tater

You're grasping here. Really. Kerbals operate their space program however I operate it. That's the point, right? I'd say that any improvised improvement by my astronauts becomes SOP. How do I do that? Abstraction is fine within something that is already abstracted at a similar level. Anything they do to "career" stuff like rep, science, funds is perfectly fine since those systems are huge abstractions to start. In FLIGHT, a couple m/s can actually distinguish success and failure. Abstracting things in flight need the fidelity of flight even as abstractions, or they break it.

Could you tell from watching which pilot was which, yes or no? If you had readouts on how much fuel they used as well, and they were identical burns, etc. That's the test. If it uses more fuel, it should use more fuel. Not have one judged "more efficient" with exactly nothing to show for it in trajectory. Set up a KSP scenario. A rendezvous and docking. Have a new player do it, and other players of various experience. The amount of fuel and RCS left gets measured for each. If you watch them, they will fly entirely different paths. You'll be able to tell en route who is doing a good job and who is not. You'll likely be able to rank them reasonably well just by watching a youtube of them doing it. What you are suggesting is the same person flying the exact rendezvous 5 times, all looking pretty much identical, then having the fuel use remaining altered arbitrarily. Our great player might actually do a little less well on one, but that might be the expert kerbal and uses less fuel, even though anyone watching would say he made a big error (say he hits "Z" by mistake). ,

The player is controlling everything, though. If the verbal is, then in a "scenario" play, with a kerbal of stellar skill, a noob should not be able to fail a simple task. If he can crash the Mun rocket, Jeb is obviously not to be considered in control at all. (or dock, whatever). Yeah, but how they do things differently matters. We can watch the ship do what it is doing. I should be able to watch 2 ships rendezvous and dock with a station (load a quicksave and do this, one of skill 1, the other skill 5) and tell FROM WATCHING THEM which is more efficient. If the two maneuvers look identical, but one uses less fuel… MAGIC. What about if you do the Skill 1 better than you do the skill 5 (accidentally hit the wrong key)? The an independent observer might think that the one that used more fuel was actually considerably more efficient. Magic. A EE is going to rewire the ship in flight? He's then going to change it back when his copilot takes the helm?

Unless you don;t control it. Make AI kerbals, then have skill actually mess with the fidelity of the orbital maneuvers, not an abstraction of "efficiency."

This kind of comment helps no one, and I think that the people that hate the new "feature" idea are not even remotely suggesting anything negative about the devs. Good game design is hard. Really. The simulation aspect is far easier, and easier to do well when more "realistic" because they then know how things are supposed to work without introducing problems. A good rule for science fiction in general is to "break as few laws of physics as possible." Usually the ante level is some sort of FTL travel. Breaking regular physics has a ripple effect, as the tech to do that easily becomes scary (any sufficiently interesting space drive is also a powerful weapon). We all would like to have Kerbals matter, the trick is to come up with constructive ideas how this can be done without breaking physics.

It's worse than that. The same actual pilot (the player) could do the exact mission 5 times… say he's taking stuff to be built in orbit… someplace. All 5 will arrive with predictably different fuel volumes. (nominal, 1% more, 2% more 3% more, etc). Make AI piloting a thing (like MechJeb), and if the player CHOSES to let Bill Kerman fly, then they see what happens when he flies, including skill level. Add uncertainty to each maneuver done by AI pilots. Maybe there is an alternate way to set mission goals instead of the player making a node, and the AI pilot doing it? What if you could set the mission goal as an orbit at 30km above the Mun, instead of designing a couple nodes? Then if the first injection is botched by AI, the AI has to do a correction… using more fuel. You see how this can work. The trade off for the player is the ability to let the astronauts do their thing, but he might have to bail them out if they screw up. Which is FUN, actually.

The engine leaves the factory floor able to do 133% of what is on the brochure. Why can Jeb use 133, and Max can only use 100%? He has longer arms and can push it forward? The mechanism for this is overheating, and is already in the game. Engines that actually overheat are really "100%" at max where they don't overheat, and anything above that is dangerous. So very well said. There is NO conflict between simulation and fun. None at all. It's not a thing. Fun has to do with game design choices, not the underlying mechanics that tell the parts which way to move.

It's magic, that's the problem. The game can be many things, but "magic" shouldn't be one of them. How about discovering alien artifacts to buff your ships… that could be fun, too. Lots of games have power-ups, right?

I've played plenty of actual RPGs before. With dice. Skill simply modifies randomness. They are adding the skill, but none of the randomness. The problem is the combination of an abstraction (skill), with a simulation (yes, simulation, the game is simulating space flight, regardless of how accurate that ends up being). The 2 just don't mix nicely with the play controlling things. I'd rather have AI kerbals to use for missions that have become tedious.

Vs making it even easier for the experienced that don't need it?

Yeah, 5thhorseman is right there. Of course if my rocket is swerving violently, I should see it swerving violently. The inefficiency needs to manifest itself. That's easy with AI pilots, you'd watch them struggle.

I'm a total noob, but I picked the right strategy at the admin building, so I leveled up before I even made it to the VAB.

More thrust is patently absurd, that's kooky (unless ole Jeb's breakfast was "off" and he's adding his own propellant to the cause out the window). Efficiency? That's really a poor choice of heading in atmosphere, but that should only be true if the rocket actually flies the heading that results in less drag, and left fuel used. So either take control from me and put the rocket in the idealized track, or don't, and no buff. In space, it's the choice of maneuvers (how close that plane change burn was to the appropriate node to match the target's plane), or the accuracy of doing those maneuvers (because if you screw up you have to correct, etc). As has been said, ALL have penalties for mistakes, not buffs for doing them better. If the top skill is 5, 5 is perfect, and all below should do WORSE. We have a boundary value here, the actual specs of the rocket. Automate node execution, have an error bar on each parameter, and have skill reduce its magnitude.

It's too deterministic to read as "skill." Bad pilots have good days, even great days, and good pilots can have a bad day, even if it is less bad than the bad pilot's bad day. Honestly, short of AI, I don't see skill as working out well in terms of flying. Not that I will ever know, because I'd turn it off.

I am fine with roleplaying the Kerbals, and don't necessarily need to see myself as the pilot. I don't see this as doing that. I'd rather see the Kerbals as NPCs, frankly. Then I assign them a task, say landing at my Mun Base, and how close they get (or if they wreck the ship) is related to their skill. I can totally live with that. I'm utterly uninterested in buffs, and would never turn it on, or not play to a point it took effect until modded out. Is there any way to actually make the Kerbal pilot skill matter short of full AI, weighted by skill level? I'm not convinced there is. I really think it requires randomness to work properly (skill as a thing in an RPG). Because they said flat out it will alter the performance. We don't have to wait, Squad said it, it's "a thing."

How do you get 4% more thrust out of a rocket engine that produces X thrust via "piloting?" How, pray tell, does someone "milk the engines?" In Star Trek… they invent new physics as SOP every episode, even using ST as an example says a lot to me in the context of even slightly realistic rockets.

Pilot skill IS decisions. That's the point. Role-playing requires that the decision actually gets possibly moderated by the game to reflect the character. The stupid kerbal would not think of the amazingly creative solution to problem X in orbital mechanics. So you should not be allowed to do it. Think real RPGs. "I hack into the Imperial Navy's computer, and take over the Tigress class battleship." GM: "your character has the IQ of a moderately intelligent labrador… Roll 10 or less on 10D20 to succeed." You are saying he should hack the computer, but only get 90% of it or something. I'm cool with RPG stuff in the game. This is not that.

The botched Mun landing is the perfect extreme example. The Kerbal pilot is great. Best possible. A legend. Jeb, but after years of experience and improvement. New, player, without two clues to rub together… should be incapable of botching a simple Mun landing with Super Jeb in control. If he can still crash, the game is NOT role playing pilot skill. If the noob player tasked AI Jeb with landing in crater X that is 50m in diameter, he'd nail it. That is modeling Kerbals as pilots.

In role playing, you have to play what you are. If your character is amazingly brilliant as a computer programmer, you get to make hacking rolls, even if you are personally challenged by upgrading your own video card drivers. I get the idea of skill altering capability. Really. I disagree with PDCWolf about "me" piloting the craft in many ways. It's Jeb (or whichever guy) driving. The thing is that it MUST be random, or it is a physics change. Random. A bad pilot can have his best day ever. Every pilot has best days, eventually his best days are better, and is worst days are usually less bad. But It's not that everything he does stinks, then everything is better. This mechanism means that every single maneuver will be XX% better. Always. Or worse. Always. On top of that, it has to interact with the player, THROUGH the player. Say I'm a middle of the road KSP space pilot. Where is middle of the road for skill buffs? normal physics? Or is the worst kerbal normal, and the best a buff? If a great KSP player pilot has a meh kerbal flying, can he still do better than ME, a meh player, but with a stud of a kerbal driving?

A better example. An "80%" pilot will always use more fuel than a 100% pilot. Always, every maneuver. He'll need to get every m/s for a critical, life-saving burn… but he will ALWAYS choke.

The suggested mechanism doesn't model bad driving. It models having a smaller gas tank.

It is violating physics if the player is flying. Because is is stated that it alters the parts/ship capabilities. If the kerbals are flying, it's just fine as long as it doesn't do anything that Mu actually said, and instead alters their piloting ability. Burn timing, duration, direction. Think MechJeb, but with error bars around every maneuver scaled to skill. Really dumb astronaut is still… an astronaut, so maybe the worst guy might be off by a couple degrees. The best is near perfect all the time. What is proposed is magic, period.

Even Mercury was controlled by a computer. The astronauts only really shine when there is a problem, and Squad has specifically said they don't like random problems. Instead we get random buffs?

I should add that WRT skills, what we are very specifically talking about is no longer speculation. Boosting thrust? Magic. Reducing heat generation? Not even sure what that means. Engines overheat now, so that means you can run an engine a few % higher without it overheating… Also magic. Increasing fuel efficiency? Magic. Boosting science? That's fine, science is abstracted. It would be like the others if he could take one soil sample, and get the biome he is in, plus another skill*XXkm away. The only way for skill to not be magic would be AI pilots, and the skill level affects THEIR PILOTING. Not thrust, not efficiency, but when they maneuver, how accurately they apply thrust, where the ship is pointed, etc.

That is exactly what it is. Every thrust is 10% more efficient means there is 10% more fuel, functionally. Period. They are identical.