And of course, there'd have to be a dedicated sub-forum for each system for the players currently playing in those systems.

Since, you know, this thread is pretty much comprised of pipe-dreams 

Nah. Random generation is not a very good system for flight sims. At least when it comes to planets. Every time the solar system would have to be proceduraly generated the game would have to make assumptions that, for example, the gas giant near the sun won't ruin the orbits of other planets/moons/potentially destructible asteroids. The game should more or less represent the real solar system (even though it's scaled down), because there's only one place in the universe with life in it and we don't know what planets in which places would influence the life in the habitable zone. It would be better to have the planets in different phase angles every time you start a save. 

We kind of do know how other planets might influence life in the habitable zone, but I hope you get my point.

I really like what @tater said about discovering the actual properties of atmospheres/gravitation of the planets through the usage of probes and actual science. I think that would work very well.

I was also refering to discovering other planets when more are added. Some sort of Saturn analogue and beyond. That way the telescopes would be actually useful when tracking them. The ones within Jool's orbit are close enough to be visible with the nakes eye.

[tater]

For real discovery, there has to be some randomness, though.

I agree that purely random would be bad. I would say have variant systems that have reasonable assumptions in terms of which types of planets are where, but that the planets themselves would change (still handmade, it just selects from a library of choices), and the solar system scaling could change as well (1.2X spacing, 4.3, 6.4, etc), with the planets changing size independent of that (as long as they scale the same or less). Best of both worlds.

[monamipierrot]

13 minutes ago, tater said:

The idea of discovering planets except perhaps one out in the kuiper belt is absurd. I think it is outside the scope of KSP to have to "discover" Saturn (I'm going to use real world examples, but assume that any I use are really made-up, randomly added planets). 

I thought we were discussing at least a bit about the scope of KSP.

 

17 minutes ago, tater said:

The starting point for KSP is roughly the Earth in the 1950s to 1960s. The planets would all be known, except, again, possible worlds that are very far away. What would NOT be known is:

The specifics of planetary atmospheres.

Should I cast a *SIMULATION BIAS ALARM* ?

I thought the entire '50-'60 thing of KSP was pure atmosphere. The hairstyle/look, the parts design, the nerdy cool jazz music. All of it perfectly blended and balanced with the silly look, the sillier technical explications of Kerbals, and the Kerbal Way of Life.

I have been a huge freak fan of flight simulators. I really respect them. And I am not going to say "this is a GAME not a SIMULATOR" cause this sentences doesn't make sense to me. Any videogame "simulates" something. I would say almost any plain GAME is. The dicothomy (if there's one) is between REALISM and... LESS REALISM. I'll stop philosophying right here and go straight to the point: the only thing I want KSP to simulate is the thrill of discovery thruogh technical skilness and risky mission which characterized those decades. They DID discover lot of things. And it must have been really rewarding for them: I wonder how rewarding would have been if they discovered - say - that there were a twin body tidally locked and hiding just behind the Moon. But there was nothing. Instead they found that the Moon had a very thin atmosphere. I wonder how rewarding would have been to stumble upon Pluto with New Horizon without really expecting there were something there. But they already knew. Since decades. No reward for that, sorry.

51 minutes ago, tater said:

There is no reason at all to ask for the added boredom of a terrestrial telescope that you have to click on, or whatever. It is needless. KSP is about spacecraft. Give the player in such a randomized "discovery" mode exactly what data they would have, and nothing more, but certainly don't give them less than they would know. 

"Boredom"? When you perform the 27th rescue in Kerbin orbit to cheaply enlarge your kerbonauts team may be boring. You do stuff to achieve something. Tired of clicking? Make it automatical. Reward is the key word in ALL games. But I would say it is even more key in discover games.

Kerbin is not Earth. It is much smaller. There's a purpose for that: Earth / Solar System would have been too difficoult to explore for standard "physics" models. Ouch, we lost a tiny bit of simulation (or better said realism). Also Kerbals are tough to die. Reason? Things are simpler (and maybe more fun). Another little realism gone. Uh, and pieces magically attack one to the other one. Fuel flows for some magik. [Realism slowly fading away]. And the best: there's a magic eye outside of your spacecraft so you can see yourself and take a selfie. Poor old realism! What did we do to her? Let me tell you this: realism had to be sacrificed for both technical reasons AND gameplay reasons. Very good ones. Still, this (and many other game) HEAVILY rely on reality (or even history, in this case) for inspiration and to model both content and gameplay. Physics is present. A solar system is present. Spacecraft are present. '50s and '60s charm is present. The only thing I miss is DISCOVER (or exploration, to me is the same word).

So if we can tweak reality and history for better gameplay purpose, please tell me why shouldn't us radically introduce discovery in the game and say we have to discover it all, and start giving rewards from the very beginning? If you ever played in freakishly manner some XXXX game such as Civilization you would know exactly what I want to say: I want a fog-of-war map instead of an entire-known-map for the same reason I prefer to slowly come to know the beauty of woman instead of see her completely naked at the 1st glance. Or for the same reason I want to slowly discover who is the killer of a misterious murder in a good thriller movie, instead of just only pursuing him on the streets like just another action movie...

Oh but you want "simulation". And you think it is just "about spacecraft". But you get bored so you want "discovery" and then you think it better and want only a very tiny little bit of it (i.e. details about atmosphere) when you could have it all.

I really don't understand. Expecifically I don't understand why this should harm your KSP experience instead of 10-fold enrichen it.

1 hour ago, tater said:

The proposal needs to be: 

Something Squad could actually do.

Oh, this is a reason, yeah. The only one.

[Mr. Scruffy]

Well, i guess there could be a compromise, with most bodies known, but some to be discovered with certain equipment coming close to them when active. Say Gilly, Pol, and Bop only show up, when you bring some sort of scanner to the planetary system. With ´random´ systems, just the possibilty that there could be something, would make you bring that equipment - and it would be discovery facilitated by flight.

[monamipierrot]

44 minutes ago, tater said:

For real discovery, there has to be some randomness, though.

I agree that purely random would be bad. I would say have variant systems that have reasonable assumptions in terms of which types of planets are where, but that the planets themselves would change (still handmade, it just selects from a library of choices), and the solar system scaling could change as well (1.2X spacing, 4.3, 6.4, etc), with the planets changing size independent of that (as long as they scale the same or less). Best of both worlds.

I perfectly agree. Only thing: handmade bodies are great, but I bet there could be a way to create entirely random generated ones which guarantee both novelty ("wow! A Earthly-type planet with one giant pangea with a giant inner sea in it with a inner continent in it!") AND compatibility with gameplay (along with minimum realism requirements). I mean the generator must know if he can create - say - satellites with vertical steepness, or planets with atmosphere lower than mountain peaks, or other weird things. I believe it is not hard to do.

The reason I want computer-generated bodies are two: 1. we don't rely on human work, so we can have infinite ones - and 2. Sometime randomness may be more interesting than human wildest fantasies.

And if the library was web-based, it could also be subject to user feedback: you may award 5 stars to some novel Kerbal system or to just one of its bodies: the system may try to offer it to other users more often, or even "learn" what players like, and try to replicate and improve it in the future.

48 minutes ago, Veeltch said:

Nah. Random generation is not a very good system for flight sims. At least when it comes to planets. Every time the solar system would have to be proceduraly generated the game would have to make assumptions that, for example, the gas giant near the sun won't ruin the orbits of other planets/moons/potentially destructible asteroids.

If a system can create procedurally generated bodies and systems of quality, it can also "test" them to assure they meet Veeltch's standards for realism and don't ruin his "flight sim" experience.

Again, you shouldn't really worry about that.

I don't really like this auto-merging post thing. I'm not used to it. I don't know if you guys get all my answers. Just in case, recheck my old comments. I'm puzzled.

[blorgon]

4 hours ago, Veeltch said:

Nah. Random generation is not a very good system for flight sims. At least when it comes to planets.

Why do you think that? What is it about planets that you think don't make for good sims? All you need (environmentally speaking) in order to simulate flight are atmospheric (if any) and gravitational conditions, both of which can be written into code.

4 hours ago, Veeltch said:

Every time the solar system would have to be proceduraly generated the game would have to make assumptions that, for example, the gas giant near the sun won't ruin the orbits of other planets/moons/potentially destructible asteroids.

I don't think you quite understood my idea. I'm talking about the kind of procedural generation that No Man's Sky and Astroneer do, which is done one time, based off of a seed with intentionally specified conditions. I'm not talking about KSP procedurally generating a planetary system on the fly every time you start a new game (I should've made that a little more clear, my bad), I'm talking about Squad "spec-ing out" a dozen or so seeds that they could then easily grow planetary systems out of, within which they could make alterations. It's not like it would be completely random; they could set parameters that would limit how many gas giants there would be, and how massive they could get, et cetera. Furthermore, with the exception of "Hot Jupiters", gas giants generally tend to form on the outer reaches of planetary systems. I don't really understand what you mean when you say that the game would have to make assumptions about "the gas giant near the sun". What do you mean by "ruin the orbits"? This is exactly what the real Jupiter does, by the way—it "ruins" the orbits of asteroids in such a way that they generally don't hit Earth and, earlier in the solar system's history, cleared out the inner solar system of planetesimals, making a relatively clean and clear path for Earth so that it could cool, and its environment could develop without constantly having to start over after a catastrophic collision.

4 hours ago, Veeltch said:

The game should more or less represent the real solar system (even though it's scaled down), because there's only one place in the universe with life in it and we don't know what planets in which places would influence the life in the habitable zone.

[emphasis mine]

Well that's a silly thing to proclaim. Not only is it highly probable that there is life elsewhere in the universe (possibly even in our own solar system, i.e., Europa), it is highly likely. So... I guess I don't understand your argument. Also, we're talking about the Kerbolar system, which, you know, has Eve. Can't really be making the argument that KSP should more or less represent the real solar system, since clearly it's only inspired by it. I don't see how you can say that it's not possible to procedurally generate stable planetary systems where life could develop since we, IRL, are very familiar with the conditions under which it developed (the only thing we don't know for certain is where it came from).

I also don't understand what you mean by "we don't know what planets in which places". If you're talking about how planets and their positions influences the possibility for life to develop in the habitable zone, then yes, we do generally know how, as I explained earlier with Jupiter.

Finally, I'm not talking about replacing the Kerbolar system entirely. I'm only talking about one game mode wherein you wouldn't be playing in the Kerbolar system—Science Mode.

Edited April 9, 2016 by blorgon
Planetesimals, not planetoids.

[tater]

1 hour ago, monamipierrot said:

I thought we were discussing at least a bit about the scope of KSP.

The scope of KSP is not going to change. We can all want it to change... doesn't matter, it's not realistic to expect that to change. If a development suggestion is to be taken seriously, it needs to be something that has a non-zero chance of actually happening, and even the less serious changes suggested are unlikely, frankly. Look at their career game additions for proof.

 

1 hour ago, monamipierrot said:

Should I cast a *SIMULATION BIAS ALARM* ?

I thought the entire '50-'60 thing of KSP was pure atmosphere. The hairstyle/look, the parts design, the nerdy cool jazz music. All of it perfectly blended and balanced with the silly look, the sillier technical explications of Kerbals, and the Kerbal Way of Life.

I have been a huge freak fan of flight simulators. I really respect them. And I am not going to say "this is a GAME not a SIMULATOR" cause this sentences doesn't make sense to me. Any videogame "simulates" something. I would say almost any plain GAME is. The dicothomy (if there's one) is between REALISM and... LESS REALISM. I'll stop philosophying right here and go straight to the point: the only thing I want KSP to simulate is the thrill of discovery thruogh technical skilness and risky mission which characterized those decades. They DID discover lot of things. And it must have been really rewarding for them: I wonder how rewarding would have been if they discovered - say - that there were a twin body tidally locked and hiding just behind the Moon. But there was nothing. Instead they found that the Moon had a very thin atmosphere. I wonder how rewarding would have been to stumble upon Pluto with New Horizon without really expecting there were something there. But they already knew. Since decades. No reward for that, sorry.

That's just reality. There could not have been a body behind the moon, because math. You cannot plausibly deny the player information that would have been available in the Age of Sail unless you plan on the game starting out in the Age of Sail.

 

1 hour ago, monamipierrot said:

"Boredom"? When you perform the 27th rescue in Kerbin orbit to cheaply enlarge your kerbonauts team may be boring. You do stuff to achieve something. Tired of clicking? Make it automatical. Reward is the key word in ALL games. But I would say it is even more key in discover games.

If it is automatic, why bother? I am only assuming they know what they should know at the point of their technology. You don't get to aircraft/rockets without having some basic physics and astronomy. Assume it's automatic... it all happened before the game starts.

 

1 hour ago, monamipierrot said:

Kerbin is not Earth. It is much smaller. There's a purpose for that: Earth / Solar System would have been too difficoult to explore for standard "physics" models. Ouch, we lost a tiny bit of simulation (or better said realism). Also Kerbals are tough to die. Reason? Things are simpler (and maybe more fun). Another little realism gone. Uh, and pieces magically attack one to the other one. Fuel flows for some magik. [Realism slowly fading away]. And the best: there's a magic eye outside of your spacecraft so you can see yourself and take a selfie. Poor old realism! What did we do to her? Let me tell you this: realism had to be sacrificed for both technical reasons AND gameplay reasons. Very good ones. Still, this (and many other game) HEAVILY rely on reality (or even history, in this case) for inspiration and to model both content and gameplay. Physics is present. A solar system is present. Spacecraft are present. '50s and '60s charm is present. The only thing I miss is DISCOVER (or exploration, to me is the same word).

Mini-Kerbin I would argue is a mistake. The reason they made it small had to do with error propagation in the code, as I recall. Rounding errors screwing stuff up with large distances. Smaller planets is actually not only unrealistic, but it also harms gameplay. The Mun has one, simple solution as a spacecraft problem in KSP. There is no balancing KOR vs MOR, vs Direct Ascent modes of landing on the Mun. If the worlds were bigger, then you'd actually have a reason to wonder as NASA and the Soviets did about what was the best choice for mode.

That said, it's not going to change.

1 hour ago, monamipierrot said:

So if we can tweak reality and history for better gameplay purpose, please tell me why shouldn't us radically introduce discovery in the game and say we have to discover it all, and start giving rewards from the very beginning? If you ever played in freakishly manner some XXXX game such as Civilization you would know exactly what I want to say: I want a fog-of-war map instead of an entire-known-map for the same reason I prefer to slowly come to know the beauty of woman instead of see her completely naked at the 1st glance. Or for the same reason I want to slowly discover who is the killer of a misterious murder in a good thriller movie, instead of just only pursuing him on the streets like just another action movie...

Oh but you want "simulation". And you think it is just "about spacecraft". But you get bored so you want "discovery" and then you think it better and want only a very tiny little bit of it (i.e. details about atmosphere) when you could have it all.

I really don't understand. Expecifically I don't understand why this should harm your KSP experience instead of 10-fold enrichen it.

Oh, this is a reason, yeah. The only one.

I have said exactly nothing about simulation. I prefer an exploration mode. I just don't think it needs to include stone-age exploration mode. I think the start date for Kerbal SPACE Program is a society of little beings capable of space flight. No civilization reaches that point without the low-level science required to establish the orbits of the planets in their solar system, and indeed even some slightly more detailed analysis.

1 hour ago, monamipierrot said:

I perfectly agree. Only thing: handmade bodies are great, but I bet there could be a way to create entirely random generated ones which guarantee both novelty ("wow! A Earthly-type planet with one giant pangea with a giant inner sea in it with a inner continent in it!") AND compatibility with gameplay (along with minimum realism requirements). I mean the generator must know if he can create - say - satellites with vertical steepness, or planets with atmosphere lower than mountain peaks, or other weird things. I believe it is not hard to do.

The reason I want computer-generated bodies are two: 1. we don't rely on human work, so we can have infinite ones - and 2. Sometime randomness may be more interesting than human wildest fantasies.

And if the library was web-based, it could also be subject to user feedback: you may award 5 stars to some novel Kerbal system or to just one of its bodies: the system may try to offer it to other users more often, or even "learn" what players like, and try to replicate and improve it in the future.

If a system can create procedurally generated bodies and systems of quality, it can also "test" them to assure they meet Veeltch's standards for realism and don't ruin his "flight sim" experience.

I'd not disagree if it can be shown that the generated worlds would be reasonable. Right now the only really good world is the Mun, IMO. I think they could have a library, and simply have people submit bodies for addition to the library. the game can go online now and again and grab new worlds to add. If you rescale, and have enough choices to pick from, it will seem pretty novel each new career.

[Mr. Scruffy]

monami: I agree with Tater, that manually scanning the sky with telescopes doesnt sound like fun. If anything this should be part of the management game, where you would have to focus an observatory on a body (that must be known beforehand) to discover its moon(s) and orbital properties automatically over time and maybe upgrade that telescope and man it. That would not add a whole lot to gameplay, though, imho. If time is to be implemented in a meaningful way, there would be enough stuff to take care of (i am thinking of research and construction times, life support etc).

Tater: I think when i was a child, Jupiter had like, i dunno 23 (this number totally coming from my behind) moons? Now it´s 62 (iirc).

Edited April 9, 2016 by Mr. Scruffy

[gasper002]

Love it!

[Pthigrivi]

I could support this thread becoming the new default 'science improvements' catchall in the frequently suggested list. The current one is pretty skimpy.  Ive seen a lot of good ideas over the last couple of years and it makes sense to consolidate things. The science system is one of the oldest career elements and one of the longest to go without major revisions. Its not surprising it that it feels lacking at the moment, but I disagree with a lot of people that the basic concept of collecting science points and then spending them on the tech tree is fundamentally flawed. It's simple to understand and flexible enough to let players go where they want to go and earn rewards without being overly proscriptive about it. And while I agree generally that discovery as an idea is great and really important to a player's sense of exploration, so far I've not been convinced that locking planets or making them blurry really adds much to the game. Part of the problem is the worlds themselves are pretty much empty, so making them blurry doesn't actually hide anything. I tend to agree with Tater that 'discovery' should have more to do with revealing information about each planet that's relevant to playing the game and interacting with them. If experiments were to unlock things like biome maps and landing site prediction they would feel much more rewarding. The other problem with science I think is with the experiments themselves. The grindiness people are feeling isn't as much due to the layout or balance of the tech tree as it is due to the fact that collecting science isn't actually very fun. Its essentially going to a place and repetitively right-clicking through a series of functionally identical parts. It doesn't feel like science, but thats not because its unrealistic, its because it doesn't feel like doing something special to collect information thats useful.

Generally:

- Each experiment should be unique, and should require the player to do something special besides clicking on a part.

- In addition to science points, each experiment should provide information that's valuable to the player and useful to playing the game outside of the tech tree. 

- As much as possible science should feed into the core fun of the game--building and flying planes and rockets--rather than distract from it. 


I've posted this before, so apologies if people have read it already, but these were the changes I suggested last year that I think might not only take the clickiness and grind out of science, but could also make conducting experiments more fun, challenging, and rewarding. In addition to providing the player with unique useful information, each experiment requires the player to in some small way fly differently. That way conducting experiments is directly involved with how missions are planned and executed. I've also suggested automating most experiments so that players can focus on the flying rather than being constantly redirected by tedious right-clicking. More experiments could be added, but until each existing experiment is given some individual attention adding more really only exacerbates the clickiness problem. 

 

Crew Reports: Gathered automatically by crewed capsules and stored for each new biome the craft enters, serving as a running log of the mission.

EVA Reports: Gathered automatically on EVA for each new biome a kerbal enters and stored when they return to the vessel. Scientists gather more valuable EVA reports, and their value can be further upgraded as they gain in levels. Kerbals cannot discern between biomes above the surface.

Surface Samples: Can be gathered on EVA by any crew member, though higher level scientists gather more valuable samples. Samples cannot be transmitted unless analyzed in a mobile processing lab, but give much more science than other sources. When a sample is analyzed either in a science lab or on Kerbin it will indicate precise ore concentrations and will become available for loading into Materials Bays (more below).

Goo Canister: First experiment available in the tech tree and acts as an introduction to gathering science. When it enters a biome with uncollected science it flashes blue for a few moments and then auto-exposes. Its one-time use unless there's a scientist on board in which case it auto-collects, stores, and then auto-resets. It draws no power. By default its set to activated, but it can be deactivated and reactivated via right click if a player wishes to hold out for more a more valuable exposure. It cannot distinguish between biomes above the surface.

Thermometer: Next experiment on the tech tree, flashes blue and then takes a reading and stores automatically when entering a new biome. Its activated by default, but draws 1.5 e/m while activated and can be deactivated to save power. It cannot distinguish between biomes above the lower atmosphere or high above a body. Vessels with a Thermometer on board show overheat bars in flight, though even without the parts will still glow red.

Barometer: Arrives shortly after in the Tech Tree and flashes blue when new science is available. The barometer is activated by default when in the atmosphere, but can be deactivated to save power via right-click. Unlike the thermometer, the barometer logs science based on the vertical swath of atmosphere it passes through while continuously running. This means it gathers a lot of data on ascent and descent, but sitting on the ground it gathers next to nothing. Later in the game, a body for which the player has completed a barometric scan will show trajectory, landing site, and aerobreak predictions factoring drag for higher level pilots.

Materials Bay: Materials Bays should be able to be loaded with materials, i.e. samples, and replace the current Mobile Lab magic science generator. When a surface or atmospheric sample is recovered, it goes into a bank of available samples. Upon launch, the materials Bay can be loaded with up to 5 of these samples, and when activated (0.5 e/s) it generates and stores science based on the value of the sample multiplied by the value of the exposure location. This means that a sample from the launchpad exposed at KSC will be worth very little, but a sample from Ike exposed on Duna will be worth a great deal. Samples generate science for 30 days and then become spent. Materials Bays can be reloaded by an adequately staffed Mobile Processing Lab, but only with samples banked at the time of the Lab's launch and with samples processed by that lab. This means bringing a lab to another body will be useful for processing and gathering science from that body over time, but samples cant be magically transported across the Kerbol System. Indeed routing samples from surface to lab to materials bays (and from planet to planet even) to maximize their value would be the real challenge.

Atmospheric analyzer: Essentially works as an atmospheric sample collector. Its deactivated by default, and once activated (1 e/s) the vessel must maintain roughly the same speed and altitude for 10 seconds to collect a viable sample. Like surface samples they may not be transmitted unless analyzed by a mobile processing lab. If atmospheric xenon collection were enabled perhaps precise concentration levels could be determined from these samples.

Surface Sample Collector: This part would replace the surface scanner, and ought really to be a small arm and drill that drops down when activated. It aught to come very late in the tech tree, but in principle enable collection of surface samples by probes. Like other surface samples these would be available for loading into Materials bays and would show ore concentrations when analyzed.

Survey Scanner: Works much as it does now, once placed in a polar orbit it generates a rough ore concentration map which can then be transmitted for additional science.

Narrow Band Scanner: Works much as it does now, but could also provide accurate distance to surface information or even a topographic overlay.

Gravoli detector: This part works 2 ways, its activated by default and draws .5 e/s, and like the thermometer automatically collects and stores data for each new biome it passes into. If however it is placed in a polar orbit it gathers all biome information for that body at that altitude, and if it is attached to a vessel that also has a survey scanner it can generate an overlay map of all biomes on that body. If a mission planner were to be added including flight time and delta-v estimates, completing a gravoli scan might unlock that body in the planner, encouraging players to send a probe first if they wanted to optimize their crewed mission.

Seismometer: This part is redesigned as an impactor experiment. Once on the surface and activated (2 e/s) a blue circle appears on the body in map mode indicating the scanning radius. The higher the level scientist on board the larger the radius. If while activated another object is slammed into the surface a red impact radius is shown, whose radius is determined by the mass and speed upon impact (I can foresee some really fun asteroid antics here ) The Seismometer generates science based on the area of overlap between the scanning and impact radii, meaning more precise collisions and bigger booms make for more science. Additionally, ore concentrations can be seen with detail within this scanned area making for better landing site decisions for mining operations.

Mobile Processing Lab: With material studies now moved over to the Materials Bays, the lab can be used primarily for processing and reloading samples. In addition new contracts could provide special samples which could either be pre-loaded or delivered to existing labs for processing and/or loading into materials bays. Unlike other data sources processing samples makes makes them transmittable, with level 1-5 scientists converting samples 20%, 40%, 60%, 80% and 100% respectively. Where most capsules can store just 3 samples, Mobile Processing Labs could store 25 samples at a time. 

Transmitting data: As almost all data is automatically logged and stored, all that would be left would be transmission. For simplicity's sake, I feel like clicking any pod or antenna ought to bring up a single data log indicating all stored data in one screen, the value of each piece of data, and giving the option to transmit. I'll be interested to see the changes Roverdude has made to the antenna system in the future, but in my mind the most straight forward solution is that all data except samples should be in principle 100% transmittable, and all losses could be controlled by quality of arrays. If surface samples could not be transmitted without processing and were worth a great deal (as they should be) then returning these samples would make 2 way trips worthwhile without the over-complication and grind of multiple transmissions. 

Biome Multipliers: Another thing that could give dimension to each planet would be to vary the science and experience reward for different biomes by 10-25%. If for instance the Polar Lowlands on the Mun were worth 20% more than landing on any old point on the Midlands players might spend more time thinking carefully about where they land rather than plopping down anywhere they haven't before. This could become especially important if easter eggs were expanded to become real, interesting surface features like mineral formations and volcanoes in small but valuable biomes, and would reward players for precise landings and using rovers to do surface exploration.

... Oh and just for reference a procedurally generated solar system has been ruled out by Squad and is on the WNTS list. Could make a cool mod if someone was feeling ambitious though.

Edited April 9, 2016 by Pthigrivi

[tater]

34 minutes ago, Mr. Scruffy said:

Tater: I think when i was a child, Jupiter had like, i dunno 23 (this number totally coming from my behind) moons? Now it´s 62 (iirc).

There were 12 known Jovian moons by 1951. 2 were discovered in 1974 and 1975. 3 were discovered by the Voyager probes. Most were later discovered by large terrestrial telescopes, interestingly enough. Had the probes been designed to look for new moons, they'd have likely found more.

In the context of KSP, I suppose you could either have more discovered (like asteroids appearing) later in the game, or you could add a space telescope part that would then discover some smaller moons---though a probe might have already found them at that point.

Note that strictly speaking, this is "unrealistic," as kerbin based scopes would have seen them, but it leaves it as something to discover on orbit for players.

Yeah, unlocking the biomes would be another useful experiment as it tells you where to land. The perfect landing site would be at a meeting point of biomes (or close enough to walk/drive/lander-hop between them. Showing biomes on the map makes that better.

BTW, "blurring" need not happen.

The extant map system places the starting view at some distance away. We could determine where the map camera is, say it's 1 million meters for example.

The camera then has a zoomed in view, that hits a "floor" at maybe 10,000 meters? (again, I'm, making this up, it might actually be higher than that).

My suggestion would be to have a simple mechanic whereby flying missions (ideally with a new camera part, but all crew vehicles can be assumed to have a (film) camera that does this once they return to kerbin) reduces the altitude "floor" for the map view. In a perfect world, it would do this based on location as well. So with a thoroughly mapped world, you might be able to zoom the map all the way in. The starting map camera position could depend on what is known from astronomy, so for Duna that start altitude might be what the planet looks like from a ship that has just hit the SoI for Duna (a reddish ball with ice caps) Other worlds might be a colored circle. Totally doable with the current map mode, it's just a camera position.

[pandaman]

I like the idea of a discovery game mechanic, though whether a random method would be practical I'm not in a position to answer that, and I don't think a fully random solar system is the right approach for stock.  Certainly the game should start with the knowledge of the planets etc maybe with inaccuracies in some details that are only corrected by visiting the planets and/or with the right experiments in the right places.

The concept of 'unseen' small moons and planetary features etc only becoming visible when you get close enough is a good one IMO.  And that wouldn't need to affect the existing solar system beyond adding new things to find.  Maybe some random element could be applied to generating small moons etc along similar lines to asteroids.

[tater]

Just checked and for Duna the map will not zoom closer than ~260,000m. It starts out just past Ike, so over 2 million. My suggestion would be for the start point to be substantially farther out, but on the "buff" side, you'd be able to zoom as close as you'd been with a craft (and some telescopic camera parts might allow zooming closer than craft altitude).

Then sending some atmospheric instrument might allow for unlocking that atmospheric prediction ability, etc. Some stuff is points, some stuff is actually useful.

Just now, pandaman said:

I like the idea of a discovery game mechanic, though whether a random method would be practical I'm not in a position to answer that, and I don't think a fully random solar system is the right approach for stock.  Certainly the game should start with the knowledge of the planets etc maybe with inaccuracies in some details that are only corrected by visiting the planets and/or with the right experiments in the right places.

The concept of 'unseen' small moons and planetary features etc only becoming visible when you get close enough is a good one IMO.  And that wouldn't need to affect the existing solar system beyond adding new things to find.  Maybe some random element could be applied to generating small moons etc along similar lines to asteroids.

I think that the default career/discovery might well be "stock," but honestly, for replay, it needs to be random. Once you know what altitude to set periapsis for a Duna landing, you'd have to forget it---that or you need a game mechanic that only allows certain settings to happen if certain science is done, which I think would be a mistake (say you cannot adjust the parachute deploy values unless you do some atmospheric science at Duna first) as it could get grindy.

[Mr. Scruffy]

Hmmm... there could be two (or even three) optional paths about this:

- Probes with a special sort of scanner would discover all moons in a planetary system they´d enter, given enough time and energy. They´d not need to be focused on something, once they reach a new planetary soi.

- Ground based telescope could do the same, if focues on that planet, but it would take much, much longer and require a scientist to work on it. Upgrades to the facility would be super-expensive but raise the cap of its exploration speed which would have to filled with scientiests again (assuming a time-relevant management game). E.g. basic version could only hold, say, 5 scientiest. Fully manned like that, it would discover 1 moon of the planet it is focused on every ~80 kerbal days - it would take just one scientiest a kerbal year. 2nd tier telescope could hold, say, 10 scientists, shorting the minimum time to 40 days (this all could be modified with the distance of the planet focused).

- Orbital telescopes would be somewhere in the middle. Moderate in price for their effeciency and automatic. It would come in the form of heavy modules you install on an orbital base, which would encourage orbital base-building. Those modules could be identical (call them ´lenses´ or whatever) and each one would work like a fully staffed tier of the ground based telecope (so are capped to its maximum possible tier per base). It needs to be focused on a planet, in order to work. Only one per base.

You could run multiple probes, one ground based telescope and multiple orbital ones all at once, all focused/used on one or multiple planets, if you wanted to. Multiple telescopes/probes focused/working on the same planet would add to each other´s effort. So you could choose one way to do it, or another, or do multiple in parallel.

The time it takes to find something should not be fixed, the numbers i gave above should be mean-time-to-happen, not absolute values.

[tater]

I think ground-based needs to simply be assumed. Since tech tree nodes are the only time progression in the game (that's an entirely different issue), when a certain node set is unlocked, you get more ground-based data.

Knowing that the moons exist doesn't alter "discovery" at all, because from telescopes, they are simply DOTS. So we know there is a retrograde dot in a certain orbit around Jool, and it's pretty small. We still have to go there to see what it looks like. Given the cool moons of Jupiter and Saturn, there is room for interesting discoveries.

[Wjolcz]

Ok, so I see my post got quoted a bunch of times and since I can't edit them because editing doesn't work for me on mobile (hooray for the new forum! yay!) I will just say few things:

There's only one place in the universe we know about that has life.

Comparing procedural generation for KSP and No Man's Sky is a bit silly. It only makes sense if you can go and visit more systems and not only one. If KSP had more solar systems then I have nothing against procedural generation outside of the main one.

Playing procedural KSP would be like staying in the same room but with missplaced furniture. Playing No Man's Sky is like going out and to a whole new room full of completely different things. Except you can come back if you don't like the paint.

Exploring the same place, but with misplaced elements would get boring really fast. If you make changes then either make them stay or expand.

[Pthigrivi]

The point is moot though anyway, as a procedural solar system is not planned. At any rate Id much, much rather see the worlds we have given more dimension and life than see endless sets of new, random, empty planets. 

[tater]

That's why I'd suggest a library of hand-made worlds to chose from.

More dimension to the existing worlds would be awesome, and that would eb valuable alone, but any mode of play where "discovery" or "exploration" is truly a thing cannot have to use the same worlds over and over.

A simple example... Minecraft. I play that rarely, and sometimes new seeds are just "cool," and you feel compelled to walk around just to see what;s around the corner.

[Pthigrivi]

How many, do you think? Like if they're going to invest time constructing interesting new planets I'd rather just get a GP2 with moons, rather than design a huge number of planets I'd only see if I opened multiple saves. I totally agree with you on that "around the corner" feeling, but Id rather get that feeling in-flight or while tooling around with a rover. Right now there isn't much of that because the worlds themselves are basically empty. If we had a basic set of surface features spread across the planets, geysers, volcanoes, pools of primordial goo, etc, and some way to use instruments to find them the way we now prospect for resources the sense of discovery wouldn't just be 'oh cool another blank planet to land on and return from', but an experience that extended in-flight and caused players to change their plans and improvise. Thats why I think making experiments truly interactive rather than jut a thing to click is so important. That's where discovery should come in, challenging players not just to go there and come back, but to engage with the surface while they're there to yield the best rewards.

Edited April 18, 2016 by Pthigrivi

[Mr. Scruffy]

Again, there could be a viable middleground between truely random or even just procedural and handmade: If that library, you talk about tater, would be split in catergories resembling properties, then due to (plausible) combination of them combined with no more procedural processing than for the mun right now added ontop of it, for detail-features only and stretching...

Eve in the place of Duna, with Laythe´s atmophere, or Ike in the orbit of Jool with the mun´s surface and Duna´s atmo. Add enough variations of these properties and mix/match...

Edited April 9, 2016 by Mr. Scruffy

[monamipierrot]

17 minutes ago, Mr. Scruffy said:

Well, i guess there could be a compromise, with most bodies known, but some to be discovered with certain equipment coming close to them when active. Say Gilly, Pol, and Bop only show up, when you bring some sort of scanner to the planetary system. With ´random´ systems, just the possibilty that there could be something, would make you bring that equipment - and it would be discovery facilitated by flight.

Yep. That was what I thought at the beginning: You don't know distant or tiny bodies till you don't have enough tech or come closer enough.

Then I realized that was just a false dicothomy. If the initial discovery rate is very fast, why bother showing *everything* at the beginning? Let the player find any body or detail by himself. This will be his "tutorial" fase about the Solar System and the planet he lives in. Seasoned player just skip it.

There's really no reason not to do keep it all unknown from the beginning - unless you are very impatient and want to know at least how the basics of the Solar System is BEFORE starting. But I bet none of you is that impatient, if he once managed to play more than 1 hour in KSP!

Another reason would be: if I don't like this Solar System / planet / whatever, I don't want to waste time playing with it and I'm going to try a new one. Again, this would be a false problem, the real one being that the procedural system didn't give you ENOUGH CONTROL in the creation process standards and variables. In other words: the creation system MUST create beautiful solar systems and bodies, and - also - understand your tastes and needs to create one that YOU like, so you can RELY on it: "Dear Player, you can't even imagine what I prepared for you. I'm thrilled to wait for you to find out a specific body that... but let's stop spoilering, and let's play!". I really bet this would be easily achievable.

 

Instead, why don't give some random/fuzzy information or even better just plain rumor? E.g.

• In the astronaut lounge, Jimmy Kerman swear he saw some shiny whity thing far South while hiking in the Mountains. However he really like lemon icecream so he may have just dream about it. [a spot or a flag in a locally darkened map is shown]
• Johnny Kerman, a tracking station analyst, reported he saw a brown body orbiting Jool, but it was too drunk with Kerbonium spirit and accidentally threw the pictures in the toilet [inclination and size of orbit of body given].
• A secret new part being tested in the VAB suffered (another) sudden Kraken attack, and was launched at 7.5x speed of light towards Eve. Moments before disappearing in the infinite vastness of the Void, it sent back a picture [picture shown] of what Scientist would swear is some liquid surface, but since this is not consistent with current knowledge of Eve's surface, it has been regarded as just another problem with Kerbonium spirit abuse.
• A Kayan prophecy believes a new star will appear in the sky at the beginning of Next Year, more or less HERE [coordinates given]. The prophecy fails to tell us when the world will come to an end, so no End of The World Parties are allowed.
• A debris of Kapollo 10 mission accidentally hit the surface of the Mun. Observing its dust pennacle, scientist were able to determine with better accuracy percentage of Ore in midland biome, but the real thrill was that the pennacle had the exact shape of a "!".
• Some anonymous Kerbal hacker/physician youngster group managed to manipulate a Kerbonium atom and discovered it simply can't exist. But it does. The silly conclusion is that we are all part of a simulated computer game. According to this bizarre theory, there should be a - sorry, I try not to laugh - dwarf planet made just for fun beyond Duna [coordinates given].
• And so on....

 

[monamipierrot]

3 hours ago, Pthigrivi said:

I could support this thread becoming the new default 'science improvements' catchall in the frequently suggested list.

 

I've posted this before, so apologies if people have read it already,  

you kidding?

Thanks for your great ideas!

I still find the Material bay thing a little bit grinding but your ideas are awesome and still simple.

Of course, I still am a fan of having fog of war.

3 hours ago, Mr. Scruffy said:

 

- Orbital telescopes would be somewhere in the middle. Moderate in price for their effeciency and automatic. It would come in the form of heavy modules you install on an orbital base, which would encourage orbital base-building. Those modules could be identical (call them ´lenses´ or whatever) and each one would work like a fully staffed tier of the ground based telecope (so are capped to its maximum possible tier per base). It needs to be focused on a planet, in order to work. Only one per base.

You could run multiple probes, one ground based telescope and multiple orbital ones all at once, all focused/used on one or multiple planets, if you wanted to. Multiple telescopes/probes focused/working on the same planet would add to each other´s effort. So you could choose one way to do it, or another, or do multiple in parallel.

The time it takes to find something should not be fixed, the numbers i gave above should be mean-time-to-happen, not absolute values.

I find just perfect these three points. I like the last one too. You put a telescope to orbit. Providing some basic coditions (power, visibility...), you can point it to a body, and then it just starts scanning and have some % chances to discover something in a given time. I don't see the grind or the boredom some other users see in this.