Making kerbals matter, without letting them fly

[LethalDose]

Given that the game has named kerbals with statistics (even though they are meaningless, at this point), it seems clear that the developers intend there to be meaningful choices to make about which kerbals are assigned to a mission.

However, I personally believe it would be a mistake to have the kerbals assigned to a mission affect how a craft flies. I feel like this would interfere with and have a detrimental effect on game play. Therefore, a system should be in place to allow kerbals to play an important part of the mission without allowing them to affect craft performance. I've outlined my initial thoughts on such a system, which includes attributes, jobs, and gear for our little green minions. Some of these ideas have been touched on in other threads, but tangentially, and I think there's room for this in the game.

Kerbal Attributes

Overview

Attributes represent a kerbal's aptitude and experience in various skills needed for space flight. Experience for each attribute is tracked separately for each kerbal, and gained by performing related tasks (indicated below). Attribute experience may be spent to improve job ranks (see below), or potentially increase the attribute value, depending on the implementation. The attributes below are not intended to be a complete listing of every possible attribute, but I don't think many more should be included in the game (total of not more than 5 or 6).

Dexterity

This attribute represents a kerbal's skill and experience controlling a craft*, and the primary attribute needed for a pilot. A kerbal would gain dexterity experience by spending time in a craft's pilot seat. Experience gained would be proportionate to the craft's mass and thrust experienced while piloting the craft, or training at flight ground school (either a KSC building, or an abstract location off-site location on Kerbin).

*I made it clear that the kerbal shouldn't affect craft handling, While the kerbal would gain experience while in the pilot's seat, the affect of a skilled pilot is explained below and does not affect craft handling.

Leadership

This attribute represents a kerbal's skill and experience commanding a crew of kerbals in space, and the primary attribute needed for a mission commander. A kerbal would gain leadership experience by being embarked on craft with multiple kerbals, and proportionate to the time and number of kerbals on the ship, or training at officer training school (OTS).

Intelligence

This attribute represents a kerbal's knowledge of physical and life sciences and is the primary attribute needed for scientists. A kerbal would gain intelligence experience by performing science in the field or training at Kerbal University and experience gained would be proportionate to the value of the science activity performed.

Endurance

This attribute represents a kerbal's physical endurance. This attribute isn't needed for any particular job, but benefits all kerbals. This attribute is gained during EVA and survival training. This attribute allows kerbals to perform longer EVAs and better resist injury*.

*I'm not clear if the devs have said absolute no to injury, or if it's something they not implementing currently.

Kerbal Mission Roles/"Jobs"

Overview

Depending on the implementation, Kerbals can either earn multiple job ratings, or be qualify to fulfill certain mission roles. Each job or role has multiple levels, junior pilot, pilot, senior pilot. Each additional level of a job/role makes the kerbal more useful.

Pilot

Pilots are one of the fundamental roles for kerbals in KSP. Better pilots are able to handle larger and more complex craft, and the size of the craft being launched is limited by the skill of the pilot as follows:

• No pilot rank is required to launch craft up to 25 tons
  
• Rank 1 pilots can launch craft up to 50 tons
  
• Rank 2 pilots can launch craft up to 100 tons
  
• Rank 3 pilots can launch craft up to 200 tons
  

Under an alternate or combined implementation, having a kerbal in a craft's pilot seat "active" would allow that craft to maneuver.

Scientist

Scientists of increasing skill ranks are able to perform science experiments of increasing complexity. Some examples are listed below:

• No Scientist rank is required to observe goo
  
• Rank 1 scientists can record temperature and barometer readings
  
• Rank 2 scientists can record seismic data and material bay results
  
• Rank 3 scientists can record gravimetric data
  

Under an alternate or combined implementation, having a scientist active in a craft equipped with science instruments would allow data recording.

Mission Commander

The exact benefits provided by a mission commander would vary by implentation. A mission commander could allow launching larger crews, allow multiple kerbals to be "active" simultaneously, or allow multiple kerbals to provide passive benefits to a ship/mission.

Gear & Equipment

Overview

Equipment increases the capabilities of kerbals in and out of the ship. In this concept, kerbal gear is divided into two categories: suits and tools. A kerbal's load-out would be determined at launch in the launch menu, or default gear loadouts can be saved in the kerbonaut facility.

Suits

Suits represent the gear that is worn by kerbals to protect them from the hostile environments they encounter. There would be several types of suits that would be developed during play. This is an idea that has been mentioned in other posts. Additionally, different suits would have varying numbers of slots available to equip tools. A ship part, the airlock, would allow crew to change suits during a mission. Some suit examples would include:

• Flight suit: the basic starting suit, has no tool slots, and limited EVA capacity
  
• Utility suit: moderate EVA and several tool slots
  
• Soft suit: extended EVA time, but only one tool slot.
  

I also have ideas for support times, but thats probably for a different post.

Tools

Tools increase the capabilities of kerbals to accomplish particular missions. They equiped in suit gear slots. Gear would be assigned on the ground (see above). Gear can be swapped out during EVA using a radially mounted gear locker, or in the ship in the gear bay. There's a lot of variation in the these tools, so its easier to give examples:

[*]Transceiver: This equipment allows a kerbal to transmit EVA reports directly back to KSC

[*]Reserve Support Unit: Increases suit support duration

[*]Repair kit: Allows kerbals to repair broken ship parts

[*]Advanced kerbal maneuvering unit (KMU): More monopropellant for EVAs

[*]Spectrometer: A science tool that measures atmospheric components. Requires Scientist lvl 2 to operate.

[*]Surface sample kit: allows collection of surface samples.

Anyway, there's a lot here, and a lot of options, but the point remains: The kerbals can be useful and individual, but not affect how the craft flies at all. It's a separate system that enhances gameplay options and choices. I'm curious to hear thoughts, and hope it gets some dev attention.

[thereaverofdarkness]

I'd like to see the scientific instruments not have a required rank to use, but instead would yield less science at the hands of someone inexperienced. The lowest level things could give high yields at almost any science level, while the highest level tools would require a high level for good scientific yield:

A.) Crew Report (science level, leadership level)

0/0: 60% - 0/1: 65% - 0/2: 70% - 0/3: 75%

1/0: 70% - 1/1: 75% - 1/2: 80% - 1/3: 85%

2/0: 80% - 2/1: 85% - 2/2: 90% - 2/3: 95%

3/0: 90% - 3/1: 95% - 3/2: 100% - 3/3: 100%

B.) Temperature (science level)

0: 60%

1: 85%

2: 95%

3: 100%

C.) Seismic Data (science level)

0: 20%

1: 45%

2: 75%

3: 100%

[Superfluous J]

Great ideas, though I'd go for a more smooth approach. Instead of having a pilot of level 3 able to pilot 200 tons, make them start around level, say, 20 and have that be able to pilot 20 tons. If he does that for a few minutes, his skill goes up to 21, then 22, etc. So no 200-ton craft until you've got a pilot of rank 200.

And the whole idea may need some thought when it comes to docking. You dock two 150 ton crafts together and then realize your best pilot is skill level 200. Uhhhh...

[LethalDose]

Great ideas, though I'd go for a more smooth approach. Instead of having a pilot of level 3 able to pilot 200 tons, make them start around level, say, 20 and have that be able to pilot 20 tons. If he does that for a few minutes, his skill goes up to 21, then 22, etc. So no 200-ton craft until you've got a pilot of rank 200.

And the whole idea may need some thought when it comes to docking. You dock two 150 ton crafts together and then realize your best pilot is skill level 200. Uhhhh...

I see where you're coming from, but I think your confusing attributes and job ranks. I'm willing to have attributes tracked on a fine scale (e.g. 1-100, though I wouldn't prefer it), but the job ranks should have a small number of levels so reduce complexity. Also, I'm really opposed to returns based on skill level for the same reasons that I'm opposed to returns based on experiment time: it gets even more 'gamey' because players begin to worry about optimization instead of achievement. I've found that a tier system with a small number of tiers is almost universally superior to more complex systems in game design because its easier for players to understand and more difficult to abuse.

Also, docking is already accounted for: The original post states that the launch weight is what is limited by pilot skill. After launch, no particular amount of skill is needed to pilot a craft. This is important because you may need to swap kerbals between craft midflight, or a pilot could be injured, stranding the entire crew. It also makes the weight check very simple to code and understand: the check only needs to be made once, and there's no confusion about whether the weight of fuel counts against the weight limit (in this case, it does). And theres no chance of a kerbal in the pilot seat becoming ineligible to pilot a craft.

Besides, the mass of many craft after launch is a small fraction of what it's launch weight, due to fuel expenditure and staging. And, practically, I don't think what gets to space is ever more than 20% of launch weight.; I don't think anyone in the payload % competition beat 20%. So there would have to be a LOT of docking going on before it really became a problem. I acknowledge that very extensive docking does happen, but I wouldn't qualify it as "frequent", and not limiting post-launch piloting avoids the issue, even in uncommon cases.

Thanks for your feedback.

[iMAniaC]

I like the general direction of these ideas. However, I think it would be more fun and "Kerbaly" to invert them:

Not Dexterity, but Bravery:

How much explosive power do you dare to light up beneath your seat? At first, maybe, say 20 tons, but when you've done that once, it's not as terrifying the next time around, so with the same amount of Bravery, you'll dare to light an even bigger rocket on fire the next time. Bravery could then be a static attribute, but you'd still have to "level up" the Kerbal to have him pilot bigger rockets, even though the Bravery attribute stays the same.

Not Intelligence, but Stupidity:

Stupid people learn from their mistakes. Smart people learn from other people's mistakes. In order to maximize learning, we need lots of stupid people to make mistakes to learn from. Stupidity is the key to progress! Thus, having a high Stupidity stat is good for progress.

Not Leaderhsip, but Confidence:

It's not the smartest who lead, nor the bravest. It's not even the one who is best fit to lead overall, but simply the one who manages to give orders most convincingly and with most confidence. Successful missions will give a confidence boost.

However, there's a very fine line between fun leveling and grinding stats. If you have to spend a lot of time grinding kerbonaut stats instead of sending rockets to new places, the game will probably lose some of its appeal, I think. Perhaps the influence to the Bravery stat could be global, so that when other Kerbals see one Kerbal survive, say 20 tons, that becomes slightly less scary to all of them. That way, you won't have to go back to basic rockets if you happen to strand Jeb on Mun or something...

Science experiments are also hard to get right. I think the way they are now, with diminishing returns, are inherently un-funny. However, I like that there is something to be learnt from doing the same experiment twice and I like that there is more to be learnt from bringing the samples back to Kerbin. Adding a kerbonaut's Stupidity into the equation makes for a really complicated way to figure out how much science is to be gained from a single experiment. That being said, I don't mean that Stupidity absolutely shouldn't count towards science, but it might be really hard to pull off in a good way.

In the end though, since the developers have already announced that there will be a kerbonaut program, I'm sure they have some good ideas of their own. I'm looking forward to seeing what they come up with.

[LethalDose]

... Wow.

I guess I have to take this apart, piece by piece, or let it stand and have it look like I agree with anything that got posted in here...

First, lets remember the point of this thread: Kerbals can matter without affecting flight characteristics. Nothing in this response came anywhere close to addressing this point.

It did, however, manage to vomit up a lot of terrible ideas. /sigh Lets get started.

I like the general direction of these ideas. However, I think it would be more fun and "Kerbaly" to invert them:

You only inverted one, which makes one improve by going down, and two improve by going up? As far as making them more "Kerbally"...

1. The game hasn't even been released yet, so who knows what the final flavor is going to be
   
2. This game has us calculating TWR's and using transfer orbits... is it so hard to expect reasonable names for attributes?
   
3. Watch this if you want an explanation of why decisions about game design based on "brand feel" lead to real disasters...
   

Not Dexterity, but Bravery:

How much explosive power do you dare to light up beneath your seat? At first, maybe, say 20 tons, but when you've done that once, it's not as terrifying the next time around, so with the same amount of Bravery, you'll dare to light an even bigger rocket on fire the next time. Bravery could then be a static attribute, but you'd still have to "level up" the Kerbal to have him pilot bigger rockets, even though the Bravery attribute stays the same.

Static attributes are terrible ideas. Period. Players will just zoom past some kerbals on recruitment if they see the candidates have poor static stats, again falling into "optimization" traps. Beyond this, the only think that was changed in this "suggestion" is regurgitating the original post with a name change.

Not Intelligence, but Stupidity:

Stupid people learn from their mistakes. Smart people learn from other people's mistakes. In order to maximize learning, we need lots of stupid people to make mistakes to learn from. Stupidity is the key to progress! Thus, having a high Stupidity stat is good for progress.

This is a game with planned educational component. Please don't tell me that I need to explain why "Stupidity" shouldn't be an appealing attribute... The fact that it's obviously a place-holder name is the only reason that this hasn't been an issue.

Also, <Insert joke about quality of the American education system here>

Also, yes, I know it's being developed in Mexico.

Not Leaderhsip, but Confidence:

It's not the smartest who lead, nor the bravest. It's not even the one who is best fit to lead overall, but simply the one who manages to give orders most convincingly and with most confidence. Successful missions will give a confidence boost.

I'm sorry you didn't like the specific word I used to present design concepts?

However, there's a very fine line between fun leveling and grinding stats. If you have to spend a lot of time grinding kerbonaut stats instead of sending rockets to new places, the game will probably lose some of its appeal, I think. Perhaps the influence to the Bravery stat could be global, so that when other Kerbals see one Kerbal survive, say 20 tons, that becomes slightly less scary to all of them. That way, you won't have to go back to basic rockets if you happen to strand Jeb on Mun or something...

Holy... where to start?

1. What makes systems "grindy" is like 5% system structure and 95% system balance. Saying the system could be implemented poorly and therefore isn't worthwhile is facile, at best.
   
2. I would have been nice if you had paid attention to the original post, in which I laid out how to improve every attribute listed without launching missions. In this case, ground school.
   
3. Stranding Kerbals is bad. As you point out later, the devs have a kerbonaut program later. If there's no penalty for stranding Jeb on the Mun, then kerbals have no game value, and then what's the point of kerbonaut program? And if you're playing smart, you wouldn't only have one competent kerbonaut capable of getting the Mun in the first place. At best, you've only demonstrated how this system would encourage wise play.
   
4. Global stats only act to homogenize the individual kerbals, and then, again, whats the point of a kerbonaut program to develop individuals if, individually, they're all the same.
   

Science experiments are also hard to get right. I think the way they are now, with diminishing returns, are inherently un-funny.

I'm sorry you find the current implementation "un-funny"? This is completely irrelevant since it's clear changes are coming. Hopefully the next implementation will make you giggle and clap your hands. Frankly, I don't see how the system's humor really relates to good game design, and frankly, I'm concerned that you think it matters.

Also, see the first comment regarding how poor implementation/balance does not equal systemic flaw.

However, I like that there is something to be learnt from doing the same experiment twice and I like that there is more to be learnt from bringing the samples back to Kerbin. Adding a kerbonaut's Stupidity into the equation makes for a really complicated way to figure out how much science is to be gained from a single experiment.

Complexity without enrichment is bad. It shouldn't be difficult for players to figure out how much science and experiment will provide.

That being said, I don't mean that Stupidity absolutely shouldn't count towards science, but it might be really hard to pull off in a good way.

Then you back-pedal into a different pothole. Stupidity absolutely shouldn't count towards science. I addressed why I'm against why intelligence should not affect science rewards in the next post.

In the end though, since the developers have already announced that there will be a kerbonaut program, I'm sure they have some good ideas of their own. I'm looking forward to seeing what they come up with.

Yes, they have a plan.

Hopefully, it doesn't involve kerbals affecting flight characteristics. Again, the whole point of this post was to demonstrate aspects of method that accomplishes that goal.

Besides, just because "they have some good ides of their own" doesn't mean we can't discuss discuss or recommend implementations of parts yet to come. If it did, this thread really wouldn't have much point...

Finally, sorry if you don't like the tone of my response. As a graduate student and researcher, I have to take apart arguments and suggestions like this on a daily basis, and i've learned that trying to sugar-coat why bad ideas are bad ideas when there are concrete facts involved doesn't do anything more than waste everyone's time and effort.

Actually, my bad, I thought I responded to this already.

This is also a bad idea:

I'd like to see the scientific instruments not have a required rank to use, but instead would yield less science at the hands of someone inexperienced. The lowest level things could give high yields at almost any science level, while the highest level tools would require a high level for good scientific yield:

A.) Crew Report (science level, leadership level)

0/0: 60% - 0/1: 65% - 0/2: 70% - 0/3: 75%

1/0: 70% - 1/1: 75% - 1/2: 80% - 1/3: 85%

2/0: 80% - 2/1: 85% - 2/2: 90% - 2/3: 95%

3/0: 90% - 3/1: 95% - 3/2: 100% - 3/3: 100%

B.) Temperature (science level)

0: 60%

1: 85%

2: 95%

3: 100%

C.) Seismic Data (science level)

0: 20%

1: 45%

2: 75%

3: 100%

Science produced by experiments should NOT be affected by the kerbal performing the experiment. This kind of design suffers from the same pitfalls as finely incremented stats that have direct play affects. Primarily, they actually encourage grindy play styles because players will max out one stat on one a small number of kerbals, then use that kerbal for all science experiments. With this kind of system, there is little to no incentive to train up multiple kerbals, and the loss of a single "optimized" kerbal in this system is a bigger penalty than the loss of an "optimized" kerbal in a system with job ranks that don't affect science production.

Edited December 2, 2013 by vexx32
Merged posts.