Paradox: this has to stop (build/upkeep cost modifiers)

[RoverStorm]

The solution is to make these bonuses stack multiplicatively, not additively.

I hate to repeat myself in the same thread, but multiplicative isn't a sollution:

Spoiler

Say I have a BASE mineral income of 1000 and a BASE mineral upkeep of 500 (total 500 income):

Additive:
Ten +10% mineral outputs is +100%. (2000 minerals, -500 upkeep = 1500 income)
A hundred +10% mineral output is +1000%. (11000 minerals, -500 upkeep = 10500 income)
This is linear scaling. Each modifier is the same "empire wide" increase to the income, which means that each modifier is PROPORTIONALLY less impactful and a smaller percentage of your income.
Thus, this process is both LINEAR and DIMINISHING, therebye it can be called "Stable Growth".

Ten -10% upkeep modifiers is -100%. (-0 minerals, total 1000 mineral income)
A hundred -10% upkeep is -1000%. (-0 minerals, total 1000 mineral income)
This is proportionally exponential capped scaling. Each modifier is stronger than the previous, BUT it is capped.
Additionally, you cannot CREATE resources, this is reducing the DRAIN. This is not growth, this is called "Efficiency", or "making better use of what you have".
This is actually still LINEAR but it is INCREASING, therebye it can be called "Capped Explosive Efficiency Gain".



Multiplicative:
Ten -10% upkeep modifiers is roughly -65%. (-174.34, total 825.66 mineral income)
A hundred -10% upkeep modifiers is (checks calculator) ...we'll just call that -99.99% because I think that's the decimal limit for Stellaris (-0.01 minerals, total 999.99 mineral income)
This is EXPONENTIAL and DIMINISHING, as each modifiers is having less of an impact than the previous. Think for a second, the first 10 modifiers did 65% of the work that the last 90 did. We will call this "Stable Efficiency Gain"

Ten +10% mineral outputs is roughly +259%. (2593.74 minerals, -500 upkeep = 2093.74 income)
A hundred +10% mineral output is roughly +13780% (13780612.34 minerals, -500 upkeep = 13780112.34)
I....don't think I need to explain this. EXPONENTIAL and INCREASING. This is EXTREMELY bad because each modifier has substationally more impact than the previous.

End results:
No modifiers = 500
Additive:
+10%*10 income = 1500
+10%*100 income = 10500
-10%*10 upkeep = 1000
-10%*100 upkeep = 1000

Multiplicative:
+10%*10 income = 2093.74
+10%*100 income = 13780112.34
-10%*10 upkeep = 825.66
-10%*100 upkeep = 999.99


When looking at the functions and scaling, this is why I don't think it's a good idea to use "multiplicative" modifiers to replace "additive" ones in Stellaris.

Put in simpler terms negative multiplicative modifiers are STILL way stronger than additive positive ones.

One -10% upkeep allows for about +11% more ships. That's also true for a +10% alloy output modifier, assuming the only thing you spend alloys on is ships.
Ten -10% upkeep, multiplicative, is roughly -65% upkeep. This allows roughly +285% (EDIT: I believe I made an error, I think this is actually +185%) more ships.
Ten +10% alloys, on the other hand, is only about +100% more ships.

Even with multiplicative, it STILL scales stronger than the proportional positive modifier. You have to change POSITIVE modifiers to Multiplicative as well, and THAT would be a COLOSSAL mistake.

 

[Kabian]

I swear half the time someone new in a Risk of Rain 2 lobby asks a question about how item modifiers stack, they are told to open a wiki page or download a mod that puts how they stack in the description. If they refuse on the grounds that a game shouldn't require external devices to play it, they get kicked. That's a good game, but a definitive flaw. That's also not a strategy game, that's an over-the-should horde shooter.


I'm not that far into XC3, so this could be totally different from what I remember once you get further in, but I don't think you have to juggle too many modifiers at once for any given character? Paradox games have way more modifiers you are dealing with, and trying to remember which modifiers are which while planning an entire build and not being able to immediately at all of them at once can get very annoying very fast if different methods of stacking are thrown in.

This should still be added, since it gives the player more accurate information! There are a tiny number of modifiers that are multiplicative, and they should be made very clear.

Sadly paradox already required a lot of external info to even play (tech tree, events, situations, but mostly the tech tree)

 

[ValinShadow]

Montu posted a video reaching 90% discount on ships in the first 30 years. Apparently 90% is the max, would had been 100% otherwise. This is a special build, but other origins can reach this later.

Simple solutions:
only the highest discount should apply from governors if more available
or
bonus after a certain point should be turned into something else, like ship xp

they'd have to also make that change for councilors as well as you can stack multiple councilor with ship built cost decrease as well I believe. even still I think the Agenda still needs to be cut in half

 

[currylambchop]

There are several approaches and many ofthem could be fun - for example The Binding of Isaac, where it is FUN to stack lots of different bonuses and make imba and broken combos. But with grand strategy games with decent multiplayer presence, there should be more careful approach in terms of stacking. And more careful, because other side is Starcraft, with patches about changing decimals in damage and overbalancing, often forgetting that the main goal is to have fun, not calcualting 0.046 shield regen after 2.57 seconds..

Exactly, stacking bonuses is fun, especially in single player.

The multiplayer community can make their own balance mods, where they remove every bonus that has reduced upkeep or increased base production, or whatever the multiplayer meta randomly decides is 'OP'.

 

[user192823]

Exactly, stacking bonuses is fun, especially in single player.

The multiplayer community can make their own balance mods, where they remove every bonus that has reduced upkeep or increased base production, or whatever the multiplayer meta randomly decides is 'OP'.

Currently the only difficulty option is civilian regardless of what is chosen, because that's what easily reachable -90% ship discount is, it so happens that even some of the single players want a difficulty tier above that. Nothing will change for players of your mindset if the ship discounts are nerfed, because there will still be a civilian difficulty option.

 

[RoverStorm]

Nothing will change for players of your mindset if the ship discounts are nerfed, because there will still be a civilian difficulty option.

Strictly speaking, finding that a tactic is easier on lower difficulties is....exactly what you expect. What makes a busted build "fun" is when it lets you do unorthodox or impossible things on deliberately difficult settings. It's almost the same feeling in a speedrun enviroment when you find a really cheesy and easy way to skip an entire region. Like duh, they COULD take the hard way and go through the region, but that's completely missing the point; breaking the game in a deliberately challenging envirtoment is the fun of it.

 

[ValinShadow]

Strictly speaking, finding that a tactic is easier on lower difficulties is....exactly what you expect. What makes a busted build "fun" is when it lets you do unorthodox or impossible things on deliberately difficult settings. It's almost the same feeling in a speedrun enviroment when you find a really cheesy and easy way to skip an entire region. Like duh, they COULD take the hard way and go through the region, but that's completely missing the point; breaking the game in a deliberately challenging envirtoment is the fun of it.

I think there is a key difference between what you're alluding to above and the ridiculous number of ship discounts we have currently, and that difference is ease of achievement.

some of MontuPlays recent videos come to mind.

the Cordyceptic Drones + Here Be Dragons combo (before they patched it) was something challenging to pull off and incredibly unique.

even the videos about Spiritualism having 0 upkeep sounds like it's the same as this current situation (that too was eventually patched) but I think there is a key difference.

with that situation Montu had to stack multiple modifiers from different sources and basically build his entire empire around that attempt. it took effort and planning and that made it unique and challenging.

with the current ship build cost reduction situation, there is nothing challenging or unique about it. any empire can do it and it requires basically no effort.

it's too easy to achieve and far far too powerful for that ease

 

[Seridor]

https://www.reddit.com/r/Stellaris/comments/13enik0

 

[Kabian]

What do you mean you need 2k alloys to increase the size of the starbase?
You can probably make do with 1000
I mean, if you really try 500 should be enough
you know what, i think 100 is a good deal
well, ill be seeing you, remember to give the 175 you will get when increasing the starbase in size by 2

 

[RoverStorm]

What do you mean you need 2k alloys to increase the size of the starbase?
You can probably make do with 1000
I mean, if you really try 500 should be enough
you know what, i think 100 is a good deal
well, ill be seeing you, remember to give the 175 you will get when increasing the starbase in size by 2

Maybe the empire is structured like an MLM...

 

[Kabian]

Maybe the empire is structured like an MLM...

now stellaris economy does make sense

 

[Oculument]

Common sense should always be resorted to as a last resort. Yes.

Why do so many different leaders have a hand in building ships? Too many cooks spoil the stew. Reducing the number of leader-sourced modifiers is one thing that should be done. Perhaps let them all have the modifiers but only the one strongest applies.

 

[RoverStorm]

Reducing the number of leader-sourced modifiers is one thing that should be done. Perhaps let them all have the modifiers but only the one strongest applies.

Some would argue that actually powerful modifiers are best restricted to leaders. Because leaders can die, and the chances of getting a perfect replacement are not super high, with the 350 new leader traits they added.

 

[Tannhäuser Cake]

Versions of this thread have appeared many times on this forum over the years, for many different Paradox games, and I'm sure many non-Paradox games as well. If The Spiffing Brit wrote a generic "games are perfectly balanced" guide without even specifying what genre of game, just giving general tips, you can bet that negative modifier stacking would be on the list somewhere.

Stacking "discount" modifiers (i.e. negative modifiers that benefit the player) additively was a mistake and always has been. You're setting up benefits that can accelerate to infinity by design, which leads to degenerate strategies: either you go all in and break the game with close to -100% cost, or if you can't stack very much of it, then it's not really worth investing in, and the bonus gets ignored. Both ends of this are bad: it's not just that you're making a game feature that minmaxers will break the game with, the feature basically becomes *only* for minimaxers, with more or less negligible effects if you don't go hard on it. Balance is basically impossible to maintain over versions as new modifiers keep getting added. Putting a hard cap on the discount can save the game from becoming broken, but it doesn't really address the issue of all-or-nothing incentives (i.e. either you get to the cap or you don't bother).

Unfortunately, it seems the devs have decided that the obvious solution, of stacking such modifiers *multiplicatively*, goes against their design principles. In some contexts, multiplicative stacking still isn't "diminishing returns", by the way: if you had multiplicative stacking of ship cost/time reduction, then it's still multiplicative stacking of how many ships you can build, i.e. you'd get an exponentially large (with respect to number of bonuses) number of ships with a given number of alloys. This sort of exponential stacking can also end in disaster if you allow too many modifiers. But exponential doesn't go off to infinity in finite time, so as long as there aren't unlimited sources of modifier, it's a lot easier to stop it reaching a game-breaking level. (Something like the governor empire size bonus is still probably a bad idea, even with multiplicative stacking.)

Rather than going the mathematically viable but casual-gamer-confusing route, of having negative modifiers be multiplicative while positive modifiers are additive, I think there are two other options that could be considered instead.

1. No negative modifiers. There can be no runaway reduction effects if there are no reductions to begin with. Rather than having a modifier that reduces the amount of alloys ship construction costs, there could instead be a bonus to how many "ship construction points" you can get for each alloy.

2. No free benefits. Benefits should come with appropriate costs to realize them. Bonuses should not create things out of thin air. Essentially, instead of "input" reductions or "output" increases, "throughput" modifiers should be used.

• The shipwrights are 10% better at building ships? That means they also use 10% more alloys.
  (This would be even better than the "ship construction points" example used above.)
• The ship crews can fire 10% more often with undiminished accuracy? This means the ship will need 10% more energy to the weapons systems.
• Something makes pops happier? They now also get greater expectations.

Combining both of the principles above can be potent, potentially.
The changes involved may require some creativity, but it should not be impossible.

 

[Oculument]

Some would argue that actually powerful modifiers are best restricted to leaders. Because leaders can die, and the chances of getting a perfect replacement are not super high, with the 350 new leader traits they added.

Previously only 2 leader modifiers could be stacked, a ruler and a sector governor. There was high drama involved in getting good ruler traits in previous versions, now design your own starting ruler and add as many governor and assistant planet governors as you want. Game start period is more critical than later when leaders get old and die.

 

[blahmaster6k]

The agenda on its days was the best agenda to get from RNG, in this version, this agenda is stronger, accessible to everyone and on demand

they'd have to also make that change for councilors as well as you can stack multiple councilor with ship built cost decrease as well I believe. even still I think the Agenda still needs to be cut in half

I don't personally have a problem with the current state of the agenda as long as they remove stacking duplicate modifiers like the governor and councilor traits. The new agenda combined with one governor trait and one councilor position would be roughly the same power level as previous ruler optimizations could give, but the new agenda can only be active for 10 out of every 40 years.

 

[Enfield_PDX]

I've talked with Kabian elsewhere but suffice to say we are following the increasingly hilarious string of silly discounts. Enjoy them while they last.
Long term, I think they're something we will try and avoid in the future.

 

[zZander56]

I've talked with Kabian elsewhere but suffice to say we are following the increasingly hilarious string of silly discounts. Enjoy them while they last.
Long term, I think they're something we will try and avoid in the future.

The site won't let me like your post so here's an ASCII thumbs up instead:

╱╱┏╮
╱╱┃┃
▉━╯┗━╮
▉┈┈┈┈┃
▉╮┈┈┈┃
╱╰━━━╯

 

[TehJumpingJawa]

What the devs have done is take this already OP agenda, DOUBLE or QUADRUPLE it's most important parts, give it to all empires, and lock it not behind a random ruler roll, but behind an easily accessible Tradition opener.

Don't forget they also locked it behind a paid DLC.

PI release cycle:

1) Release DLC with ludicrous power creep.
2) Earn $$$
3) Nerf DLC features
4) Fix bugs [optional]
5) Repeat

 

[Dr Pippy]

Good day to everyone and welcome to the PRE-DLC release rant.

In the new DLC, we are getting A LOT of new sources for both -ship cost and ship upkeep, to the point that passing 50% is trivial and reaching 90% is not even hard to do, and we already had problems last time culture workers got added (-100% pop upkeep/amenities) remember?

Please, paradox, build/upkeep costs modifiers are a bane of balance and should be used either never, or on a single, localized part of the game so they cant be stacked.

Why this rant? well, because how due the build/upkeep modifiers cost. The cost tends to 0 as the bonus tends to 100%, which means the following:

As long as the cost/upkeep bonus remains small the effects on balance are fairly limited and controlled, but the moment the modifier surpasses a set value (around 30 in my opinion), the effect of further stacking is amplified so much that the game balance rapidly breaks down:

Some math for all to see, assuming we are going to build 100 units of anything, the cost modifier increases as follows.
with 10% cost you can make 11 more(11% increase)
with 20% cost you can make 25 more (25% increase)
with 30% cost you can make 42 more (42% increase)
with 40% cost you can make 66 more (66% increase)
with 50% cost you can make 100 more (100% increase!)
with 60% you can make 150 more (150% increase)
with 70% you can make 233 more (233% increase!)
with 80% you can make 400 more (400% increase)
with 90% you can make 900 more (900% increase)
with 100% you can make infinite

As you can see, even static augments of the modifier (10% steps) rapidly breaks down the bonus, since a jump from 0-10% has a wildly different effect than going from 50% to 60%.

What does this mean for stellaris? well, it has a easy way of seeing how it affects the economy.

The % increase of stuff you can build can be applied directly to the job that produces the raw material to see the effects.
Per example, if you have 50% build cost, you practically have a +100% miner output after all bonuses are calculated (so a miner that produces 12 minerals, practically produces 24 for building purposes).

This also applies retroactively: if you made 1000 alloys, which is enough to build 100 ships normally, and get 50% build cost, your alloy storage effectively doubled from 1000 to 2000.

Or, per example, if you have 60% build cost, that means your alloy workers have a 150% alloy output, so a job that produces 10 alloys, for practical purposes is making 25 alloys.

And this just gets worse, and its going to get way worse.

As the new dlc comes, and future are planned, the standard of overabusing cost/upkeep modifiers (which is a problem you can see in ALL paradox games), is just going to make the game worse, since the best route to play stellaris is going to stack as much build/upkeep modifiers until your buildings are free, your ships are free and your economy is irrelevant.

I know, build costs are fun, they make you feel powerful (because they are), but they should relegated only to civic bonuses, specially since civics are hard capped to 3, and the most powerful ones are locked to be picked at game start

A (mathematically) straightforward approach to fixing this is to make negative modifiers divide the original quantity, so -100% ship cost means that you pay 50%, not zero. Given a multiplier X% to some quantity Y, the game currently calculates the result as Y*(1+X/100) whether X is positive or negative. This is fine when X>0, but should be replaced with Y/(1-X/100) if X<100. (See below for a graph.)

In addition to avoiding shenanigans like free ships or alloys, you'd also avoid the problem of ships taking years to get across a system that has two -50% ship speed modifiers.