Part balancing standards the IKAO and you!

[Johnman]

I applied these formulas in a tank I\'ve been making today and I must say they\'re spot-on. It weights a lot, 15.4, but 3,500 units of fuel allow for a lot more than common tanks do!

[nivvydaskrl]

I applied these formulas in a tank I\'ve been making today and I must say they\'re spot-on. It weights a lot, 15.4, but 3,500 units of fuel allow for a lot more than common tanks do!

Thanks! I\'m good at formulas; however, actually formalizing them and putting them into forum markup took an hour.

It\'s good to have them down though. Maybe a wiki page would be good.

[Wintersniper]

We can probably get one of the artsy people on the forum to make an original pic. And your java stuff stays in your post until we have properly decided on something, so don\'t worry about that.

What\'s this? something here I actually have the correct talents to offer to help with? Amazing! if you still need help with this I\'ll gladly do so. I\'m a bit of an artist, and currently taking classes for art related stuff as well.

[NovaSilisko]

I\'ll still insist we convert the base game\'s balance into my version of it.

[Johnman]

What\'s this? something here I actually have the correct talents to offer to help with? Amazing! if you still need help with this I\'ll gladly do so. I\'m a bit of an artist, and currently taking classes for art related stuff as well.

Go crazy! Feel free to make proposals for an IKAO logo. Best if it\'s semi-official looking imho, like the game\'s logo.

[Zeroignite]

I\'ll still insist we convert the base game\'s balance into my version of it.

I would like this.

[Boulway]

Hm i was thinking about this Problem yesterday evening without knowing about your Thread.

I use several Mods together and its really pain that its so unbalanced between each other.

I made a small list to show the problem. (I divided Max Thrust / Consumption)

KSP Vanilla LV-T30 25

KSP Vanilla LV-T45 25

C7 C7AX-301 25

Saturn V Medium Bertha 18,51851852

Saturn V TG-175 29,16666667

Saturn V Lander Module 60

It would be great if there would be a balancing for the mods or if the modders dont want it maybe a converting tool.

So you choose for example that liquid engines have max thrust / consumption = 25 (like in vanilla modules).

You let the max thrust form the mod and alternate the consumption ratio so that you got a Thrust/Consumption of 25.

I think a converting tool could look up the parts and normalize them in the way you want.

My programming skills are sadly to weak for such a converter. I only can hardly write C for Atmega MCUs.

But if you need some help feel free to give me something to do

Regards,

Boulway

[Tiberion]

So there is some rumbling about perhaps resurrecting SIDR and Wobbly if they go completely dormant, combining them and balancing them as a whole pack.

So I set down and ran the numbers through the calculator thing to see what it would look like, thought I would share the spreadsheet to see what you think about it, what might need changed, added, etc.

I am pondering how cutting all of the values by 75% after balancing them might work, to escape the Heavy-mass issues the game engine has. The Diameters/Heights wouldn\'t change, and everything would be 75% of 'real' values - but as long as they\'re balanced together it should still be alright, ya?

https://docs.google.com/spreadsheet/pub?hl=en_US&hl=en_US&key=0Ah0q6AoAI_W6dE5fUHo3NkhEUG9CODYzTFpCRTU2OWc&output=html

Also, I guess you guys haven\'t tackled solid boosters entirely, yet?

[Killerhurtz]

I\'ll still insist we convert the base game\'s balance into my version of it.

While admittedly your pack is pretty good, remember the game is still in early development. That\'s the main reason nothing\'s really balanced yet: as I believe Harv said, the mechanics have priority over content. Anyhow, just look at the Gimbaling engine\'s troubles: just a little fix had him rectifying the only part doing it. Now try to see if on top of that he had to make sure everything was still in balance to a standard. That\'s EXACTLY why game balance is one of the last issues once everything has been ironed out. But it\'s going to happen sooner or later. Be patient, and soon you might shine...

[Boulway]

@Killerhurtz,

You are absolutly right that the balancing can only be solved if we know how the final version will behave.

But i dont think all guys here like to wait so long

And it can not hurt to think about the balancing before the final is released.

A balanced/normalized MOD can easier be rebalanced if the changes are needed.

Perhaps this could also help the devs with completing this great game.

@Balancing

One aspect in balancing that is not used in game yet is the costs of a part.

Maybe it is not bad to think about this while editing the stats.

For example if you have a 1x1 diameter tank with really much fuel inside you can also rebalance them with costs recalculations.

I can imagine that the KSP-Scientists think of different fuels and better tank materials. Of cause they need to cost more.

Same with the engines. A Engine made of junk and an old soldering-iron is much cheaper and more ineffective then a high end engine from kermany ;P.

Of course the costs modules are not working now but it could rebalance some parts in the future.

Regards

Boulway

[Tiberion]

So I plugged in the numbers to the parts for some testing. Looks pretty good.

Question on the heat generation; How we do know how much to scale up the tolerance on the engines and tanks?

With the value in place, and the bumped heat numbers the calculator produces, the overheat within seconds.

Is heat best used for part banding, to make sure 3m engines are only used on 3m parts (by setting 2m tanks too low to tolerate the heat)

Or should we drill down and set up the heat numbers so that engines overheat after a minute of redline usage?

[Deusoverkill]

well this is the calculator i made for the engines and tanks, its a very simple one, for the engines just plug in the maxthrust and efficiency and it gives you the fuel consumption; in the tank part you plug the size (diameter and height), and it gives you the mass and fuel amount. the constant values where calculated from the vanilla parts.

it also calculates the downscale of all the results in 75% 50% 25% and 10%,

its an excel file, just put the blue box values in, and thats it. hope you find it usefull.

cheers!

[Boulway]

Hi Tiberion,

The Heat mechanics are not totaly clear to me.

The standart engine LV-T30 got a heat production of 500 at maxthrust from 200.

So LV-T30 produces 2.5 heat for 1 thrust.

The Omni Bertha from your List got 1500 thrust.

1500 * 2.5 would be 3750 like in your list.

Maybe we should take a max heat production level?

Or maybe another way for calculating the heat production?

Like a droping value table.

It would be much flatter if you take 35* root of (thrust) = heat

That would give us following values.

35* root of (200 thurst) = 494.975 heat

35* root of (1500 thrust) = 1355.544 heat

We should then make some rounding. I think in steps of 50 heat.

494.974heat would be 500 heat

1355.544heat would be 1350heat

What do you think of it?

Best Regards

Boulway

[Tiberion]

I dunno.. I bumped up the Maximum temp on the engine by 10x and they don\'t explode, but the heat transfer to connected parts is wicked, any tanks or couplers that could possibly be attached would need to be heat-proofed too.

I rolled back to the previous heat settings for now, til I can run some closed-system tests on a few parts.

[Festivefire]

once you get all of the ratios in place, you could probably add a tolerence equation for the thrust/consumption ratio. like if you want a +5 tolerence for 30, you could have a equation like (((heat tolerence / heat production) * mass) /10) - (cost/100) = positive tolerence. this also means you can have finicky engines for more thrust or efficiency, and you could ramp the cost through the roof for cheat parts, but this equation still needs tweaks cause it makes the lv-t30 have a -7 point something tolerence, meaning its more efficient for its price, but it should be the baseline. if someone could fix it and explain what i did wrong, or just post something else that works and explain why i\'m wrong that would be great.

[Killerhurtz]

once you get all of the ratios in place, you could probably add a tolerence equation for the thrust/consumption ratio. like if you want a +5 tolerence for 30, you could have a equation like (((heat tolerence / heat production) * mass) /10) - (cost/100) = positive tolerence. this also means you can have finicky engines for more thrust or efficiency, and you could ramp the cost through the roof for cheat parts, but this equation still needs tweaks cause it makes the lv-t30 have a -7 point something tolerence, meaning its more efficient for its price, but it should be the baseline. if someone could fix it and explain what i did wrong, or just post something else that works and explain why i\'m wrong that would be great.

If you want the LV-T30 to be the reference, your equation should be ((heatTolerance/heatProduction)*mass/10)-(cost-costLV/100). That way, if the engine costs the same as the LV-T30, the numerator will be zero, and well 0/x=0.

Although I\'m throwing it a quick idea (which will need a few additions to the engine for it to work) for an equation regulating the cost according to stats:

((thrust+heatTolerance+(y*gimbalRange))/(heatProduction+weight+fuelConsumption))*manfRate*factorX=cost

which would imply commercial preferencial rates (although that can be skipped) and a preset multiplier (which could be used to make parts cheaper as newer, more reliable or more efficient engines are made, or some other criteria, to manually balance the costs). In my opinion it would be the most complete and global equation: high thrust, heat tolerance and gimballing would make a part more expensive while higher weight, heat production and fuel consumption would reduce the cost of the part (although that would be only for the part, it would require planning, as a high heat production is going to require higher heat tolerance in a few other parts, higher consumption will require more fuel, yadda yadda). That\'s without mentioning that it would auto-regulate the cheat parts to be nigh unusable due to high cost (as most cheat parts either have very little fuel consumption or absurd amounts of thrust). Thoughts?

[Festivefire]

i think that is a great eqation for the balancing and marketing of parts. unfortunately, a ridiculous price will not deter players until campaign mode is added. however, the area were balance is currently most important, challenges, can have budgets or limits on parts packs.

[Killerhurtz]

i think that is a great eqation for the balancing and marketing of parts. unfortunately, a ridiculous price will not deter players until campaign mode is added. however, the area were balance is currently most important, challenges, can have budgets or limits on parts packs.

Until campaign mode, NOTHING can deter cheat parts.

[Tiberion]

There\'s no reason TO deter cheat parts until then either. Though prototyping the expense equations to use in challenges would be pretty swell.

So I have been testing the numbers quite a bit ingame. Having all of the engines with the same efficiency feels a little strange - one long 1.75m tank and a bertha engine can lift a munlander+orbital engine pack into orbit, seems too effective. But the lander engine can\'t handle descent and ascent on a lander-sized tank, even when I fudged the numbers a little.

I guess I need some other piloting styles to test them out. I didn\'t hear back from Nova or Sundaypunch yet, so I won\'t post them, but I will give the link out to a few people if you PM me, if you wanna see how these numbers are working out.

[Killerhurtz]

If today we had the same engine efficiency as the earlier ones, we probably would have never reached the moon NOR have a space station. The goal is not equivalent efficiency; it\'s to have it balanced with the future campaign. Otherwise, if all mods were as efficient, we\'d be designing only one phase of it.

[Festivefire]

i think that we need to find a way to incorporate mass of the engine into the efficiency equation, cause you could make a perfectly efficient engine of 5000 thrust, and have it weigh 0.000000001 mass units, making it way overpowered, or have an engine with an efficiency of 30 and max thrust of 30, but make it weigh 40 mass units, rendering it useless, even though the equation says it overpowered. this way we can have more or less efficient engines without having them be accused of being cheat parts for challenges and campaign mode.

OR

we could balance weight of the engine dependent on size, crash resistance, and heat tolerance, and make heat production dependent on efficiency and weight. this way you get a triangle of values, but they don\'t all have to be in the same equation.

[zweerify]

You\'re doing a great and very important work here, guys! I beleive that your researches should become some sort of canonical thing in nearest future.

But maybe it\'s too early to mess with costs of parts? We have some behavioral & unmeasurable characteristics, like \'overall coolness\', \'design\' and \'non-red-texture\' for example. Also, it somewhat hard for me to include simultaneously \'drag\', \'tolerance\', \'heat proof or-whatever-it-named\' and all listed above fuel-capacities et cetera in one big formula/equation.

I mean, it\'s great to have coefficient of vanility, and model balanced parts depending on it, but don\'t you think that computing cost depending on it is too complicated and, er, not needed at all?

i think that we need to find a way to incorporate mass of the engine into the efficiency equation, cause you could make a perfectly efficient engine of 5000 thrust, and have it weigh 0.000000001 mass units, making it way overpowered, or have an engine with an efficiency of 30 and max thrust of 30, but make it weigh 40 mass units, rendering it useless, even though the equation says it overpowered. this way we can have more or less efficient engines without having them be accused of being cheat parts for challenges and campaign mode.

OR

we could balance weight of the engine dependent on size, crash resistance, and heat tolerance, and make heat production dependent on efficiency and weight. this way you get a triangle of values, but they don\'t all have to be in the same equation.

Good point! To split equations it\'s a good way to achieve flexibility and still have objective values, which measures how much part is well-balanced.

I find it hard to express my thoughts in english without knowledge both in equations and language at all. Sincerely, your poorly-speaking russian comrade.

[Festivefire]

You\'re doing a great and very important work here, guys! I beleive that your researches should become some sort of canonical thing in nearest future.

But maybe it\'s too early to mess with costs of parts? We have some behavioral & unmeasurable characteristics, like \'overall coolness\', \'design\' and \'non-red-texture\' for example. Also, it somewhat hard for me to include simultaneously \'drag\', \'tolerance\', \'heat proof or-whatever-it-named\' and all listed above fuel-capacities et cetera in one big formula/equation.

I mean, it\'s great to have coefficient of vanility, and model balanced parts depending on it, but don\'t you think that computing cost depending on it is too complicated and, er, not needed at all?

I assume tour talking abiut this equation right? when I use it I use a baselibe value for price abd useualy fiddle wuth heat productuon and mass. Price is there for when we gey campaign mode

If you want the LV-T30 to be the reference, your equation should be ((heatTolerance/heatProduction)*mass/10)-(cost-costLV/100). That way, if the engine costs the same as the LV-T30, the numerator will be zero,

[Festivefire]

just had a brainwave. for tech-tree progressions of engines and boosters, we can have a chart that gives different efficiencies for different generations of engines. like a generation 1 engine will have an efficiency of 25, to match the lv-30, and a generation two engine could have a base efficiency of 26 and so on and so forth. this way we can use all of the same equations, and just change a value, and still have balanced mods.

[Killerhurtz]

Guys, I just had an idea to help with balancing parts in the current build.

We use the equation which relates to price, and we use 'price levels' as baselines for general guidance - something similar to this:

Efficiency level 1: 200+

Level 2: 400+

3: 600+

And so on and so on. What do you guys think?