Was KSP a somewhat accurate space flight simulator?

[Xforce]

KSP was one of the most famous science games as it simulates a complete space program, as well as physics such as aerodynamics and orbital mechanics. Such science focused games are rare, and something favourable amongst nerds. To make it even better, it is easy to modify the game and add almost any features you want.
However like everything, this game have it’s limitations in terms of physics. I’m not accounting mods like Principia, I’m just counting the stock game. Here are some flaws I have found.
1. Single body gravitation: KSP only accounts the gravity influence on a spacecraft from a single celestial body. Where this is very different from reality, where in KSP you can instantly change your trajectory as you encounter different celestial bodies.
2. Planets on fixed orbits: planets in KSP appears to be running on rails. In reality their orbits can shift due to the gravitational pull of another planetary body. Also, no matter how much force you exert on a planet or moon (like Gilly), you can’t change their orbit even the slightest bit .
3. No GR/SR. For the special relativity part, you can examine it simply by enabling infinite fuel and accelerate to c. As you approach the speed of light, there are no time dilation, no mass change at all. You can freely go FTL using just rockets.
4. Limitation of aerodynamics: you don’t hear a loud boom as your aircraft break the sound barrier. And the color of shock doesn’t change according to black body radiation as you go very fast. Even an entry at 10 times the orbital velocity (around 20km/s, achieved by opposite orbits from kerbin, the first picture) the shock is still red and yellow. In reality it would appear to be violet (see the Japanese Huyabusa spacecraft high speed entry, which is the second picture )
Should these physics be improved in the stock game? Will them add more fun and challenging, or becomes disastrous for the game?

 

[Drakkith]

Should these physics be improved in the stock game?

No, I don't think they should. They are either entirely inconsequential in the scope of the game (for things like the planetary orbits) or they are too computationally expensive to perform in real-time (things like accurate aerodynamics and the side effects thereof). The game already struggles with the physics when launching spacecraft consisting of a few dozen parts or more, depending on your machine, so I think they've gone with a good compromise of accuracy and game speed.

Will them add more fun and challenging, or becomes disastrous for the game?

Most likely they would just add unnecessary complexity that the player just doesn't care much about. I don't play the game to get a super accurate simulation of spaceflight, I play it to have fun. It is just the right amount of realism for my taste.

Now, after saying all that, I have to throw in the caveat that I'm not against adding things to the game as long as they make it useful and/or fun. By all means, add the planets to the gravitational physics engine if you also throw in some way for that to be meaningful and not just a waste of processing power.

 

[Rive]

Should these physics be improved in the stock game?

I don't think so. Just remember to those attempts to make space sim flight based on the most elemental Newtonian real physics - they died out pretty fast and today in most of the popular space games spacecraft s are are still like submarines in the endless waters of space...
If something is too far off from the usual/average physics experience then it is doomed to be fun for some geeks only => not enough players/buyers.

 

[Xforce]

I don't think so. Just remember to those attempts to make space sim flight based on the most elemental Newtonian real physics - they died out pretty fast and today in most of the popular space games spacecraft s are are still like submarines in the endless waters of space...
If something is too far off from the usual/average physics experience then it is doomed to be fun for some geeks only => not enough players/buyers.

Actually, simulating a giant spaceplane inside the atmosphere was intense physics simulation(since it have to consider the air flow over every part externally) , so I think advanced aerodynamics will be unnecessary.
But orbital mechanics appears to be easy for the computer to handle as I think in this case, it thinks hundreds of parts as a whole...

 

[member 656954]

But orbital mechanics appears to be easy for the computer to handle

Download AstroGrav - you can use it free for a week or so - and set the date to some time in the future / past and animate the planetary orbits of just the main solar system objects and it manages about one year of orbital time every 5 seconds. I've a 16GB current Gen i7 with a 265GB SSD and it is clearly not easy, computationally. When I do this with asteroids loaded the PC starts to slow down and the fan spins up. Correctly calculating a lot of orbiting objects and rendering them takes some grunt.