How do you justify part selection?

[TimePeriod]

Hello guys, TP here.

Before you read on you have to understand a few things about the way I play this game. I have 3 rules which are cast onto everything I do:

1. The costs have to be kept down to absolute minimal.
2. Part count has the same policy  - minimal.
3. Weight must be pushed to the absolute minimal.

These rules are "liquid" so to speak, they can give sway to each other if the benefit from breaking them supports the final goal or target.

You might quickly draw the conclusion that I am not having a lot of fun, that'd be where you are very wrong. I love to challenge myself with minimalistic builds that looks absolutely appalling but from a functional point of view are masterpieces. Functionality is the name of my game and as such I often find myself having severe headaches whenever I search "Spacecraft Exchange" for the above reasons. Yes Rune, I am looking at you, don't even get me started on Whackjob...

 

With that out of the way I'll get straight to the topic at hand, how do you justify the way you build?

Let me give the example which started this whole nonsense

 

So inside my mind I keep tally of the various statics I see when I build these two objects, with A(left) and B(right):

A Has 1 more part but weighs 1.3 tons less then B, costs 140$ less then B and is shorter then B.

B Has one less part then A, is a lot more rigid, higher crash tolerances and is a lot more appealing.   

Do I choose option A or B? How do I justify spending more cash, more weight, more size(space/volume)? Do I even have to? Will the Kraken be more likely to eat A or B?

 

Here is another example I also conjured up:

This is a question regarding functionality again, about costs savings, weight and part count. A is on the left and B is on the right.

A has the advantage in terms of weight but looses to cost, part count and functionality. However it has no need for a pilot to remain in the seat and can be freely sacrificed if needed. No lives are lost.

B wins on costs, same performance as A, costs are lower and has a lower part count. However B looses out on weight, requires a pilot (default use here, not as storage) to be operational and has an overall larger volume. 

 

Okay, I think you get the jest here. I have my reasons to select the various parts I use but this is by no means a definitive and absolute way of playing a videogame. Everybody has their own options and decisions, which all are valid and equal.

My question to you is this: How do you justify the things you decide to build? Is there some kind of deeper logic behind how you make your decisions?

(This is not a showcase thread, please don't post 12 pictures of your absolute best works of all time ever. Just the very essence of what makes you decide on option A-B or C)

[Gaarst]

Take two random assemblies A and B.

If A looks better, then use A.
If B looks better, then use B.
If none look (or looks?) better, then grab an assembly C and restart.

Also, if part count gets over 150 start to watch it, I don't like slideshows.

[FancyMouse]

I don't care about cost at all.

Part-count and weight - I optimize for weight most of the time. It's especially true for payload as that will dramatically change your rocket. When I want to count parts, physicsless parts I count them half.

When hard to choose - then I consider none of them and then see which one looks better.

[Geschosskopf]

To fully understand what you're saying, you need to tell us what you picked in your examples above :).

But the way I do things, ship design is a function of mission design.  No point in building something if it's not for a purpose.  So first I draw up the overall mission in terms of general objectives.  Then I break the objectives down into specific tasks.  These tasks all require certain payloads to perform them, so now I have a list of payloads I need.  There's no getting around these payloads if I want to do my objectives, and there's usually no real way to reduce the size or cost of them, either, due to the required level of functionality requring certain specific parts.  :Like there's only 1 choice (in stock) if you want to take Mystery Goo, etc.  Next step is to decide if these payloads need to be in separate ships or if it's better to combine some or all of them in the same ship.  At the end of this process, I have the main components for the whole mission laid out.  Then it's just a question of getting them to the target, which basicaly means building transfer stages and lifters for however many separate ships I ended up with.  I go for the simplest, chepest rockets that will do the job.

[Randazzo]

18 minutes ago, TimePeriod said:

With that out of the way I'll get straight to the topic at hand, how do you justify the way you build?

That's part of the beauty of KSP. You don't have to justify anything.

I really only have three major considerations:

1. Can it complete the contract I'm building it for?
2. Can I launch it without massive headaches?
3. Assuming 2 is "Yes", then: Can I afford to launch it?

Beyond that I do whatever strikes my fancy at the moment.

[Hotaru]

I certainly wouldn't come to the conclusion you're not having fun: I enjoy the challenge of optimizing my designs too. In general my priorities are functionality first, cost second, part count third (with the order of the last two flipped once the part count starts going much over 100). Looks are a relatively distant fourth, but I don't ignore them completely: all things being equal I'll tend to pick the option that looks the most elegant.

One of the reasons I enjoy career mode is that it gives me a valid gameplay reason to optimize for cost. In sandbox I never feel compelled to optimize for anything (except maybe looks), and I find the lack of a gameplay challenge a little boring.

Have to disagree about getting headaches from Spacecraft Exchange, however. Although on the whole I tend to prefer "form over function," which often results in fairly bland, utilitarian designs, I do like my ships to have nice lines if at all possible. I think a lot of the designs I've seen posted here are extremely cool, even if they don't share my personal design priorities.

[Stratzenblitz75]

These days, I'm almost always optimizing for part-count. If option A is more efficient and costs less, but option B has significantly fewer parts, then I'll go for option B. Its the only way I can prevent the game from becoming completely unplayable.

Beyond that, I usually choose function over form. I don't care how it looks, I just care that it works.

[Francois424]

6 minutes ago, Stratzenblitz75 said:

These days, I'm almost always optimizing for part-count. If option A is more efficient and costs less, but option B has significantly fewer parts, then I'll go for option B. Its the only way I can prevent the game from becoming completely unplayable.

Beyond that, I usually choose function over form. I don't care how it looks, I just care that it works.

^^

That Really is the real thing.  Eversince 0.90, where I can't dip over 250 parts or go in yellow, or reach 350-400 and spend 3 seconds for every one second in-game...  I try to limit myself to 250 parts.
It really thought me a lot about optimizing my designs and about the importance of "TweakScale" (Mod).  
I also find myself editing new engines sometimes (made-up: say I need 30 engines for a total of 1800 thrust and 90 tons, I'll edit a mainsail to be 90t, 1800 thrust, and keep same ISP and total heat generation... BOOM saved 29 parts !)

But I'll usually sacrifice a bit of part counts for lights, struts, additional cargo bays/RTGs, and looks.

[steve_v]

31 minutes ago, Stratzenblitz75 said:

These days, I'm almost always optimizing for part-count.

Yeah, this. Part count first, efficiency second and aesthetics a distant third. Also all docking performed on the day side because lights add too many parts.
Since 1.0 hit it's been Kerbal part-count-reduction program.

[GoSlash27]

TimePeriod,

 I build my stuff like it's intended to be used in career, so I don't spare any simplicity, reliability, or economy for "cool factor". That makes my stuff about as exciting as a maytag washer.

Best,

-Slashy

[Geschosskopf]

38 minutes ago, GoSlash27 said:

 I build my stuff like it's intended to be used in career, so I don't spare any simplicity, reliability, or economy for "cool factor". That makes my stuff about as exciting as a maytag washer.

Except your stuff still has a certain je n'est sais quoi that comes from really thinking about it and keeping it true to your idiom.  You make art.

I, OTOH, rely on brute force and ignorance.  MY stuff has all the attractiveness of a 32-pound maul, but I like 32-pound mauls so I find them attractive, and any problem that can't be solved with surgical application of high explosive can usually be finished off with a 32-pound maul

 

[r_rolo1]

Well, regarding the OP ... always, ALWAYS the lighter option needed for the mission profile, that might not be the ligher option  Say, on your second example, if the objective is simply to send a ship to X, A is clearly my favourite option, but if part of a mission to get science, B is obviously better ( you can do always more science of a crewed mission  ) .Cost is a mostly irrelevant issue for payload ( the cost of the mission, though, might not be  )

That said, on a somewhat related issue, what is the deal with Vectors and Mammoths prices? I could buy Mammoths, hack them to take the Vectors out and make a hefty profit ... if the game allowed that

Edited January 6, 2016 by r_rolo1

[TimePeriod]

54 minutes ago, GoSlash27 said:

TimePeriod,

 I build my stuff like it's intended to be used in career, so I don't spare any simplicity, reliability, or economy for "cool factor". That makes my stuff about as exciting as a maytag washer.

Best,

-Slashy

If I can take 1 part, make it serve 3 functions but it looks total garbage I will use that part.

In my eyes, its also about making the most of less.

[MagicFireCaster]

Mostly i build my crafts trying to balance between cost and functionality, since most contracts need to be done so i can get more funds than i spend is usally common for me to do it the cheaper way, depending on the task, i choose to send either probes or kerbals, also i always pack extra resources, sometimes they help to do more things, or just to compensate errors and so.

[Bill Zarr]

Part count, cost, weight, aesthetics are all considerations. Strangely enough when I think about it aesthetics is probably the overriding consideration. I'll generally add a few parts here and there to make things look vaguely more realistic. ie if kerbals are meant to hypothetically move through it I'd probably use a proper hub rather than slapping docking ports on a fuel tank. I tend towards capsules not lawn chairs. I'll use things that are less than optimum, ie nose cones even when they're of dubious benefit. My design criteria are numerous, variable and likely to change on a whim. I suppose if one examined most of the things I've built one could probably get an idea of what sort of balance I find between things.

 

[Gojira1000]

My criteria usually results in big, terrible to fly and amusing in some way. I think I design like a drunken Victorian steam-fitter.

[Temstar]

I optimise for:
 

1. Meet mission objective - else no point going
2. Low part count - if the ship can't be simulated, the mission can't be done
3. Overall low cost - something to aim for. Although I take a very long term view on cost so I'm bias towards reusability at the expense of upfront cost, within reason

There's one other condition that's difficult to place in the above list - plausibility. This has to do with mapping between the perfect simulation of KSP and real life. On one extreme of the scale is space shuttle replica in KSP. We all know it's hard to design, hard to fly and not very practical solution for reusability in KSP, yet you could take up the position that it's a worth while endeavor to do in KSP because it was done in real life.

On the other extreme, you could send a single Kerbal to land on Duna and back with the entire trip done ridding in a command chair on the side of the rocket. It's perfectly valid in KSP but you wouldn't be able to do that in real life going to Mars, some people would find that acceptable.

Most people including myself probably sits somewhere between these two extremes. It's a fine line to walk though and has the power to override any of the above requirement depending on how far a certain solution crosses this line.

[Norcalplanner]

I named my design philosophy - it's "Cheap and Cheerful". Click the link in my signature for my guide, which uses many, many words to fully describe it.

[Xavven]

12 hours ago, TimePeriod said:

A has the advantage in terms of weight but looses to cost, part count and functionality. However it has no need for a pilot to remain in the seat and can be freely sacrificed if needed. No lives are lost.

B wins on costs, same performance as A, costs are lower and has a lower part count. However B looses out on weight, requires a pilot (default use here, not as storage) to be operational and has an overall larger volume. 

Hey, TimePeriod! In your second example, I'm guessing you're designing a payload. Optimizing for weight, not cost, on the payload will make your first stage much, much smaller, and thus much, much less expensive and most times you'll have less parts overall too. It's penny-wise but pound-foolish to save $360 on your uppermost stage, only to pay $10,000 more to get it to orbit because it weighs more.

Anyway, put me in the function over form category. My rockets have no style but accomplish the mission efficiently.

[Mikki]

You can build cheap and cheerfull, nature works the same way. I use parts as they come in stock, i refuse to clip, at least i clip very minor amounts of fuselages for aerodynamic purpose, somewhat 5% of the volume max.

[Malich]

I'm pretty much with most of the above. I need a craft that can do the job and most of the rest is secondary. I have a few rules I impose on myself such as "Each Kerbal on a craft going out for more than 24 days needs at least 2 seats worth of space". Generally i don't find cost restrictions to be a major factor after the early stages of a career game though early on I never launch a mission unless I can get at least 150% of the launch costs back from completion fees. 

This often leads to craft that are anything less than optimal. I often launch ships with huge amounts of reserve crew space as I plan to detour into orbit on the way out / back to collect a gaggle of Kerbanauts who've managed to get themselves stranded or with unnecessary parts (docking ports usually) added on because i have a station or part test mission do do that will let me gain some extra funds for the mission I would not have gotten funded otherwise. Recovering 3 Kerbals from orbit and landing a new "station" on Minmus will usually cover the launch costs for a slightly over engineered ship that will complete the missions and gather me a bucket load of science rather than having to launch 3 1-Kerbal recover vessels as each contract appears and a 4th craft for the station / science part.

[Blackhog]

Well, I am a really new player, so some of my designs had quite a bit of extra stuff in case I screw up, but overall I'd say I value simplicity and aesthetics above all else,

Simplicity includes both mass and part count so my list would be:

1. Mass & efficiency 
2. Aesthetics
3. Part count

Although they are really close.

One of my recent "Challenge" was to land on Minmus.

I revised my ship like 5 times before I was satisfied with it, constantly cutting stuff off.

I kinda have the soviet mentality in designing rockets.

 

[rocketbuilder]

I try to base my builds around aesthetics and will go a few funds futher to make it look nice, but mass usually wins out. A a, B b because it also can carry a kerbal.

[Just Jim]

On ‎1‎/‎6‎/‎2016 at 1:20 PM, Stratzenblitz75 said:

These days, I'm almost always optimizing for part-count. If option A is more efficient and costs less, but option B has significantly fewer parts, then I'll go for option B. Its the only way I can prevent the game from becoming completely unplayable.

Beyond that, I usually choose function over form. I don't care how it looks, I just care that it works.

 

Unfortunately, I have to do the same.  I love aesthetics, and if it's a smaller vehicle I'll go all out crazy making it look cool.
But I have a somewhat older and slower computer, and if I go over 200-300 parts it starts to complain.
So larger things like space stations, I have to balance coolness with part counts, whether I like it or not.... 

[monstah]

In early career, funds are my main constraint, but unfortunately, like some said above, once I get the money I design for part count. I play on a laptop, and it suffers when I get around 300-400.