Engine balancing issues in ARM?

[Red Iron Crown]

I don't object the use of graphs in general or the general principle that high TWR should go with low I_sp, at least for engines of similar mass and size.

What I object to is the use of a power law for it. That is simply factually wrong.

How is it factually wrong? The curve he uses seems to fit the balance of the older stock engines.

[Kasuha]

I did my own analysis of new engines and my conclusion is that the only single engine which is seriously out of balance is LV-N. And I would hate if that one was nerfed because interplanetary travel would become way less fun. Adjusting other engines to match it is also out of question.

In general, not just "more TWR -> less Isp" but more importantly "bigger engine -> better TWR & better Isp" apply. And new engines follow that rule nicely. When I tried to put them in line by adjusting their Isp according to their mass+thrust, I ended up with actually increasing Isp of S3-KS 25x4 to 374, for instance.

No, I don't think the new engines are unbalanced. They're fine.

[Red Iron Crown]

I did my own analysis of new engines and my conclusion is that the only single engine which is seriously out of balance is LV-N. And I would hate if that one was nerfed because interplanetary travel would become way less fun. Adjusting other engines to match it is also out of question.

In general, not just "more TWR -> less Isp" but more importantly "bigger engine -> better TWR & better Isp" apply. And new engines follow that rule nicely. When I tried to put them in line by adjusting their Isp according to their mass+thrust, I ended up with actually increasing Isp of S3-KS 25x4 to 374, for instance.

No, I don't think the new engines are unbalanced. They're fine.

Wouldn't the 48-7S be woefully unbalanced by those metrics too? It is a small engine with very high TWR and decent Isp, which goes against your balance criteria if I understand them correctly.

Is your analysis in a format that is sharable? I'd like to see how you took engine size into account.

[stupid_chris]

I don't object the use of graphs in general or the general principle that high TWR should go with low I_sp, at least for engines of similar mass and size.

What I object to is the use of a power law for it. That is simply factually wrong.

It's not a power law. It's an inverse function.

I did my own analysis of new engines and my conclusion is that the only single engine which is seriously out of balance is LV-N. And I would hate if that one was nerfed because interplanetary travel would become way less fun. Adjusting other engines to match it is also out of question.

In general, not just "more TWR -> less Isp" but more importantly "bigger engine -> better TWR & better Isp" apply. And new engines follow that rule nicely. When I tried to put them in line by adjusting their Isp according to their mass+thrust, I ended up with actually increasing Isp of S3-KS 25x4 to 374, for instance.

No, I don't think the new engines are unbalanced. They're fine.

If by mass+thrust you mean mass added to thrust, this is pretty wrong on the numbers side, as this number is not relevant of the performance of the engines

And again, the size of the engine is not relevant.

When comparing the balance of engine, what you need to take into account is it's specific impulse and it's TWR. The size, mass, thrust, yadda yadda, and the other stuff, is no relevant.

In this case, I'm pretty sure that I made no error making that graph, so I'm highly suspicious that you could get different results from the same data field.

I've even taken the time to remove that abbering data. The LV-N is not a chemical engine, so it should not be on this curve. The RAPIER is two engines in one, and therefore does have a much lower TWR than it should. Miniature engines also do not follow this curve, which englobe the LV-1, LV-1R, 24-77, 48-7s. And finally, I don't need to debate to say that the MK. 55 is terribly unbalanced, and that including it in the calculations would simply put everything off by a significant factor.

I'll map the KW Rocketry and NovaPunch engines to this graph to show how the curve is also followed by them.

Edited April 4, 2014 by stupid_chris

[The Error]

Guys. This is a *single player* game. Balance seems irrelevant... anyone can just edit .cfg files or mod in whatever they wish. So I feel like "balance" discussions are...well... a tad pointless.

Just saying.

[Red Iron Crown]

Guys. This is a *single player* game. Balance seems irrelevant... anyone can just edit .cfg files or mod in whatever they wish. So I feel like "balance" discussions are...well... a tad pointless.

Just saying.

It is and it isn't. The game itself is single player (for now), but people like sharing their craft, participating in challenges, helping each other with their designs, all of which become more difficult with a bunch of custom .cfgs.

[The Error]

It is and it isn't. The game itself is single player (for now), but people like sharing their craft, participating in challenges, helping each other with their designs, all of which become more difficult with a bunch of custom .cfgs.

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.

[Red Iron Crown]

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.

I think having parts balanced in the stock game is a desirable goal in and of itself. Unbalanced parts belong in mods, IMO. One should have to edit .cgfs to make parts unbalanced, not balanced.

[The Error]

Balancing stock parts is fine, but really, enforcing said balance is pointless.

Balance your stock as much as you feel like, just don't expect that to hold, is what I mean.

[Kasuha]

If by mass+thrust you mean mass added to thrust, this is pretty wrong on the numbers side, as this number is not relevant of the performance of the engines

And again, the size of the engine is not relevant.

By "mass + thrust" I meant mass combined with thrust. Or rather, we need to use 3D metric, not just 2D metric. Because by combining mass and thrust into TWR you're losing one whole dimension.

Size of the engine is very relevant. Products are in general manufactured in factories because it is orders of magnitude more efficient than if people make them at home. Similarly, airplanes use small numbers of huge engines rather than large numbers of small engines because huge engines can be built more efficient. If that wasn't true, using large counts of small engines would not mean any disadvantage and would be in fact safer as losing a few would not mean any threat.

Is your analysis in a format that is sharable? I'd like to see how you took engine size into account.

I just created a 3D scatterplot of ln(mass), ln(thrust), and Isp. I still have not figured out how to look correctly at that graph (and maybe the metric is not good at all) but all engines lie approximately in single plane in that graph, except for LV-N which makes a huge spike.

Edited April 4, 2014 by Kasuha

[Slouch]

Balancing stock parts is fine, but really, enforcing said balance is pointless.

Balance your stock as much as you feel like, just don't expect that to hold, is what I mean.

I think the stock parts are fine, but saying that since you can edit a config file makes this a moot point is IMO a bit daft. I for one don't want to spend time messing with the config file because I don't like the balance of a game, I just want to play the game. Also not everyone could/wants to/will try to edit there config file.

[Dodgey]

Balancing stock parts is fine, but really, enforcing said balance is pointless.

Balance your stock as much as you feel like, just don't expect that to hold, is what I mean.

What do you mean enforcing? Like it or not due to the community KSP is not a purely single player game due to the massive community sharing ship creations (spacecraft exchange). Everybody editing their configs to suit their own desires would just destroy the spacecraft exchange and the whole concept, as what might be really cool to someone else who plays with different confi settings would just be stupid witty someone else. And since multiplayer is on its way the single player argument is now null and void. It just isn't a single player game anymore, at least in the sense that Skyrim is a single player game.

[stupid_chris]

By "mass + thrust" I meant mass combined with thrust. Or rather, we need to use 3D metric, not just 2D metric. Because by combining mass and thrust into TWR you're losing one whole dimension.

Size of the engine is very relevant. Products are in general manufactured in factories because it is orders of magnitude more efficient than if people make them at home. Similarly, airplanes use small numbers of huge engines rather than large numbers of small engines because huge engines can be built more efficient. If that wasn't true, using large counts of small engines would not mean any disadvantage and would be in fact safer as losing a few would not mean any threat.

I just created a 3D scatterplot of ln(mass), ln(thrust), and Isp. I still have not figured out how to look correctly at that graph (and maybe the metric is not good at all) but all engines lie approximately in single plane in that graph, except for LV-N which makes a huge spike.

Except this is not relevant of the performance of the engine at all :/

[Kasuha]

Except this is not relevant of the performance of the engine at all :/

I wonder what part of my post was this related to.

Anyway, I made a quick survey of real rocket engines and to me they appear ever more unbalanced than KSP engines.

[Red Iron Crown]

I just created a 3D scatterplot of ln(mass), ln(thrust), and Isp. I still have not figured out how to look correctly at that graph (and maybe the metric is not good at all) but all engines lie approximately in single plane in that graph, except for LV-N which makes a huge spike.

I think using the logarithms of mass and thrust minimizes their impact on your metric, while keeping Isp linear maximizes its impact. Which would explain the LV-N being an outlier as Isp is its claim to fame.

Anyway, I made a quick survey of real rocket engines and to me they appear ever more unbalanced than KSP engines.

I don't think accurate simulation of real world engines is the goal with KSP.

[Themohawkninja]

Why do people keep bringing up the increased thrust? It's a 3.75 meter engine, designed to lift 3.75 meter tanks, which are going to be much heavier. The increased ISP is fairly minimal seeing as what you can do with a Skipper engine with the 2.5 meter parts (and nobody thinks that it's OP), and the tech tree is irrelevant, because Squad just decided to compress the tree, instead of adding on to it. That's not imbalanced, it just means you get the parts a bit faster. In the end, the largest parts in the game are still found at he same place in the tree.

[Red Iron Crown]

Why do people keep bringing up the increased thrust? It's a 3.75 meter engine, designed to lift 3.75 meter tanks, which are going to be much heavier. The increased ISP is fairly minimal seeing as what you can do with a Skipper engine with the 2.5 meter parts (and nobody thinks that it's OP), and the tech tree is irrelevant, because Squad just decided to compress the tree, instead of adding on to it. That's not imbalanced, it just means you get the parts a bit faster. In the end, the largest parts in the game are still found at he same place in the tree.

I don't think anyone is complaining about increased thrust. It's the higher combination of TWR and Isp that makes them unbalanced.

[maccollo]

Why do people keep bringing up the increased thrust?

TWR is not the same as thrust.

[Themohawkninja]

TWR is not the same as thrust.

I am well aware that a ratio of thrust is not thrust. The OP explicitly (in bold) stated that the engines used more thrust as a point that the new engines are OP, which I am disagreeing with.

- - - Updated - - -

I don't think anyone is complaining about increased thrust. It's the higher combination of TWR and Isp that makes them unbalanced.

The OP complained (in bold at that).

[stupid_chris]

I wonder what part of my post was this related to.

Anyway, I made a quick survey of real rocket engines and to me they appear ever more unbalanced than KSP engines.

Mainly the first part. Mass and thrust don't mean anything on their own. If you don't combine them, you don't get to see how much the engine can really lift, you just get to see that it's heavy and has a high thrust. The numbers don't inform you on the performance of the engine. What will is what we find in the rocket equation. ISP is there, and TWR appears indirectly. The higher it is, the higher the trust you can have for the lowest dry mass because the engine contribution to it reduces.

[Kasuha]

I think using the logarithms of mass and thrust minimizes their impact on your metric, while keeping Isp linear maximizes its impact. Which would explain the LV-N being an outlier as Isp is its claim to fame.

I don't claim my metric to be best but I still believe we need to use (at least) three parameters to put new engines into context properly. So if you know better approach, fire away.

I don't think accurate simulation of real world engines is the goal with KSP.

I don't think having engines better "balanced" than in real world is the goal with KSP.

[Red Iron Crown]

The OP complained (in bold at that).

Reread the OP. He's complaining that an engine that is the same mass as the Mainsail has 33% more thrust. Effectively complaining about the new engine's high TWR.

[Red Iron Crown]

I don't claim my metric to be best but I still believe we need to use (at least) three parameters to put new engines into context properly. So if you know better approach, fire away.

I believe stupid_chris' metric of TWR vs Isp is better, and does use three parameters.

I don't think having engines better "balanced" than in real world is the goal with KSP.

Balanced engines would mean more variety in construction, and more quantitatively valid choices for the player. More realistic engines would result in NTRs for everything.

[Rakaydos]

Is there any good "engine rebalancing" mods to fix the nasa engines, the 55, and the rockomax probe?

[stupid_chris]

Is there any good "engine rebalancing" mods to fix the nasa engines, the 55, and the rockomax probe?

I'm planning on releasing something like this maybe this weekend.