Pilots: uses for them past early career?

[John FX]

You have not seen this yet Gliph?

Thanks for posting that, while I have seen it before I did read it again as I have been wondering for a while how a probe would do a capture burn on the Far Side of a planet where it would be out of signal and in the article you posted apparently a large probe core can have autonomous action which solve that problem quite nicely.

Although when speaking with RoverDude in a Devnotes thread he was rather cryptic about the solution that they had found to that particular problem when just saying 'you can use a large probe core' does not seem to warrant the secrecy.

[stratvox]

Yah. Since the problem is "probes make pilots obsolete", then any answer is going to have to be either "make pilots better somehow" or "make probes worse somehow".

Adding a remote-control requirement to probes, as they're doing in 1.1, might help with that-- would help even further if only pilots have the ability to remote-control probes that don't have LOS back to Kerbin.

Adding a lightspeed control lag to probes (as RemoteTech does) would help even more, though I suspect that would be too complex to add for the stock game-- would be too hard for beginning players, would require adding at least a limited autopilot, etc. So that's probably not an option for stock.

Actually, that makes a lot of sense. If you want to run a rover on the surface when you first reach a new body, you need to have a pilot in orbit above to drive the damned thing. You don't even necessarily need line of sight. If you were to model command delay because of c then having a pilot able to be there and exploit the nearly instant control they'd have. With command delay, you don't really get to drive a remote rover, you get to say "apply x force to y wheel for z seconds" and then wait for the results to come back to figure out your next move. I mean, it's okay for the guys running the mars rovers 'cause they've got no choice, but this is a game. Want to not have to do that? Send the rover with an orbital station and the station must have a pilot and command module for kerbals (i.e. a probe core isn't enough, they need a cockpit), or else you get to run the rover with a delay. Right now if I had one on Eeloo I'd be looking at a nearly fourteen minute delay between sending the command and finding out what happened. That kind of thing can be acceptable when you're talking about programming orbital maneuvers but makes docking and rovers completely impossible.

That would make pilots completely indispensable. Yeah, they're kind of payload in their own ships, but they'll be driving all the rest of the stuff around when you get them there for you. Make pilots necessary for "presence" when moving stuff; have the presence of a pilot within the SOI means that your control delay is zero... assuming the pilot has LoC to the probe/rover in question (that is, that they deployed a comm network around the body so they can control it on the other side of the planet. Even if they haven't, being able to drive the thing when you are in LoS is still a massive help overall.

There's another possibility to consider too. Right now, both pilots and probe cores use a bit of a naive approach to changing heading. With very large heavy ships this leads to problems (massive swings back and forth over your desired heading). Give pilots a newer, better algorithm that accounts for things like ship mass, rotation momentum and rcs twr to make them better at changing the heading of the ship. That's something that could even be improved as they go up to level five (up the fineness of control, which should increase the efficiency in time and monoprop etc to change the heading) so that they end up being far better at controlling the ships than the probe cores are. This makes pilots better in a way that makes sense from the PoV of a game. Yeah, the best probe will do maneuver node holding, but the pilot will get you there much faster and more efficiently in terms of resources (battery for the reaction wheels, monoprop for the RCS).

[Snark]

I'd argue that it's not simple at all-- find me anything else in KSP right now where "having X somewhere on your ship means that Y won't work." Parts do what they do, and the interdependence of parts is minimal.

Man, think one more time. Fuel. Electricity. They aren't here, something around there stops working. It's, like, the core-est of all the core mechanics.

No, that's a completely different thing. Read what I wrote more carefully (I've added emphasis above for clarity). Sure, everyone expects that "you have to have electricity to power electrical things", or "you have to have fuel to power engines". That's just resources, it's a core mechanic, and the game does various things to call your attention to it (like having fuel gauges and such).

This is a completely different thing: having a particular component somewhere-- anywhere-- on your ship will stop something critical from working. There's nothing like that in KSP, and for good reason: it would be complicated, confusing, and difficult to discover.

It's also a completely unintuitive thing that no player would expect, because it makes no logical sense at all. "My engine doesn't work, maybe it's out of fuel" is something that might occur to anyone, it follows logically from how rockets work. But "This probe core worked on this ship, but on this ship it doesn't seem to do anything!" is mysterious and frustrating. It goes completely against the real world: components stop working because they're either damaged or they're missing some resource they need to run. To magically stop working just because some other thing happens to be present is just arbitrary-- it's like reading a book where a character gets hit by a truck not for any reason connected to the plot, but because the author simply needed the character to be gone.

Sorry? If you dock a ship with a little crew cabin to something, it means the ship has a pilot. Thus, if both ships were controllable before docking, they're still controllable after docking. Unless you've managed to remove the pilot after docking... which is technically possible but seems completely random, overcomplicated and unneeded.

Not at all. I use space stations as parking lots for docked ships (crewed, uncrewed, and crew-capable-but-empty) all the time. Sure would be annoying and frustrating if a player was happily using such for a long time, and then just happened to leave an empty-but-crew-capable ship parked there and suddenly all of the uncrewed ships (and the space station itself!) go completely dead and uncontrollable and the player has no idea why. Especially since a long time and lots of missions might elapse in between last-kerbal-leaving and try-to-control-something-on-the-station, which would mean the player would have no idea when or why it went dead. It would just mysteriously not work.

Also: Imagine the frustration of the player who tries to build, say, a rescue ship that has an empty crew pod and a probe core. "Why won't this work?" Would be incredibly frustrating to debug, because no player would expect that the crew capsule would stop the probe core from working. I know I sure wouldn't. Do I have electricity? ...yes... Maybe I need to put an antenna on? ...no, doesn't do anything... Perhaps after a lengthy and frustrating trial-and-error process I'd come down to "Hey! It looks like adding a crew pod disables the probe core!" And the very next thing I'd do would be to log a bug, since that would clearly be buggy behavior. (And I imagine Squad getting pretty darn tired of receiving the N-hundredth bug report from a well-meaning player who has just "discovered" the "bug".)

Tell me, which would players want more-- ships that work as they expect them to, or arbitrary rules that force them to have pilots along? It would be a cure far worse than the disease, IMHO.

Like I said, it's not completely counter-intuitive.

Yes, it is. It's utterly nonsensical that a component would stop working because of the presence of a crew-capable part. It's not like my PC, house thermostat, answering machine, and everything else shuts down because I park my empty, crew-capable car in the garage and then drive away in my other car.

It's going to confuse players, and it's going to cause frustration. (I base this on all the other things I see players get confused by, from the past year participating in the "Gameplay Questions" tutorial. Awkward bits in the game far less confusing and arbitrary than that cause grief for players.)

I think it just comes down to what Red Iron Crown described earlier in this thread as the "elephant in the room": the plain truth of the matter is that technology makes pilots obsolete. You can have a game design that deliberately limits technology to keep the game fun (i.e. the fact that MechJeb isn't part of the stock game)... but you can't really get away from the fact that once you have a computer-controlled gizmo that controls a ship, you've just taken away a major reason why you need a live pilot. You can come up with other uses for the pilot; but I don't think you can put the genie back in the bottle.

In any case, I think we're gonna have to agree to disagree there-- I think we've reached the

point now, unless someone else wants to chime in on this issue.

- - - Updated - - -

There's another possibility to consider too. Right now, both pilots and probe cores use a bit of a naive approach to changing heading. With very large heavy ships this leads to problems (massive swings back and forth over your desired heading). Give pilots a newer, better algorithm that accounts for things like ship mass, rotation momentum and rcs twr to make them better at changing the heading of the ship. That's something that could even be improved as they go up to level five (up the fineness of control, which should increase the efficiency in time and monoprop etc to change the heading) so that they end up being far better at controlling the ships than the probe cores are. This makes pilots better in a way that makes sense from the PoV of a game. Yeah, the best probe will do maneuver node holding, but the pilot will get you there much faster and more efficiently in terms of resources (battery for the reaction wheels, monoprop for the RCS).

I like this idea.

It doesn't really solve the "useless pilot" problem-- it's still a pretty minor convenience, I can get along without it just fine-- but it's a really nice thing to work in there, and ought to be available for pilots from level 0.

By the way, the succinct term for what you just described is: "Pilots tune the PID parameters of SAS."

Edited October 22, 2015 by Snark

[monstah]

Pilot skill could affect PID tuning for SAS?

We all know it's crappy sometimes, and there are some methods for figuring out the best parameters based on mass, inertia moment, etc., and also methods for continually improving you parameters as the system handles better/worse. Pilot level could be the meta-parameter that directs how well the parameters converge.

Or, alternatively, level zero pilots have P controllers, level one have PD, level two have PID, three onwards feature algorithms for improving the parameters.

- - - Updated - - -

Also, it seems to me, based on its behaviour, that the SAS controller runs on pitch+yaw+roll instead of quaternions. RT2's computer does quats, and rotations are WAY smoother as a result (although it's been some months since I've used that mod, and my memory could be playing tricks).

That paradigm switch could also happen based on pilot level.

I'm in for trajectory predictions! Tho I still think a better SAS is in order, too.

I think a pilot is a "Flight Specialist". KSC has its flight operators on wire, but the pilot feels the ship.

To me that means better ship handling, I guess. [snip]

There's another possibility to consider too. Right now, both pilots and probe cores use a bit of a naive approach to changing heading. With very large heavy ships this leads to problems (massive swings back and forth over your desired heading). Give pilots a newer, better algorithm that accounts for things like ship mass, rotation momentum and rcs twr to make them better at changing the heading of the ship. That's something that could even be improved as they go up to level five (up the fineness of control, which should increase the efficiency in time and monoprop etc to change the heading) so that they end up being far better at controlling the ships than the probe cores are. This makes pilots better in a way that makes sense from the PoV of a game. Yeah, the best probe will do maneuver node holding, but the pilot will get you there much faster and more efficiently in terms of resources (battery for the reaction wheels, monoprop for the RCS).

By the way, the succinct term for what you just described is: "Pilots tune the PID parameters of SAS."

No offense, but am I talking to a wall?

[stratvox]

I like this idea.

It doesn't really solve the "useless pilot" problem-- it's still a pretty minor convenience, I can get along without it just fine-- but it's a really nice thing to work in there, and ought to be available for pilots from level 0.

By the way, the succinct term for what you just described is: "Pilots tune the PID parameters of SAS."

Thanks for the head's up on the proper term. I'll go digging in there a bit. A quick (and probably naive) take on it would be to up the precision of the error value as the pilot progresses in level so they get better at getting to the heading in short and efficient order. This is something that can really matter when you're trying to wrestle a class E asteroid in LKO during the time between eg AN and Pe.

I also think that the remote control aspect works really well; speed of light command delay is a real thing, it could be implemented at higher than standard difficulties, and would give an extremely compelling reason to stash a pilot in a station in a given system to eliminate the delay. Note that this won't change the need to build a comm network inside the system in question, but would mean that you could build one local to the system, with the idea being maybe actually hooking it up to the KSC on the next transfer window, due to cost/lift/tech reasons. Won't matter very much on easier difficulties, but will make pilots compelling at higher difficulties when exploring other planets. If you think of the player as being the KerbalHiveMind, having one of the members of the hive mind present during remote control actually makes a lot of sense... you're not controlling someone who's 7 light minutes away from the action, instead you're controlling one that's only a few light ms away. Having it tied into the difficulty settings in the way I described also makes it worth while imho.

You know, I mean, the hive mind can put itself anywhere without regard to c propagation delays because quantum entanglement mumble mumble hand wave

- - - Updated - - -

No offense, but am I talking to a wall?

Yeah, I got to your post after I made mine...

[fairytalefox]

I use space stations as parking lots for docked ships (crewed, uncrewed, and crew-capable-but-empty) all the time.

Got it. Okay, I was wrong, forget that idea. Thank you for the discussion.

[cantab]

Light speed delay really wants a good programming/automation system unless Squad are happy to cripple interplanetary probes. Such a system is obviously possible (NASA have done it!) but it's not a trivial task. Remote Tech's flight computer fails abjectly in my view, while kOS does well but is daunting for those without a programming background. Smart Parts looks good but I think it would get messy for more complex missions. I think the way to go would be a visual programming language, maybe something like what Lego Mindstorms uses, or perhaps more like Scratch. However it's done that's a big addition to the game and a big new direction for the gameplay.

[stratvox]

Light speed delay really wants a good programming/automation system unless Squad are happy to cripple interplanetary probes. Such a system is obviously possible (NASA have done it!) but it's not a trivial task. Remote Tech's flight computer fails abjectly in my view, while kOS does well but is daunting for those without a programming background. Smart Parts looks good but I think it would get messy for more complex missions. I think the way to go would be a visual programming language, maybe something like what Lego Mindstorms uses, or perhaps more like Scratch. However it's done that's a big addition to the game and a big new direction for the gameplay.

Well part of my thinking is that that only shows up on the highest difficulty levels, above the difficulty level that requires a satellite relay in the first place. But yes, an automation system that is simple to use and works is a completely necessary part of that kind of feature. It could be simpler than you say, though... while the probe is in LoC, set up the maneuver (and this won't require modelling c delay... you are setting it up on the computers at the KSC) and starting the upload with the maneuver before the satellite goes behind a celestial body and can't be reached, or more specifically the upload must be started before occlusion time - (time required to get the message there + time required to actually transmit the data). Dealing with say doing scientific measurements would be harder... an integrated system that would be able to handle all of the possible actions that could be taken would be more work. A visual front end to kOS maybe? I've not looked at it (though I've been tempted....) so I don't know how much of a task that would be.

And yeah, it's pretty big, but I get the vibe we're blue-skying here.

Edited October 23, 2015 by stratvox

[Gilph]

You have not seen this yet Gliph?

I have...

That's was my point when I said I did not care so much about LOS and relay. It is a hard problem to try and make pilots relevant when the probes have so much functionality. Everything I have in Jool is unmanned, because it's easy. Time lag, even with LOS, may be an option.

[Javster]

It's like real life: nobody pilots rockets, everything is controlled by computers, humans are just a payload

I'd suggest to make automated probe cores to be able to only control ships with so many science parts. If you want more experiments in one trip, you have to have a real pilot. Not very realistic but kinda solves the issue. Should solve.

"This craft is too complex (has too many parts) for an automated probe, get a Kerbal!"

[Waxing_Kibbous]

Pilots are tricky, definitely a tough thing to both balance and make them appealing. The OP idea seems interesting, as does giving pilots better sas capabilities, so the ability to pilot unwieldy craft better. Here's an idea: probes have mass limits on what they can control, better ones can go bigger, but even the highest level one has a limit.

(on a more extreme nerf level take out their reaction wheels and electric charge)

Edit: is there a way to hack a file to implement mass limits and disable the above features on probe cores? I wouldn't mind trying it out. Think about it, not only is Jeb the coolest character in video games, he is also the least needed! Jeb and Val need to be included more just because they are hyper-manic

Edited November 3, 2015 by Waxing_Kibbous

[Rath]

Pilots have (for me, at least) an interestingly difficult problem: they have a shelf-life with a rigidly-defined (and early) expiration date.

Up until the moment I unlock SAS-capable probe cores, pilots are utterly indispensable and I use them to the exclusion of everything else-- it's just too painful to try to control a ship that doesn't have SAS capability. It's not that engineers and scientists have no use, it's just that I have to have a pilot, and in the early pre-probe part of career, it's awkward to build multi-kerbal craft (the Mk1 pod really doesn't stack well).

However, the moment that I unlock the OKTO, suddenly pilots are totally obsolete. I have no use for them at all. Yes, it's kind-of-sort-of nice to have the advanced SAS modes available before I have the top-end probe cores unlocked, but that's only a nice-to-have, not a have-to-have; whereas I really need what scientists can do (reset experiments) and engineers (mining). So for the 90% of my career that happens post-OKTO, pilots are just taking up space. It becomes particularly frustrating for me, since I like to populate my program with rescued kerbals, and fully 1/3 of them are useless pilots.

In general, I really like what Squad has done with the various professions, but pilots are in an interesting hole. Engineers and scientists have a use the entire game, but pilots get left out in the cold. I'd love to see some sort of mechanic that makes pilots have some use throughout the game.

One idea that occurs to me: The upcoming patch will add the RemoteTech-like need for probe cores to be able to talk back to KSC via line-of-sight or relays in order to be controllable. It will also make it possible for kerbals in a ship to remote-control probes. What if you make it so that only pilots have remote-control ability, and also that controlling a probe core in your own craft counts as "remote control". Thus, if I send a scientist to the Mun with a probe core on the ship, and then the ship goes around behind the Mun and loses LOS to Kerbin... suddenly it's an uncontrolled probe core, and the scientist has no SAS. That would give a reason to send pilots along: "well, either I can add a pilot to the ship, or else I have to build a relay network of satellites."

Anyway, that's just one idea, I'm curious what other folks have thought about this. Discuss?

I like it

[passinglurker]

I'm still sticking by the "give pilots facility feats as an alternative to investing in infrastructure" idea earlier in the thread. it makes pilots useful if you want them to be useful thereby injecting an element of choice into career.

[NWM]

And what if the experienced pilot can do something extra, what the probe cores not: ?

Auto-execution of nodes

Docking

Navigating rovers independently to destination (long rover trips can be fatiguing)

Or even fly an aircraft...

[TrainEngie]

Maybe the science experiments and resetting experiments should take time, according to who is aboard the spacecraft.

Probe Core, 10 sec

Pilots, 8 sec

Engineers, 8 sec

Pilots + Engineers, 6 sec

Scientists, 4 sec

Scientists + Pilots, 2 sec

Scientists + Engineers, 2 sec

Scientists + Scientists, instant

Scientists + Engineers + Pilots, instant

Maybe this would make pilots and engineers contribute to science a little more.

[passinglurker]

And what if the experienced pilot can do something extra, what the probe cores not: ?

Auto-execution of nodes

Docking

Navigating rovers independently to destination (long rover trips can be fatiguing)

Or even fly an aircraft...

There are people who are vigorously opposed to autopilot and either way it isn't a viable solution for them cause they fly by hand so they are still stuck with the useless pilot conundrum whether they like it or not

Seriously lvl1 get maneuver node planing, lvl2 get patched conics on the ship the pilot flys before upgrading buildings. This is the best if not only viable option all others turn pilots into mech jeb(has a vigorous opposing faction), magic isp generators(like mechjeb this also has a vigorous opposing faction), or arbitrary roadblocks to annoyingly force you to use them and is usually poor game design(I am vigorously opposed...).

Only other solution I've seen in this thread that wasn't honestly terrible is using pilots as a general buff generator for the other classes making engineers yield more isru efficiency and scientists yield more science but being unable to do the jobs themselves.

[KerbMav]

I am also still against pilots making engines more efficient. The effects of scientists and engineers is somewhat different. Scientists can react to arising opportunities during an experiment and engineers are merely keeping the drills running more smoothly to not overheat them to the point of inefficiency.

"make pilots better somehow" or "make probes worse somehow".

Exactly the problem/question here.

Maybe the RCS could be only controlled by pilots until you have the most advanced core?

Or just "better" control?

I think I like the OP's suggestion best so far, make pilots required to control things remotely. Though I guess with no lightspeed lag we could just park a pilot at KSC or in LKO and run everything from there.

Line of sight?

"This craft is too complex (has too many parts) for an automated probe, get a Kerbal!"

Like the idea!

And what if the experienced pilot can do something extra, what the probe cores not: ?

Auto-execution of nodes

Docking

Navigating rovers independently to destination (long rover trips can be fatiguing)

Or even fly an aircraft...

And this one too!

----------

To summarize a bit - well, the parts of the thread or rather the ideas I liked the most and some of my own 0.02 €:

Probecores should always be inferior to pilots.

SAS-autopilot - when asked to point to a specific node on the navball - will tend to sway back and forth before it settles on the node, especially with heavier crafts. Pilots should always be better, more efficient, waste less energy/RCS-fuel by not pushing the stick to the max continuesly, but initiating a spin/rotation and waiting for it to reach the node, starting to stop the rotation in advance and sway less before settling again. (Granted, this can be avoided by the player steering the ship manually.)

Probecores should be uncontrollable without line of sight connection to Kerbin/KSC.

Quoting from the dev-blog:

"... a probe must establish a connection back to Kerbin or another ‘control point’ via an antenna part in order to operate ... control points are the planet Kerbin, or a craft with an antenna, a pilot Kerbal, and optionally a large probe core."

Pilots flying a craft in atmosphere should have different nodes available to them.

e.g. normal/radial is only of use in orbit. Instead they should get switches to: hold airspeed, hold altitude above sea level, hold attitude towards horizon.

Probecores should have a parts/weight (w/o fuel?) limit to have control over a ship.

Leads to difficulties regarding necessary larger launch vehicles - maybe limited to a certain altitude above Kerbin?

Further musings:

Ships out of range (NOT line of sight!) from Kerbin (KSC) could only have trajectory projections available to them while they have a pilot AND a probe core/advance computer on-board.

Pilots could be the daring species Jeb represents, being able to go on EVA earlier than the building upgrades would allow.

Somewhat representing the commanders of a ship, they could enable lower ranking crew members to operate on a higher rank.

e.g. a rank three pilot enabling scientists/engineers to work as if they were on level one?

[Snark]

Pilots are tricky, definitely a tough thing to both balance and make them appealing. The OP idea seems interesting, as does giving pilots better sas capabilities, so the ability to pilot unwieldy craft better. Here's an idea: probes have mass limits on what they can control, better ones can go bigger, but even the highest level one has a limit.

(on a more extreme nerf level take out their reaction wheels and electric charge)

Edit: is there a way to hack a file to implement mass limits and disable the above features on probe cores? I wouldn't mind trying it out. Think about it, not only is Jeb the coolest character in video games, he is also the least needed! Jeb and Val need to be included more just because they are hyper-manic

Yes, it's quite easy to write your own "mini-mod" that will do most of what you want. I don't think there's any way to implement your "maximum mass limit" requirement without actually writing code, but everything else is a trivially simple snippet of config using ModuleManager.

1. Install the ModuleManager mod, if you don't have it already.
   
   • Although if you're running any other mods at all, you almost certainly have it already. Practically everything uses ModuleManager.
     

[*]Go into your game's GameData folder and create a file named with the ".cfg" extension.

• You can call it whatever you want and put it anywhere in GameData (or subfolder thereof), it just needs to have ".cfg"
  
  

[*]Add appropriate config to that file (see example below)

[*]Start up KSP.EXE

[*]Enjoy the misery of worse probe cores!

ModuleManager syntax is documented here and here and here.

Here's an example of a config file for making probe cores worse. It removes electricity storage and reaction wheel capability, doubles the mass and electricity consumption, and dumbs down the cores with most advanced SAS capability by one level. (This config will work on everything in the stock game. If you're running a mod that adds new probe cores and want the config to apply to those as well, you'd need to add an "AFTER" qualifier to the config; see ModuleManager documentation.)

// Example of how to make probe cores worse

// Changes to all probe cores
@PART
[*]:HAS[@MODULE[ModuleCommand]:HAS[#minimumCrew[0]]]
{
!RESOURCE[ElectricCharge] {} // No electric charge
!MODULE[ModuleReactionWheel] {} // No reaction wheel
@mass *= 2 // Double the mass
@MODULE[ModuleCommand]
{
@RESOURCE[ElectricCharge]
{
@rate *= 2 // Double the electricity consumption
}
}
}

// Changes to all parts with high-level SAS
@PART
[*]:HAS[@MODULE[ModuleSAS]:HAS[#SASServiceLevel[3]]]
{
@MODULE[ModuleSAS]
{
@SASServiceLevel -= 1 // dumb it down
}
}

Personally, I'm not a fan of this as the "answer" to giving pilots a use in late game-- I like how Squad has balanced the probe cores, for the most part. So I wouldn't be in favor of this as a change to stock behavior. But if you want to play around with it on your own, have fun!

[Waxing_Kibbous]

Thanks for the pointers Snark, I think I may try this as a preliminary trial, but instead of being heavier I may make them a bit lighter due to no reaction wheels/internal battery, and keep the electric charge usage the same.

I understand that if my changes were to be made to stock KSP people would flock to Mexico City with their pitchforks and torches, but again IMO the probes are overpowered because of the simple reason that as soon as you unlock the most basic one you really don't need a pilot any more. This particular nerf will take a bit of thought, when I come up with a decent outline I'll head over to the modding forums. Thanks again!

Edited November 3, 2015 by Waxing_Kibbous

[fireblade274]

Ive been thinking about this: What if 5 star pilots can use the nav ball to automatically find the best trajectory to a planet or intercept?

Its a long shot but, yea

I know how the nav ball works (2k hours in) but it would be nice to have a auto-trajectory feature for those intersepts with things zooming past a planet that cant get into orbit (runway ships that ran out of fuel) with out mecjeb or other mods. I figured this would kill 2 birds with one stone. ill probably make a suggestion post about this too

[Sanic]

Giving pilots a less overpowered version of MechJeb seems like a good idea to me.

[panzer1b]

I use pilots in every military vessel, since you dont exactly stick scientists in a starfighter and expect them to live very long let alone kill anything hostile.

Engineers are used to man vehicles like tanks, and to man ground emplacements, ofc you dont expect a pilot or scientist to drive a tank very effectively.

As for scientits, they are mostly used as civilians, man research stations to develop better weapons/armor, and are rescued from hostage situations.

And yes, i really should stop thinking of KSP as a space sci-fi military combat simulator

[322997am]

Pilots are good for small rovers without space for a probe core.

[fireblade274]

Giving pilots a less overpowered version of MechJeb seems like a good idea to me.

YES This

[Wjolcz]

Giving pilots a less overpowered version of MechJeb seems like a good idea to me.

No auto pilot, please.

Can we all agree that Pilots should be able to predict the landing site and check CoM, CoL and CoT in-flight?