Engine balancing issues in ARM?

[Jouni]

Have some better plotted data then

I would include the 24-77 engine among the balanced engines. It's a small, high-TWR engine that's easy to attach almost anywhere. I use it a lot in landers, SSTOs, utility tugs, and other ships that need a small rocket stage capable of delivering 500-1500 m/s of delta-v.

[Scoundrel]

I shall offer my (admittedly worthless) opinion on the subject in the vain hope of steering this thread away from devolving into another pointless discussion that doesn't bring anything to the community except perhaps a sour taste.

Firstly, there's no point trying to use statistics and make comparative claims when the engines are very clearly the product of arbitrary values. Assigning arbitrary values to a system always makes the system seem arbitrary and inconsistent because, well, it's arbitrary. Thusly, arguing that arbitrary values feel arbitrary - then arguing about by how many degrees it feels arbitrary - doesn't strike me as particularly constructive. Then again, maybe it is, and I'm just not enlightened enough to see it.

Secondly, arguing over a system that exists in its own discrete reality by comparing it to real world values and systems (of which KSP functions as a poor - but fun - facsimile thereof) is also not particularly constructive. To correct that, Squad would have to create a particularly complex formula to determine mass, thrust, and fuel consumption that mimics real world rocket science, which I suspect is beyond the amount of time and effort they want to invest in "balancing" engines. Thus we have the ubiquitous "power creep" that comes in because game designers want a "feel" rather than actually having to engineer a system so that current and future parts are consistent with each other.

Lastly, we can't have engines have "roles" because ISP is done wrong - it is used as a weird form of SFC (specific fuel consumption) mostly because it is easier for us non-rocket scientist types to figure out how to adjust the design so it can go farther. If ISP were to be done properly, it would be fixed so THRUST changes with atmospheric pressure, not SFC (and yes, both would change with throttle setting). Not only that, but ISP would be a curve, because an engine's peak ISP is determined by its nozzle shape (which, of course, KSP does backwards:huh:) which is optimized for a certain altitude... and now we know why aerospikes were invented!

As it sits now with ISP acting as SFC - and since the goal of any engine is to go up - all engines get better as they race to vacuum where they operate the best. It is why we never really pay attention to atm ISP value of engines and focus exclusively on vacuum ISP figures, because atm ISP numbers get to vacuum ISP numbers real quick.

Now, if Squad changed ISP to affect thrust and put it on a curve so that some engines had decent ISP (thrust) at sea level, high ISP (thrust) at medium to high altitude, and low ISP at vacuum; and others were optimized for vacuum and simply produced less thrust in atmosphere, and so on; we would actually have engines that had roles and would only be good at their specific job (except aerospikes, which are always good at all jobs, though why no thrust vectoring Harv? Why?)... but BUT! we would pay for that by making it a little bit harder to overcome that barrier to getting things into orbit, and I think making thrust variable due to altitude would possibly make a lot of players mad enough to fill the suggestions forum with "Mak ISP be lik old SIP or I kwit forevar!!1!"

So, in closing, I humbly and respectfully plead that the discussion shift to solutions and other constructive goodness so that we may hopefully affect positive changes to the game, and leave the negativity to the electrons!

Edited April 6, 2014 by Scoundrel

[Technical Ben]

I was just trying to point out almost anybody can completely and totally unbalance a part on a whim, so trying to enforce "balance" in the game doesn't really accomplish that much.

A simulation first, a game second. Perhaps? Real life does not care about "balance" or mechanical designs. For instance, we have adjusted the mass of planets in KSP for gameplay reasons, but that's to avoid grind. The game very rarely, if ever, adds changes for cost/point/mechanical balance reasons. Only time, such as the mass/size of planets, the ion engine power/efficiency etc. Most other balances will be done on cost/availability of parts, not on their mechanical function. Why? The sim part requires some parts TO be more powerful, just as in real life new tech IS better than the old. Not all parts mind, but some will be, and the player will go "wow, new tech can be better". Sometimes it can be worse too (looks at the old ion engine).

PS, also as already mentioned. Sepratrons etc show that unbalanced gameplay IS possible. Take another real world + gameplay example. Any rocket with real power and fuel efficiency can be used providing you have the funds. We would have hundreds of space hotels already providing we had infinite funds (like Sandbox mode), as current tech and fuel and TTW and ISP are all find in real life. Same goes for KSP, it could have the most silly rocket and thrust ever, providing you had the "science" and "kerbal bucks". In Sandbox on any game, everything is OP by design, in SP/missions/career, it's balanced via other metrics.

Edited April 6, 2014 by Technical Ben

[m4v]

These charts are not intended to necessarily prove anything particular about the engines, but merely to demonstrate that there's many ways to make "math" say whatever you want it too....so don't think that just because it's math, that it's beyond dispute.

We all know that anyone can abuse maths and support whatever argument you want, but for gauging the overall performance of a rocket you need to account its Isp and TWR so stupid_christ is on the right.

Here's another factor where the size-3 ARM engines really don't bother me, ....the 25x4 needs more fuel just to move itself than the smaller ones....yes, even the mainsail.

This finding is meaningless because your plot doesn't account thrust, this is the equivalent of saying that the mainsail is better when there's no payload mass.

Include the payload mass and you will see that the Mainsail and 25x4 will tie with a payload of ~12.3t. So the 25x4 will be better for payloads of >12.3t, will the mainsail better for smaller payloads? If we had only those two rockets then yes, but in truth we have a lot of other options, so more math is needed for verify it.

My point is that the graph is not useful for get conclusions about balance.

Edited April 6, 2014 by m4v
removed an assumption by my part.

[Hodo]

And anyone who has had statistics classes knows that systematic rationale is the most important part of model specification; none of which has been provided here.

I've already said I'm not discussing it further. Especially not with a confrontational mod.

Wow, you really know how to debate something don't you? Chris just brought a counter point to your figures and you answer with that?!?! What are you 6? "I dont like you don't agree with me so I am not talking to you."

On topic,

I still haven't bothered to use those engines, they don't seem all that bad. People cried about the RAPIER, the SABRE, the nukes, and everything else. The only thing I think people should complain about is, why aren't there more people playing this!

Let us remember it is ONLY .235 not 1.235 not 23.5, it is VERSION 0.23.5, which means it is still .765 versions away from being a live complete version.

[Red Iron Crown]

Let us remember it is ONLY .235 not 1.235 not 23.5, it is VERSION 0.23.5, which means it is still .765 versions away from being a live complete version.

I sincerely hope we don't have to go through seventy odd more releases before 1.0. Good thing version numbers don't work that way.

[Ruinsage]

I feel that there is one key factor that people have forgotten in this whole debate: The new engines are not supposed to be a bigger version of the old engines, they are supposed to be heavy lifters of rockets too heavy to be lifted without asparagus designs. This, I believe, is because Squad is moving to a more realistic air resistance model in the near future, just as they have recently improved the joints. Squad is pushing for more realistic rockets, and asparagus rockets are simply not viable in the real world, so they made new rockets. In fact, in order to create a more efficient rocket in teams of delta-V per ton than the large new engine with the largest tank attached, you just need to make an orange tank, one mailsail at the bottom of that, and attach two Rockomax X200-32 Fuel Tanks with mainsails to the sides, feeding their fuel into the orange tank. This weighs almost exactly the same, and has a larger amount of delta-V. So you see, the new engines are absolutely not overpowered for what they are intended for, they are just not supposed to be larger versions of the old engines.

For my other views on this matter, ill be brief:

-The game should be balanced for career, so cost is also going to be a balancing issue when that is implemented. Until then there really isn't a reason to nerf the engines, especially not this close to the implementation of money. Saying the game should be balanced for sandbox is like saying that League of Legends should be balanced for coop-vs-ai.

-Yes, they should move it further back on the tech tree, and I assume they will soon.

-Research should provide better parts, just like how research in CIV provides better soldiers. Research providing larger but other wise identical parts is ridiculous.

-You will still need the same skills, except for asparagus building skills. Don't whine about it, just move on.

-I expect new parts of the larger diameter being made soon, hopefully something that will make station building useful.

-No, mainsails are not obsolete, they are just used for smaller payloads. This will be more apparent when costs are implemented.

Edited April 7, 2014 by Ruinsage

[Z-Man]

The models presented by Stupid_chris are not useful.

Glad someone else noticed that, too.

[shadowsutekh]

Wow, you really know how to debate something don't you? Chris just brought a counter point to your figures and you answer with that?!?! What are you 6? "I dont like you don't agree with me so I am not talking to you."

On topic,

I still haven't bothered to use those engines, they don't seem all that bad. People cried about the RAPIER, the SABRE, the nukes, and everything else. The only thing I think people should complain about is, why aren't there more people playing this!

Let us remember it is ONLY .235 not 1.235 not 23.5, it is VERSION 0.23.5, which means it is still .765 versions away from being a live complete version.

That's not how version numbers work...

Edited April 6, 2014 by shadowsutekh

[UberFuber]

As it sits now with ISP acting as SFC - and since the goal of any engine is to go up - all engines get better as they race to vacuum where they operate the best. It is why we never really pay attention to atm ISP value of engines and focus exclusively on vacuum ISP figures, because atm ISP numbers get to vacuum ISP numbers real quick.

Now, if Squad changed ISP to affect thrust and put it on a curve so that some engines had decent ISP (thrust) at sea level, high ISP (thrust) at medium to high altitude, and low ISP at vacuum; and others were optimized for vacuum and simply produced less thrust in atmosphere, and so on; we would actually have engines that had roles and would only be good at their specific job (except aerospikes, which are always good at all jobs, though why no thrust vectoring Harv? Why?)... but BUT! we would pay for that by making it a little bit harder to overcome that barrier to getting things into orbit, and I think making thrust variable due to altitude would possibly make a lot of players mad enough to fill the suggestions forum with "Mak ISP be lik old SIP or I kwit forevar!!1!"

As you stated, variable thrust can be a bit too complicated, we can focus on fuel consumption behavior.

Since it's based off fuel consumption, we can still have engine that has high ISP and sea level and low ISP at vaccum. Basically have engine "consume" more fuel as the air gets thinner.

Think of it as the Kerbal engines are designed to have controller that maintains a constant thrust, but will consume more and more fuel to "reach" that thrust.

[Scoundrel]

As you stated, variable thrust can be a bit too complicated, we can focus on fuel consumption behavior.

Since it's based off fuel consumption, we can still have engine that has high ISP and sea level and low ISP at vaccum. Basically have engine "consume" more fuel as the air gets thinner.

Think of it as the Kerbal engines are designed to have controller that maintains a constant thrust, but will consume more and more fuel to "reach" that thrust.

It is a good suggestion for a cognitive rationalization for variable SFC, as I too once thought as you did. However, once I started doing some experiments with engines (I'm slowly working on a parts pack for landers) to create a lift/boost atm optimized engine, an optimized upper stage engine, and a vacuum optimized engine, I discovered that the vacuum optimized engine was the best engine out of all of them even when I reduced its atm Isp to 200, as by 5000m they were as good as the atm engines, which I had set as 388 atm Isp and 250 vac Isp.

Worse, I discovered that there wasn't a peak altitude performance number, which meant that no matter the engine, the one with the higher vac Isp would always be the superior lift engine as long as the gross mass:thrust ratio remained close. Needless to say I was incredibly disappointed that my idea for an Eve lander engine was such a failure, and have since had to rethink my mod concept, simply because Isp isn't Isp.

But you shouldn't take my word for it. Try copying the Squad engine folder and mess around with the engine config files, and you'll see what I mean. It seems to me that only by marrying thrust to atm pressure rather than fuel consumption and having a peak atmospheric pressure number that we can have engines for various roles besides size.