Realism in KSP

[juanml82]

So you want Isp to stay wrong? You want reentry to remain a placeholder? The atmosphere to be a placeholder?

The main issues:

1. fix Isp definition.

2. When the admitted placeholder atmosphere is fixed, should it be fixed with something the same as it is now, less realistic than the one now, or more realistic. Devs apparently mean to change it. Realism people want it to be better (more realistic) than the current atmosphere... the anti crowd wants it worse? The same? (tongue firmly in cheek)

3. Reentry to matter (the game descriptions clearly suggests it should matter, the landers say they cannot reenter, explicitly). Devs apparently want this done, too. So the realism people want it to be more realistic than zero damage now, and the anti-realism want what, exactly? Less realistic than the current "no effect?" Maybe reentering can still have no effect, but you get a "buff" for your next launch? (same tongue position )

So of the big three issues, one is a plain mistake that should be fixed, and 2 are things that are implicitly or explicitly things the devs mean to change. In that latter case, should the changes be more, or less relistic than the placeholders? If your answer is "more" then you are in fact for more realism.

If you go back to the first post I've made in this thread, it starts with what do we talk about when we talk about realism. And if you check the posts I've been writing today, they are an answer to the idea that KSP should be accurate enough to pretty much replace text books at universities.

[m4v]

m4v: It'd be nice if it were 0.2%. It usually isn't, and it's hard to ignore when there are giant advice threads on the forum that say "Don't use nose cones! They make drag worse!"

I just checked and in my rockets the dV loses is always between 1% - 2%, so yeah, not 0.2% but still hardly something to cry over. The advice you quote is just a blind statement that says "anything that adds mass adds dV loses", and while true nosecones weight next to nothing so the usefulness of that advice is very low, something more useful would be "always add 20% more of the dV you think you will need".

Running short of dV because of nosecones means your dV budget was very tight in the first place.

It's also pretty counterintuitive that wide, flat noseconessless rockets perform better than long, spindly streamlined ones. And I for one thing it's bad if KSP teaches us wrong, counterintuitive things. Especially when it's sold as a learning tool.

EDIT: ninja'd by Franklin.

KSP would be hopelessly useless at teaching aerodynamics, why the heck are you assuming that KSP is used for that? is obviously not the case.

[juanml82]

If you can't understand, as a teaching tool for aerodynamics, that nosecones not being essential and arguably unnecessary, is a game design problem, then I don't know what to tell you.

If you want a videogame to teach actual aerodynamics, you're probably better off finding old copies of MS Flight Simulator. And books, lots of them.

It's funny because your post really shows the problem Nathan just talked about.

Ippo, if the Shuttle took 70 minutes to reenter, then any spaceplane made in a realistic KSP would need to take roughly the same time to reenter.

[tntristan12]

And if you check the posts I've been writing today, they are an answer to the idea that KSP should be accurate enough to pretty much replace text books at universities.

Can we avoid this kind of hyperbolic rhetoric? Literally nobody is saying that it should...

[Ippo]

... the idea that KSP should be accurate enough to pretty much replace text books at universities.

... which is something only you seemed to notice in the thread, and I am sure nobody has ever thought about except for you.

[Franklin]

KSP would be hopelessly useless at teaching aerodynamics, why the heck are you assuming that KSP is used for that? is obviously not the case.

Honestly I think you and Nathan are on the same side, Nathan's arguing that the game should be able to be used as an aero teaching aid, and Squad has expressed that it will want to re-evaluate the current aero model down the pipe. So you may be running up the down escalator here if you think KSP shouldn't try to incorporate a teachable aero model.

[Ippo]

Ippo, if the Shuttle took 70 minutes to reenter, then any spaceplane made in a realistic KSP would need to take roughly the same time to reenter.

And if KSP was a better teaching tool, you'd know that not all reentries need to be lifting reentries like the shuttle's.

Also, yes,

, but nobody wants RSS to be stock (not even me or regex) so I really don't understand why you keep talking about it.

[tntristan12]

Ippo, if the Shuttle took 70 minutes to reenter, then any spaceplane made in a realistic KSP would need to take roughly the same time to reenter.

You're confusing realistic with identical. Nobody is saying that KSP should be a mirror image of reality, or even a second version of Orbiter. That's why Rowsdower changed the title, and we should really be backing away from this kind of sensationalism. Realistic physics principles can and should be implemented without sacrificing the playability of the game. Hardware as depicted in the game should.work as expected, not according to some arbitrary set of rules imposed by placeholder physics. That does not, I repeat does NOT mean that ascent and re-entry would or should take the same amounts of time as real-world counterparts. Can we STOP using this as an argument? Please?

[juanml82]

I disagree entirely. There are certain things that you can learn better from experience, which is why entry level physics and astronomy classes have demonstrations, and/or labs. The non-intuitive nature of orbital rendezvous, for example is great to see for yourself (even if in a game). You might not learn history from the games you mention, but if they were done properly, you might actually learn something about why they used the tactics they used (Empire: Total War, for example could have done this, or virtually any age of sail game (all of which fail to model sailing outcomes accurately). Another example is ww2 flight sims. I was always a ww2 aircraft buff, but I honestly "didn't get it" intuitively until I had played good ww2 flight sims, particularly vs other people. Now when I read first hand accounts, I see them as I read, and actually understand (in an internalized way) exactly what they are talking about.

So I think a game like this is a fun tool. I'd have used this with kids at the campus observatory back in the day in a heartbeat (as long as I didn't have to unteach them things that the game got wrong).

Trust me, learning is simple without books.

It won't take anywhere near 70mins to re-enter. Use some mods and find out.

If you can't understand, as a teaching tool for aerodynamics, that nosecones not being essential and arguably unnecessary, is a game design problem, then I don't know what to tell you.

Well, there's also the rationale "you really shouldn't be using planets with insane densities when you're teaching kids about space"--Kerbal.EDU should have the real system at least as an option.

Also, I think a reasonable case can be made for something like 6.4x Kerbin (or maybe only 3x? who knows). Basically, juuust large enough that parts can have realistic sizes and masses and retain the same payload fraction they do now. If that happens, then we can have reasonable Isp, reasonable dry mass fractions, etc., but keep the same "easy sandbox" feel.

That said, given the sense of the community (although one might question just how many have *tried* a smaller variant of RSS, but eh) I don't think that the second will be happening, and I won't spend time and effort arguing for it.

EDIT:

Yeah, if all these "TOTAL REALISM!!!" folks would actually *name* one person who has said they *do* want total realism in stock KSP, that would be great. Heck, even the thread title was edited to get away from that strawman.

Why? when is planet density even a variable when launching something or in space? In education the important thing is to transmit concepts and ideas, not the magnitude of numbers. There's a lot of stuff in school that is taught in a simplified way or outright wrong (how wings generate lift is one that comes to mind) for the sake of it.

And you are ok with that? Seriously: why should we just surrender to mediocrity? "Other people are wrong: let's be wrong too!".

There is a difference between "simplified game mechanic for the sake of fun" and "contradicts everything we know about the universe", especially in a game about space exploration.

Well...

Can we avoid this kind of hyperbolic rhetoric? Literally nobody is saying that it should...

... which is something only you seemed to notice in the thread, and I am sure nobody has ever thought about except for you.

Check the bolded parts

[juanml82]

You're confusing realistic with identical. Nobody is saying that KSP should be a mirror image of reality, or even a second version of Orbiter. That's why Rowsdower changed the title, and we should really be backing away from this kind of sensationalism. Realistic physics principles can and should be implemented without sacrificing the playability of the game. Hardware as depicted in the game should.work as expected, not according to some arbitrary set of rules imposed by placeholder physics. That does not, I repeat does NOT mean that ascent and re-entry would or should take the same amounts of time as real-world counterparts. Can we STOP using this as an argument? Please?

If you add realistic physics, including realistic planetary densities, aerodynamics so realistic they can be used as a teaching tool, a realistic atmosphere that leads to realistic reentries, then you'll end up with realistic results: ascent and re-entry would have to take a similar time as in the real world.

[Franklin]

My post in no way suggested KSP could replace texts, I don't know how you could've even shoehorned that reasoning in there.

But I like posts like that because it gives me a glimpse into how people skim material and take what they want the writers to've meant.

[Ippo]

Check the bolded parts

Yep, sure. Totally what we meant there, buddy. I especially liked how you also bolded my "are you ok with that" which is really a long stretch.

You know what? I'm so done with this thread. You point out there are obvious errors in KSP's physics and you get replies telling you to go play orbiter. I won't tell you where to go because I'd be permabanned.

Over and out.

[Franklin]

Easy now. I'd hate to see this thread devolve into actual fighting. Clearly this is a big enough issue to warrant 40 pages, I'd hate to see it locked & sunk just because some can't be reasonable.

[tntristan12]

Just because something can be used as a teaching tool does not mean it is a picture perfect example. Many dynamics and control problems I study are simplified and/or caricatured to make the underlying fundamentals more important than the surface details, but if the fundamentals are wrong then the surface details are meaningless. What people are lobbying the devs for are more accurate fundamental physics, NOT a surface level representation of reality.

[The Jedi Master]

This might be relevant. That's the only thing I can reasonably add to this discussion.

[ferram4]

The only person arguing for a planet with a not insane density said this, which you conveniently ignored the last part of:

"you really shouldn't be using planets with insane densities when you're teaching kids about space"--Kerbal.EDU should have the real system at least as an option.

No, the version of the game for teaching people about real space shouldn't have the realistic solar system as even an option.

Everyone else hasn't even argued for such a thing. Once you account for that, realistic aerodynamics don't result in realistic reentries because the orbital velocities are lower. Then again, realistic aerodynamics here just means, "streamlining actually accomplishes something rather than being detrimental."

You are too focused on the scale issue, like pretty much every other person arguing against realism. It's frankly disheartening, since it means that rather than arguing about how realist we want things we end up on tangents about 70 minute reentries (for only one specific type of vehicle, even) that will never be in the stock game because (get this) no one wants them in the stock game!

[Guest]

This might be relevant. That's the only thing I can reasonably add to this discussion.

We've already gone over that, but thanks for posting.

How so? Most of the suggestions are based around how realism aspects would make standard intuition more useful in the game, which should make it more accessible to new players. A lot of the current unrealistic stuff just makes the game more confusing (do you remember when you learned nose cones were useless? Does that seem logical? Did it make you more confused about how the game worked?), not accessible.

It's not. None of the things that have been discussed increase the level of complexity of play within KSP (save adding an Engineer-style information panel maybe, which a lot of people want). They do, however, fix behind-the-scenes inaccuracies. Even FAR can be used without the data analysis tools (I do it all the time); look at NEAR, for instance, which is basically just a rewritten, more forgiving FAR without the "confusing" tools. NEAR is a good example of what a stock aero overhaul could look like. It's accessible and not intimidating in the slightest.

E: ferram4 puts it even better than I did. Confusing the player's possible pre-existing knowledge is just ... backwards.

We already had a discussion on the topic of complexity. Regex managed to ninja me, and I fully agree with what he said just above me, but here's one of my responses to that, probably the most relevant one:

For those people claiming that the game is too difficult for a new players: Keep in mind that majority of learning curve comes from some illogical choices made by the developers and lack of proper information given to the player.

For me learning the game was a breeze once I found a wiki that explains oddities this game has. Like massless parts' date=' lack of n-body physics, nose cones that are nothing more than a crippling parts, lack of any in-game comparison of the components (especially important for engines), lack of in-game Delta V screen or any other basic information (eg. GUI never tells you what's the aerodynamic pressure (so you know if you ascend too fast or too slow) and G-force gauge is so tiny and badly scaled that you can't get anything useful out of it).

New players really struggle to get anywhere because of all the oddities that KSP has (eg. that counting on a gravity turn is an idiotic thing to do - instead you should ascend till ~10km and then make 45 degree turn - game NEVER bothers to tell you that). In comparison getting a basic understanding of how orbital mechanic work with current navigation nodes is a breeze! (I did try KSP over a year ago - got discouraged by random guesswork after getting into the orbit - when I played it with the nav nodes - I instantly decided to purchase the game for myself - again an example that explaining stuff to the player makes an enormous difference)

Same with docking - docking wouldn't be even nearly as much of a problem as it is now if we'd have a proper docking camera with proper data displays (eg. velocity on XYZ axis in comparison to the target) - for me it's still MORE difficult to dock when I use "set target as" option on a docking port than doing it manually by observing the craft and moving the camera around - I automatically switch my focus the the gui which gives you some info, but there's just so many gaps between what I can seen and the info I need for successful docking that I usually end up either crashing or spinning off course.

KSP does relatively little to explain anything - hoping that players will "figure it out" (which ends with: read the forum / wiki / watch youtube or in most of the cases: RAGE QUIT) - and at the same time it does A LOT to add crapton of it's own oddities that are NOWHERE to be explained in the game itself.[/quote']

[tntristan12]

This might be relevant. That's the only thing I can reasonably add to this discussion.

*sigh* We are not talking about complexity. We are talking about accuracy. With the exception of life support none of the big points being discussed here even expand the GUI. In fact, the case has already been made that a better aerodynamic model makes the game simpler since players won't gave to memorize silly little tidbits like nosecones hurting rocket performance or engines thrusting at a constant rate despite the ambient pressure.

[The Jedi Master]

We've already gone over that, but thanks for posting.

[sincerity] I can't tell if this is sarcastic or not. [/sincerity]

[Guest]

[sincerity] I can't tell if this is sarcastic or not. [/sincerity]

Those replies were in the last few pages of the thread. Generally, when you come into a thread and post something like "This might be relevant. That's the only thing I can reasonably add to this discussion.", it's polite to actually figure out what you are adding to the discussion and whether your addition is actually relevant. As tntristan12 correctly points out, the recent discussion topics (aside from the sniping we've all been doing, anyway) have nothing to do with adding complexity to the game.

[NathanKell]

So, there are a quite amaznig number of misconceptions here. (Maybe because KSP teaches you misocnceptions? Nawwww.....)

Let's go through some of them. And note that you should not construe this as endorsing rescaling the system for stock KSP since apparently I have to add that disclaimer. I'm just talking about misconceptions with how Real-KSP might play.

1. Levelord: Unless you are going out of your way to replicate a launcher designed fro GTO shots (like Atlas V or Delta IV) an ascent simply does not take that long unless you *literally* burn straight up and then circularize at apogee. You can, in fact, make orbit *easily* in six minutes, you just need to be somewhat inefficient (i.e. have a high-thrust upper stage). If you play it like stock KSP and don't care about G forces, then you could do it faster still. I encourage you to actually try this stuff rather than making assumptons or parroting what (equally unfounded) critics have to say.

2. ferram already answered the point about aerodynamic stress. I encourage you to try taking off in a

Learjet at the same speed you do in KSP, and pitching up as fast. Actually I don't, since you'll die. That said, since the consensus even among the die-hard realismers is that stock KSP shouldn't have that level of aero stress, it's kinda a moot point. Although I will say it's kinda funny that often the very people who proclaim how fun, how ~kerbal~ explosions are, somehow don't like it when their planes do.

3. juanml82: Yes! It's very true! An Orbiter's (aka "Space Shuttle"'s) reentry takes 70 minutes. Good thing no one has to make that exact craft, or fly it that exact way. Ballistic reentries can be very fast, and even lifting reentries don't take very long. Note, again, that these are different things, and there are an infinite variety of lifting entries. The Shuttle flew a particular track due to all sorts of reasons (that I'm happy to get into, but is OT for this thread), and there's no reason you couldn't fly a lifting reentry that's not much (if any) slower than a six minute ascent. Once again, this problem would be solved if, y'know, you actually tried it (as you assume all of us have tried the version of KSP *you're* promoting...). And if KSP had a reasonably real reentry model and aerodynamics model (hint: not uber-real, not perfect, just *reasonable*) you would know this distinction, and not think that all real life reentries take 70 minutes.

4. m4v: well, it's very convenient that a game can't teach you something unless it sets out to! Oh wait, the follow points (above) should...pretty much put paid to that idea. KSP teaches you lots of things. Some are good (orbital mechanics), some are bad (wrong aerodynamics, wrong ideas about reentry) and some are indifferent. You don't get to pick what KSP teaches people; you don't get to pick what people learn or don't learn from KSP. What people learn or don't learn from KSP is dependent on (1) them and (2) KSP.

ALSO: What tntristan12 said.

Edited September 6, 2014 by NathanKell
typos, woo

[Levelord]

This is not a problem of unintuitive aerodynamics, but a problem of scale. Did you take off at anything above 100 m/s? Yes? Then you were going faster than every single real life plane ever on a takeoff roll. Go and watch planes take off at an airport and the fastest you will find is ~80 m/s, with many taking off at slower speeds.

Didn't FAR nerf the jet engines? This goes against my previous flight sim experience where you are supposed to take off on full throttle. FAR really shouldn't be allowing jet engines to push you beyond 100m/s on ground level.

That's funny, it never took me fifteen minutes to get to orbit using RSS... Am I doing it wrong?

Either your game isn't as 'realistic' as you claim it to be with the ships you use, aerodynamics, physics, time warp, G-forces... or you happen to be better at piloting than a real life astronaut. Take your pick.

Did you know Scott Manley scaled down the harshness of DRE because he believed that DRE was not set up correctly for use with FAR? DRE under FAR actually makes reentries easier, apparently. He even made several spaceplane reentries before scaling it back, all without the Ablative Shielding resource that DRE provides.

That sounds more like a mod conflict than actual 'realism' where both aerodynamic forces apply. But that's besides the point on whether DRE should be added to KSP without re-balancing it first and all this proves is that it's incompatible with FAR.

That's not the only reason it should be in the game, that was just a side benefit. How about illustrating the difficulties of space flight? How about adding additional challenge? How about adding it as a difficulty option?

Your last sentence said the magic words. Difficulty option, which is what I'm supporting.

Can you link it? Never mind, I'll do it myself, and quote what I wrote:

Quote Originally Posted by regex View Post

Realistic fuels and such would be awesome, but they're not really needed (besides, gives us something to mod). OTOH, the nuclear engine really only needs to use Liquid Fuel.

Saying if they are needed or not was not the context we were talking about, we were talking about the context of saying if it was the multiple types of fuel or... you know what, nevermind.

[The Jedi Master]

Those replies were in the last few pages of the thread. Generally, when you come into a thread and post something like "This might be relevant. That's the only thing I can reasonably add to this discussion.", it's polite to actually figure out what you are adding to the discussion and whether your addition is actually relevant. As tntristan12 correctly points out, the recent discussion topics (aside from the sniping we've all been doing, anyway) have nothing to do with adding complexity to the game.

Understood. Sorry, it's a long thread. No excuse, I guess.

[Guest]

Either your game isn't as 'realistic' as you claim it to be with the ships you use, aerodynamics, physics, time warp, G-forces... or you happen to be better at piloting than a real life astronaut. Take your pick.

Nathankell already answered.

That sounds more like a mod conflict than actual 'realism' where both aerodynamic forces apply. But that's besides the point on whether DRE should be added to KSP without re-balancing it first and all this proves is that it's incompatible with FAR.

The point being is that it's not nearly as hard as you make it out to be and that people expect it. I was really disappointed when I learned that reentry heat meant literally nothing in KSP back when I started playing.

Your last sentence said the magic words. Difficulty option, which is what I'm supporting.

Sure, SQUAD's adding a selector for it next version, no reason not to use it. I don't think anyone would argue with that anyway.

Saying if they are needed or not was not the context we were talking about, we were talking about the context of saying if it was the multiple types of fuel or... you know what, nevermind.

Exactly.

Understood. Sorry, it's a long thread. No excuse, I guess.

It's cool, sorry if I've come off harsh but that's been the general flow of this thread.

[ferram4]

Didn't FAR nerf the jet engines? This goes against my previous flight sim experience where you are supposed to take off on full throttle. FAR really shouldn't be allowing jet engines to push you beyond 100m/s on ground level.

If a jet engine wasn't able to push you beyond 100 m/s at sea level, how would jet fighters be able to reach Mach 1.2 (408 m/s) at sea level? Besides that, if you didn't have the excess thrust to pass 100 m/s, how would your vehicle be able to produce the work necessary for it to gain altitude? After all, something needs to be able to produce the energy to fight losses to drag / gravity, and if drag is already eating all of that energy on the runway, there's no way for it to get into the air.

That sounds more like a mod conflict than actual 'realism' where both aerodynamic forces apply. But that's besides the point on whether DRE should be added to KSP without re-balancing it first and all this proves is that it's incompatible with FAR.

DRE is set up to model realistic aerodynamic heating and FAR is set up to model realistic aerodynamic forces; how are you getting "incompatibility" out of that?

Edited September 6, 2014 by ferram4