Realism in KSP

[Sky_walker]

I don't have a problem with DRE either, but it's very unforgiving for others, which is why it's remained a mod so far. The fact that you have to quote Scott Manley, a person who has a PhD in physics and who is also on the extreme end of people who are skilled at KSP serves to reinforce the point that DRE incorporated into the game is too difficult as it stands. I'm not against having re-entry effects in game, but it needs to be scaled down in harshness the same way the planets are scaled down.

Bad implementation of deadly reentry - as it is in DRE mod now.

Good implementation of deadly reentry:

- Flight path colored red when reentering atmosphere in too steep angle

- Clear warning in a HUD telling you are approaching too fast at a wrong angle

- When clicked game shows you a popup window with possible solutions to the situation

Let me repeat it again:

Difficulty in a games is directly proportional the the user interface.

Man, I loved those sims. Jane's F/A-18 was my favorite, though...

Wow. First time I see someone mentioning that game! Man, I spent months playing it back in a day. Mission editor it had was outstanding. So much fun!

[Levelord]

"Full throttle?" Try that in something like an F4U in a game that actually models torque some time. BTW, virtually no one uses full power on takeoff, it reduces engine life. Just like military aircraft don't use afterburners on takeoff. You use as much power as you need. In your other sims, the jets were not designed by YOU, in 5 minutes, using "legos."

You're mixing up throttling with afterburners. Thats like mixing up maxing a gas pedal with NO in cars. Commercial aircraft lift off using full throttle and then throttle back when they've finished their climb. The only times I'm aware of them throttling back during this time is for noise reduction but it is dangerous and against FAA regulations.

Military aircraft actually use full throttle to take off and its actually a requirement for aircraft carrier takeoff and landing.

[tater]

You're mixing up throttling with afterburners. Thats like mixing up maxing a gas pedal with NO in cars. Commercial aircraft lift off using full throttle and then throttle back when they've finished their climb. The only times I'm aware of them throttling back during this time is for noise reduction but it is dangerous and against FAA regulations.

Military aircraft actually use full throttle to take off and its actually a requirement for aircraft carrier takeoff and landing.

I'm not confusing burners with full (or even military) power. Airliners don't use the maximum capability of their engines on takeoff as SOP. They typically "derate" the thrust of the engines depending on the load of the plane, and almost never use full power (what would be "military" power on military aircraft). Afterburners are in addition to that (reburning).

Regardless, KSP planes have just engines tacked on and have no real balancing as real aircraft do. If a jet would be ideal with X fuel load, you'd not design it with 1.35X fuel because that's what size the tank happens to be.

I'm sort of amazed that aircraft work at all in this game

[Yukon0009]

Personally, I think the current realism fits KSP, for example the fact that flying into the atmosphere won't kill you. Atmospheric reentry damage would be seriously annoying while aerobraking, i.e. when your ship implodes during the aerobraking. But there are lots of good things about adding realism, for example the docking cam, which would be extremely useful.

[Red Iron Crown]

Personally, I think the current realism fits KSP, for example the fact that flying into the atmosphere won't kill you. Atmospheric reentry damage would be seriously annoying while aerobraking, i.e. when your ship implodes during the aerobraking. But there are lots of good things about adding realism, for example the docking cam, which would be extremely useful.

It needn't be all that annoying, just that it would require a bit of planning and accepting some limits on how deeply one can aerobrake. The aerobraking ship would need heat shielding and a shallower aerobraking angle (no more deep diving into an atmosphere at interplanetary speeds), but aerobraking would still be feasible and useful. Heat from aerodynamic friction could be difficulty scaled, too, from almost nothing to fully realistic in steps. Personally, I'd like Deadly Reentry Lite (Threatening Reentry?), somewhere in the middle.

I also think aerodynamic heating belongs as part of the aero model, not a separate thing as it is currently handled by mods. It is, after all, an aerodynamic effect.

[Whirligig Girl]

-snip-

Ah damn. Now you've convinced me. Almost.

Problem is that if you increase the Kerbol System's size, you also make in-space maneuvers much harder, as they require more delta-v for a Munar transfer, landing, etc. Launch Vehicles from StockSize+FAR might still work, but their payloads won't be able to do their jobs so you need larger launch vehicles anyway. If they were to do this, it would probably also merit 5 meter parts.

Of course, NovaSilisko (Former Developer! He made the PLANETS!) said that KSP is in an alternate universe anyway!

Things that definitely need to be added for authenticity:

--Shock Heating, with reflective and ablative heatshields. Best implementation idea I've heard for spaceplanes is a sort of "spray-paint" tool which allows you to paint on the underside of your plane to give it a reflective heatshield.

--NEAR Aerodynamics + Changes in physics with Mach number (This clearly goes under Flying and Crashing properly)

--Engine thrust scales with Isp.

--Isp scales down to zero. Have fun landing on Eve with an Isp of 180 or so

--Jet Engines. That is all.

[Whirligig Girl]

I'm just going to list the biggest gripes I have about realism, because I see that some of these haven't actually been mentioned.

1. Nuclear engines are underpowered and use LFO instead of LiqFuel + fissile fuel (for the fission reaction).

2. Reaction wheels are overpowered.

3. Ion engines are overpowered.

4. The air is more akin to water than air.

5. Planets are too small, too high density, and too close together.

6. Kerbals don't immediately die upon EVA on Eve or Moho.

7. Random failures don't happen.

8. Career mode gives out funds like candy.

9. Asparagus Staging is a thing.

10. Fuselages have fuel in them and wings don't.

11. Jets are overpowered.

12. There is no re-entry heating.

13. Kerbals are somehow immune to G-forces.

14. No life support.

15. Asteroids have no gravity whatsoever.

And now, I'm going to say this:

I want Realism Mode to be an option when starting a new game, that way people who don't like realism aren't forced to deal with it, but it'll be there and fully supported (unlike mods, which tend to break every major update).

1: sure.

2: yeah.

3: erm... Do you want to spend 5 hours on a burn with an ion engine? Even back then the Ion was overpowered compared to real life, so burns that are currently 2 minutes might take 10 hours. If they add Timewarpable engines that would be another story.

4: probably.

5: probably not. I've already gone into detail on rescaling the planets. If you make densities for planets correct without scaling them up, you end up with Kerbin that has no gravity. Planets are as close as they should be for their scale. They just look closer because of the scale. I'm working on a game with some friends and The home planet is 1% the size of Earth, the moon is 1% the size of The Moon, and the distance is 1% the distance to The Moon. They look much closer than the Earth and Moon do, but the scale is still correct. NovaSilisko says KSP is in an alternate universe, so problems with planets are explainable

6: They wouldn't die on Moho in an upgraded lunar space-suit. On Eve, they'd need an upgraded suit, sure, but who's to say they DON'T have upgraded suits already?

7: Random failures are the WORST. Why have failures that you can't predict and can't solve in a GAME. The only enemy is the laws of physics and your design flaws are your weaknesses. Random Failures would not be hard, they'd just ruin your mission.

8: 0.25 will allow you to set funds income.

9: In theory it's a thing IRL. I think they ought to make fuel pumps a lot more expensive because of how hard crossfeeding is IRL, but if you have Soyuz-style shape for boosters Asparagus staging works fine. I even did a science fair project on this last year, using FAR. Obviously my tests had some amount of usefulness, as I won a medal from the U.S. Navy, which do/I] work with aerospace if you didn't realize this.

10: I agree that wings ought to have toggleable/tweakable fuel tanks.

11: Yup.

12: I agree.

13: While I think Kerbal anatomy helps with gee-forces, they ought to die at 30 gees or more.

14: Simple like support such as "Snacks!" mod is the best way to go with this.

15: As far as your game is concerned, they don't. Their mass would give them gravity so low the game wouldn't simulate it anyway. That's even assuming that gravity in the kerbal universe is 11 times stronger (as NovaSilisko suggests). Asteroids IRL that are targeted for Asteroid Redirect Mission don't have measurable gravity either, aside from the biggest ones for Asteroid Boulder Retrieval Mission, which in KSP terms would be G-class asteroids, and even then it would take days to fall from 10 meters up.

[nine_iron]

The greatest step towards realism would be to remove the time warping options...

[Red Iron Crown]

The greatest step towards realism would be to remove the time warping options...

Can't tell if trolling or serious. Why would removing time warp be desirable for anyone?

[Xyrus]

Can't tell if trolling or serious. Why would removing time warp be desirable for anyone?

Because by the time I get to Eeloo I want to be playing this game on a HOLODECK!

I think he was just being a smart@55. I don't think this game would have many players if you had to wait days/months/years IRL to do anything.

[NonWonderDog]

Fighter pilots being what they are, fighter jets quite often take off in afterburner, especially if they're trying to look cool for an audience (or are taking off from a carrier). An F-15 at full afterburner can take off in a VERY short distance. For instance:

What they absolutely don't do, is go from one end of a 15,000 foot Shuttle Landing Facility runway to the other at full afterburner before rotating. The tires would explode and the plane would probably spin out of control at more than half the speed of sound and possibly disintegrate due to aerodynamic forces before exploding in a gigantic fireball and strewing a several mile area in front of the runway with flaming wreckage. The fact that such a takeoff is standard procedure in KSP for many players is slightly odd.

[Flixxbeatz]

So then I stumbled on this thread. Reminds me of a post I made about a somewhat related thread a year ago...

EDIT: And yea. That's always my reply when someone pushes further realism in KSP.

[I_Killed_Jeb]

So then I stumbled on this thread. Reminds me of a post I made about a somewhat related thread a year ago...

EDIT: And yea. That's always my reply when someone pushes further realism in KSP.

Yes let's just reduce all well-reasoned opinions to the absurd and then pat ourselves on the back for a job well-done

[DaMichel]

Can't tell if trolling or serious. Why would removing time warp be desirable for anyone?

I believe he is using sarcasm.

Btw i think i got that stone about the Kerbin rescaling issue rolling. Sorry about that That was definitely not meant as suggestion to make it full blown earth-like. The point was to increase DV required to reach orbit under FAR.

Different topic, regarding FAR and real aircraft and flight sims. It is great how KSP+FAR+DRE teaches you how to design in addition to how to fly. On the other hand, i should give Ferram4 and the other experts credit for this. Much thanks! With flight sims you get a pre-built well designed planes and never have to wonder why it performs the way it does. The exact numbers are off because KSP has strangely OP engines, unrealistic mass densities and so on but i guess nobody would argue that FAR+DRE get the fundamental physics better than the stock game.

Ultimate what is being implemented depends on Squads vision of the game and financial consideration i suppose. I actually wonder if we can reasonably expect Squad to fix these things no matter how much we complain. After all, half of their user base doesn't seem to care, they put respective issues on a not-to-suggest list and instead, they focus on implementing "gamey" stuff like contracts and "strategies" and more parts. I seriously doubt that they will implement any realism improvements at this stage. If they they wanted to do that, they would have already announced something like "Hey guys, we hired this aerospace-guy to implement nice physics models for us. It will be ready in a couple of months, stay tuned!". Yes, unless they want to delay release a few years, we would have a definite answer about this already. It is not something to take lightly.

[Zander]

...

I think this is because of the limitations of the engine. This engine is woefully inadequate for what its being used for, and I think the devs realize that trying to squeeze any more realism out of it is futile.

If you want total realism your only option is Outerra. Not only is it physics based unlike Orbiter, but it includes a proper aerodynamic simulation engine. The devs there regularly state they want to make everything as realistic as possible, case in point the painted lines on the road actually have a physical thickness of a few millimeters in that game and the undersea topography of the earth is mapped so you can actually visit the 11km deep Marianas trench (but its pitch black because the game vehicles don't have lights yet).

I play ksp because outerra's not finished yet so KSP is the best there is.

Edited September 6, 2014 by Zander

[Red Iron Crown]

Ultimate what is being implemented depends on Squads vision of the game and financial consideration i suppose. I actually wonder if we can reasonably expect Squad to fix these things no matter how much we complain. After all, half of their user base doesn't seem to care, they put respective issues on a not-to-suggest list and instead, they focus on implementing "gamey" stuff like contracts and "strategies" and more parts. I seriously doubt that they will implement any realism improvements at this stage. If they they wanted to do that, they would have already announced something like "Hey guys, we hired this aerospace-guy to implement nice physics models for us. It will be ready in a couple of months, stay tuned!". Yes, unless they want to delay release a few years, we would have a definite answer about this already. It is not something to take lightly.

That's a good point. As I understand it, Squad is focusing on getting all the planned features into the game before refining existing ones, so I'm hopeful that at least some of these realism features are revisited after the career-mode gameplay mechanics are fleshed out. We'll have to wait and see, I suppose.

[DaMichel]

... Awesome video and stuff ...

Exactly. And in KSP+FAR you can do the same thing. Provided you design for it. Who cares about TWR? Just rotate at the appropriate speed (around 100 m/s seems good) and go straight up vertically like this pilot does

I think this is because of the limitations of the engine. This engine is woefully inadequate for what its being used for, and I think the devs realize that trying to squeeze any more realism out of it is futile.

I don't buy this because i talked about stuff that has been implemented as plugin already. Which might be the reason why they don't look into it.

That's a good point. As I understand it, Squad is focusing on getting all the planned features into the game before refining existing ones, so I'm hopeful that at least some of these realism features are revisited after the career-mode gameplay mechanics are fleshed out. We'll have to wait and see, I suppose.

I cannot really buy this either. The problem is when adding parts for instance they have to do the config twice. Now for stock - and later for whatever is improved. Consider this, now that Sp+ will be included in stock. With respect to this it is also very interesting to see the following statement from the B9 (a very popular mod) developers. They seem to have decided already. "The stock drag model is no longer supported. FAR or NEAR are now required. While the parts will function in the stock drag model and are balanced, we won't accept bug reports relating to performance under simulated soup conditions. :^)"

[cantab]

Well the KSP runway's somewhat shorter, about 1.8 km or 5900 feet, which is on the short side for airliners, but then few KSP aircraft approach airliners in size and mass. Being unable to takeoff until passing the end of the runway, and more generally needing excessive speed, is often down to poor gear or elevator placement preventing rotation.

[Red Iron Crown]

I cannot really buy this either. The problem is when adding parts for instance they have to do the config twice. Now for stock - and later for whatever is improved. Consider this, now that Sp+ will be included in stock. With respect to this it is also very interesting to see the following statement from the B9 (a very popular mod) developers. They seem to have decided already. "The stock drag model is no longer supported. FAR or NEAR are now required. While the parts will function in the stock drag model and are balanced, we won't accept bug reports relating to performance under simulated soup conditions. :^)"

There's going to be further balance passes one way or the other. If KSP were being developed using the traditional closed model, it would definitely make sense to have just a few tester parts until all features were implemented to reduce the .cfg editing work. But with the early-access model, they almost have to include at least a couple of parts with each release to satisfy the endless hunger for them. It definitely does make implementing features more labor-intensive, which is one of the causes of the "development asymptote" that HarvesteR talks about regarding the pace of development.

As I say, I remain hopeful but we'll have to wait and see.

[lincourtl]

I actually wonder if we can reasonably expect Squad to fix these things no matter how much we complain. After all, half of their user base doesn't seem to care, they put respective issues on a not-to-suggest list and instead, they focus on implementing "gamey" stuff like contracts and "strategies" and more parts. I seriously doubt that they will implement any realism improvements at this stage.

I agree. As I've said before, just look at how KSP is being marketed via social media. The emphasis is on absurd creations and explosions. I think you have to assume this is intentional.

Mind you, I'm not judging. I've said many times now that I understand the tradeoffs Squad is making in trying to produce a fun game that will enjoy mass appeal, and yet features fairly realistic spaceflight. Even as it stands, I think what HarvesteR and everybody else at Squad have created is really special. In the end though it's a game so if people just want to create explosive pancakes, that's fine with me.

It's just frustrating because it would take so little to have something that is both fun and much more realistic than present.

Now let's see how much of what I just typed can be taken out of context.

[nine_iron]

I think he was just being a smart@55. I don't think this game would have many players if you had to wait days/months/years IRL to do anything.

Pretty much, I love simulations and I love smart @55 comments

But really I am hotly anticipating the option to control realism, I probably wont touch it but its good to know its there for those who want it.

[regex]

That's a good point. As I understand it, Squad is focusing on getting all the planned features into the game before refining existing ones, so I'm hopeful that at least some of these realism features are revisited after the career-mode gameplay mechanics are fleshed out. We'll have to wait and see, I suppose.

What'll actually happen is SQUAD will drop the ball (so to speak) by claiming the feature(s) "aren't fun". Then they'll use a "it really doesn't matter to the average player" excuse for the little corrections proposed here, and then claim that a mod does it better in the case of aerodynamics. Basically we're going to continue getting a mediocre arcade game instead of what could have been a great scaled-down simulator (and it really wouldn't take much to get there, IMO).

SQUAD has a history of disappointment with me, and I presume others, and I don't have any reason to expect that to change.

[Red Iron Crown]

What'll actually happen is SQUAD will drop the ball (so to speak) by claiming the feature(s) "aren't fun". Then they'll use a "it really doesn't matter to the average player" excuse for the little corrections proposed here, and then claim that a mod does it better in the case of aerodynamics. Basically we're going to continue getting a mediocre arcade game instead of what could have been a great scaled-down simulator (and it really wouldn't take much to get there, IMO).

SQUAD has a history of disappointment with me, and I presume others, and I don't have any reason to expect that to change.

I'm trying to stay optimistic. Hopeful, even. Maybe I'll be disappointed, but Squad have done well so far with KSP in my book so I'm willing to give them the benefit of the doubt.

[Sky_walker]

Well, to be fair - something did change.

When I first came in, which wasn't long ago, SQUAD was notoriously bad in it's communication with the community. One of worst in all of the "early access" projects i ever supported. And look how they improved in last month or so. We get screenshots, we know nearly everything about next patch (and still Devs left a door open for one surprise to come in from them - which sounds like a beautiful balance), we see one or another dev popping in to leave a comment on a forum, they even got 2 Q&A sessions quite recently. And look how their blog posts improved. Or heck - they've even skipped the nearly-mandatory cinematic!

Now it's actually interesting to wait for a next patch, not frustrating.

I know, I know, I'm new here, perhaps too optimistic, you regex post here since 2013 so surely got more experience, but what I seen happening in last month really looks like SQUAD can change for a better.

[tntristan12]

What'll actually happen is SQUAD will drop the ball (so to speak) by claiming the feature(s) "aren't fun". Then they'll use a "it really doesn't matter to the average player" excuse for the little corrections proposed here, and then claim that a mod does it better in the case of aerodynamics. Basically we're going to continue getting a mediocre arcade game instead of what could have been a great scaled-down simulator (and it really wouldn't take much to get there, IMO).

SQUAD has a history of disappointment with me, and I presume others, and I don't have any reason to expect that to change.

*cough*resources*cough cough*

Seriously though. If squad didn't have a history of being wishy washy (perhaps unpredictable is s less loaded word) about previously held positons I'd say this whole debate is cut and dry. I was so shocked when.HarvesteR blatantly said that they were not considering delta-v readouts because "that took the magic out of trial and error" (paraphrasing: the exact link is necessary), and decided resource mining would be boring because it'd be grindy (totally unlike the science system we have now) in favor of multiplayer. It seriously set a dangerous precedent for discussions like this, since the old guard (myself included) are still operating under the assumption that at least the appearance of realism is important to them.