(9-VT) Higher, Balanced Upgrade Costs

[Alpakinator]

Problem:
Building cheaper units before unlocking the upgrade, then upgrading them, is too cheap. More efficient to conversion than buying units, or investing in buildings. The best way to use gold, a no-. It's cause the formula is:
upgrade cost = (

of target unit -

of current unit) + 10.

eg. Knight to Lancer:

195 = ( 350

- 165

) + 10.
So we save 185 by paying 195, and get full XP.

Meanwhile it takes 2x morethan to make early units.
It's ridiculously bad for balance and makes "decisions" trivial. Allows for too fast Spearman and Archer rushes on pre-wall Capitals. Later rushes too. Partly why warmongering is too easy.

I made a website to show the problem and allow you to find a good solution:

https://alpakinator.github.io/vox-populi-mechanics-visualizer/

Spoiler Gold costs to production costs ratio graph of the website :

The lower on the graph, the less

it takes to save up on 1

.
The green "upgrade" curve is way lower than red or blue until modern era.

Adjust the formula with sliders on the left so that the yellow curve is reasonable. Then propose your formula if looks better than mine.
My solution:
Change the upgrade cost formula to:
1.3 * cost difference + 30

Spoiler Graph with this formula :

Pros:

1. Early upgrades tiny bit less efficient than buying. Tradeoff with pre-building caused time saving, important vs pre-wall cities.
2. Upgrade cost discounts matter more; strengthens underpowered Imperialism.
3. XP Discrepancy between bought and pre-build units is higest for post Military Academy 60XP+ units, so value of upgrading units rises vs buying them. Late game, risinng upgrade cost discrepancy due to *1.3 counters that.

I'm against making the formula more complex. Simple formula lets players easily predict their future upgade costs (cause they're not shown in the UI). And despite simplicity it has nice side effects, like the counter to late game full XP.
Interactions with other proposals:
I think none directly.

I'll help explain the graphs if they're confusing.

 

[azum4roll]

If reworking the invest/purchase gold formula is out of scope for this congress, I don't see how this is allowed.

 

[Alpakinator]

I want to place a bit of a one-off restriction on this session and limit it to "balance". Specifically I won't be allowing any major overhaul proposals to give the devs more time to improve stability and fix bugs as we embark on the 5.0 adventure! So things like the integration of big modmods like Enlightenment Era, revamps to the Promotion system, swapping out all the AI with ChatGPT, and really anything that would require speculative changes to the DLL will not be allowed in the first 5.0 session. Hopefully that feels sensible to everyone. The MAGI and I will have to assess on a proposal-by-proposal basis but I don't think it's beyond the wit of man to achieve this.

Yeah. Both sound like a balance to me and don't seem major enough to be an 'overhaul'. But purchase formula touches buildings, units, wonders and more policies so it approaches it.

 

[azum4roll]

Personally I think we should standardize upgrade cost from A to B to max(10, base purchase cost of B - base purchase cost of A), whatever base purchase cost formula is. There shouldn't be a difference between buying A then upgrading to B and straight up buying B.

 

[sbw]

If reworking the invest/purchase gold formula is out of scope for this congress, I don't see how this is allowed.

Seems a bit silly to me that we can have entire systems (corporations, religion) reworked or entire policy trees rewritten but a small self-contained formula change is too speculative.

In any case, strongly support this proposal and the reasons behind it are well articulated. Upgrading units is super efficient in terms of money per poduction, XP saved, and time gained (by not having to build the current unit after it unlocks). An alternate proposal could also be to tie the upgrade cost to the current era's investment efficiency, but that's a bit more complex.

 

[FPXI]

I kinda like the early upgrade efficiency, but ya this seems fine.

Personally I think we should standardize upgrade cost from A to B to max(10, base purchase cost of B - base purchase cost of A), whatever base purchase cost formula is. There shouldn't be a difference between buying A then upgrading to B and straight up buying B.

This also seems good. Would support either approach.

 

[hokath]

That's a very beautiful piece of interactive graph wizardry. Impressive. Very nice.
I am of course completely in favor, in principle.

How would it interact with

S

Thread '(9-VT) Simplify Gold Rush Buy Formula and Increase Late Game Costs'

Jan 3, 2026

Motivation:
The gold hurrycost formula is a bit complicated and consists of 4 different factors which are then multiplied together. It has some unexplained mechanics and becomes very powerful late game when combined with bonuses from Banks, Stock Exchanges, or Imperialism. The intent of this proposal is twofold

1. Simplify the number of factors going into the calculation of how much it costs to rush buy something
2. Rework the cost scaling to increase the costs for late game rush-buying.

For context, here is the existing HurryCost formula for units:

Gold Cost = 0.8 * (30 *...

• sbw
• Replies: 17
• Forum: Failed Proposals

• 

Would it be sensible to make the two changes together and test them out-of-congress?

 

[Alpakinator]

That's a very beautiful piece of interactive graph wizardry. Impressive. Very nice.
I am of course completely in favor, in principle.

How would it interact with

S

Thread '(9-VT) Simplify Gold Rush Buy Formula and Increase Late Game Costs'

Jan 3, 2026

Motivation:
The gold hurrycost formula is a bit complicated and consists of 4 different factors which are then multiplied together. It has some unexplained mechanics and becomes very powerful late game when combined with bonuses from Banks, Stock Exchanges, or Imperialism. The intent of this proposal is twofold

1. Simplify the number of factors going into the calculation of how much it costs to rush buy something
2. Rework the cost scaling to increase the costs for late game rush-buying.

For context, here is the existing HurryCost formula for units:

Gold Cost = 0.8 * (30 *...

• sbw
• Replies: 17
• Forum: Failed Proposals

• 

Would it be sensible to make the two changes together and test them out-of-congress?

Thank you! It was inspired by this @sbw 9-VT proposal.

Mine would interact directly with it, cause upgrade curve should trace along the purchase/invest curves that that 9-VT aims to change.

If 9-VT passes in the future with the linear or intermediate approach, and 9-120 doesn't, then upgrading becomes even more of a obviously correct choice.

If 9-120 passes and 9-VT doesn't then mine is fine.

If both pass, then 9-120 should be modified to fit 9-VT, by making upgrade costs even more expensive, especially late game, probably by raising the 1.3 coefficient.
So yes, it's best to do both together imo.

EDIT: to be clear, I'm against full linear 2/1 of the original 9-VT. i think currentbalance relies on some non-linearity as costs increase. I'd only support it with a gradual buffs to gold yields as the game progresses. I feel something like 60% of the current curve and 40% of the 9-VT linear curve is good, but it's just vibes.

 

[Alpakinator]

Good

balance means that

buildings, specialists, and improvements should be similarly strong vs their

counterparts throughout the game assuming no extra civ or policy bonuses that skew such balance. Is that the right goal to aim for?

I think parts of current flawed

balance is on life support due to mechanics we want to fix, and some parts suffer from it.
Example:

Plantations, forest-less camps, merchant specialist yields are weak if we judge them with early-mid game purchase/invest atio. But late game when it's 1/1, they become good. And if the player uses all their gold early game for unit upgrades, then they almost use 1

/1

, thus making these

sources strong even early game.

But there are other sources that are strong even assuming 2

/1

, like some International Trade Routes, Villages, Thrift , God of the Opne Skies, Fealty etc. that become too powerful if player uses gold for upgrades a lot, and purchases armies late game. And that seems like a bigger issue. And not to mention the purchase discount stacking test-able on my website.

So the issues we want to fix are actaully good things for some elements of the balance and bad for others. I think more of the second. The game impact of both our proposals (assuming a not fully linear 9-VT) will not be harmful for balance, while addressing some problems of warmongering, giving interesting use choices to players, making gold costs clearer and not overly powerful late game.

 

[redrum68]

Good

balance means that

buildings, specialists, and improvements should be similarly strong vs their

counterparts throughout the game assuming no extra civ or policy bonuses that skew such balance. Is that the right goal to aim for?

I think parts of current flawed

balance is on life support due to mechanics we want to fix, and some parts suffer from it.
Example:

Plantations, forest-less camps, merchant specialist yields are weak if we judge them with early-mid game purchase/invest atio. But late game when it's 1/1, they become good. And if the player uses all their gold early game for unit upgrades, then they almost use 1

/1

, thus making these

sources strong even early game.

But there are other sources that are strong even assuming 2

/1

, like some International Trade Routes, Villages, Thrift , God of the Opne Skies, Fealty etc. that become too powerful if player uses gold for upgrades a lot, and purchases armies late game. And that seems like a bigger issue. And not to mention the purchase discount stacking test-able on my website.

So the issues we want to fix are actaully good things for some elements of the balance and bad for others. I think more of the second. The game impact of both our proposals (assuming a not fully linear 9-VT) will not be harmful for balance, while addressing some problems of warmongering, giving interesting use choices to players, making gold costs clearer and not overly powerful late game.

FWIW, I think some of these gold sources need balanced a bit as well. For example, engineer vs merchant yields early game are quite unbalanced.

 

[sbw]

That's a very beautiful piece of interactive graph wizardry. Impressive. Very nice.
I am of course completely in favor, in principle.

How would it interact with

S

Thread '(9-VT) Simplify Gold Rush Buy Formula and Increase Late Game Costs'

Jan 3, 2026

Motivation:
The gold hurrycost formula is a bit complicated and consists of 4 different factors which are then multiplied together. It has some unexplained mechanics and becomes very powerful late game when combined with bonuses from Banks, Stock Exchanges, or Imperialism. The intent of this proposal is twofold

1. Simplify the number of factors going into the calculation of how much it costs to rush buy something
2. Rework the cost scaling to increase the costs for late game rush-buying.

For context, here is the existing HurryCost formula for units:

Gold Cost = 0.8 * (30 *...

• sbw
• Replies: 17
• Forum: Failed Proposals

• 

Would it be sensible to make the two changes together and test them out-of-congress?

There's two different efficiencies here (as shown on the neat website). Rush purchasing starts out somewhat inefficient and ends up hyperefficient late game. Upgrade efficiency is basically flat throughout the game, but that means it's *relatively* much more efficient in the early game (since rush buying is weak) and less efficient late game (since rush buying is strong).

My proposal nerfed the change in rush buying efficiency (i.e. not making it more efficient late game). This proposal doesn't really affect the change in upgrade efficiency over the game, just increases it across the board

On a separate note, I think maybe a counterproposal could be to make cost scale with XP earned. Something like +1% upgrade cost per XP earned on the unit or +10% upgrade cost per level of the unit could increase costs to accurately compensate for XP carryover of upgraded units being so strong. Thoughts on this?

 

[paul_soryu]

Wouldn't this introduce a problem in early to mid-game when you just wouldn't have enough Gold per Turn to upgrade all your units between tech unlocks? Or even if you would have enought GPT, it all would be going towards upgrading your units, leaving you effectively with no choice where to spend your gold? So for example you just upgraded your last Warrior into a Spearman, but you also already unlocked Pikeman and need to upgrade all your Spearman again.

I am not saying it will, just asking if this was considered, because you have no other way to upgrade your units other than by spending Gold, so you can not not spend it here (unlike with buildings where you can). You can disband units and build/buy new ones (effectively "investing" into new unit's production), but you lose all XP. Disbanding your units for money currently isn't worth it, I think it would go well with this proposal if disbanding's money return was buffed to compensate for the loss of XP. So like Gold return = 50% of unit's purchase cost + (10% * [current level - 1]). So at level 6 it is a full refund. It would make disbanding high-XP units to replace them with new ones not as bad as it currently is, so an alternative to direct Gold upgrades.

I did some calculations, it seems like the change isn't that big in numbers (well duh, it's ~33% more, but I don't have unit Prod costs and Upgrade costs memorized), I think there should be an additional graph showcasing how much units cost to upgrage now vs. how much they will with your proposed formula, to make it easier and faster to understand.

I think late-game units already cost a lot to upgrade (upgrading big armies without Imperialism can cost you tens of thousands of Gold), but I am also of an opinion that there's a lot of free Gold in late game, so I don't mind late game upgrade costs being increased, you're also saving a lot of accumulated XP with upgrades so seems logical. It actually also kinda nerfs Rationalism, because with how quickly they can progress in late-game tech tree with increased Science and GS bulbs, they might not have enough Gold to keep up with upgrading all their units. I like that, Rationalism should struggle with every other yield in exchange for getting a lot of Science.

 

[paul_soryu]

On a separate note, I think maybe a counterproposal could be to make cost scale with XP earned. Something like +1% upgrade cost per XP earned on the unit or +10% upgrade cost per level of the unit could increase costs to accurately compensate for XP carryover of upgraded units being so strong. Thoughts on this?

Haha I just wrote my reply without reading yours and I used the same idea but for unit disbanding, funny how both of us got this idea nearly at the same time

I think actually both ideas complement each other and this proposal, you can upgrading high-level units is costly, but disbanding them also gives a lot of gold, so if you're struggling with gold (or you need it for other things) you can disband a unit (or some of them which are not so high-leveled) to upgrade a higher-level unit while also building a replacement. Or you can upgrade more mid-level units and disband high-level for more gold return. It sounds like introducing "high army" (fewer but high-level, upgraded) and "wide army" (more but low-level, built anew) concepts A cherry on top could be that high-level units also have higher Gold Maintenance (it is too low currently anyway compared to building maintenance in my opinion)

So this way at each new tech upgrade a natural thing would be to make your army smaller (but more "concentrated") by disbanding mid-level units and upgrading high-level units, while also building newer units (which consumes instead of for "upgrading" which is the best part in my opinion) to replace the disbanded ones. I think it represents real history much better than current "upgrade all units indefinitely", at least on a mechanical level. Disbanding should be more important than it currently is imo, one of my biggest issues with units is how production is never required past initial unit construction, it heavily benefits the side of the war which doesn't lose units, allowing them to not spend production on building new units. Moreover, often after building Military Academies and/or Brandenburg Gate many of my units, especially Ranged, have less XP than the newly build ones would have, so I just disband old ones and build new ones. So this "disbanding-upgrading" strategy is already present in the game to a small degree.

Also why I like this is that it makes upgrading more smooth instead of "all at once". If you're in an active war, you might want to upgrade powerful units and keep weaker units unupgraded for now and use them as support (so that you can sell them after the war is over) instead of frontline assault troopers, which would be the upgraded ones.

The only problem is I don't know if AI can handle all of those ideas and strategies.

 

[hokath]

After discussing with the authorities, I think we should handle this and sbw's proposal together, in a joined-up way.
Otherwise it will get quite messy and there will be a lot of reduplication of effort.

I'm going to make a channel on the discord, where we can coordinate this implementation.
Then, we can distribute as a mod in the official release thread. At the next congress we can ratify the final version we arrive at.

Proposal Vetoed

 

[redrum68]

After discussing with the authorities, I think we should handle this and sbw's proposal together, in a joined-up way.
Otherwise it will get quite messy and there will be a lot of reduplication of effort.

I'm going to make a channel on the discord, where we can coordinate this implementation.
Then, we can distribute as a mod in the official release thread. At the next congress we can ratify the final version we arrive at.

Proposal Vetoed

Does this mean any proposal to change unit upgrade cost will be vetoed?

 

[hokath]

Most likely, yes.

 

[Alpakinator]

FYI, both proposals are being tested as a modmod, that will be released for testing and feedback after some fine tuning of formulas. Then It'll become a proposal on the next congress.

It's a shame this proposal is vetoed since there are other non-vetoed proposals rn also with 'speculative' implications whole game while being less beneficial to game balance.

 

[Alpakinator]

Wouldn't this introduce a problem in early to mid-game when you just wouldn't have enough Gold per Turn to upgrade all your units between tech unlocks?

I'm testing a much more brutal version as Ottomans on Discord (#playthroughs) with:
upgrade cost= 2* ( of target unit - of current unit) +10.

Together with the constant 2/1 of the other vetoed proposal. I think it's clearly too harsh making gold not valuable enough but it's fine.
Eg. Archer to Comp. Bow turns from 40 to 90 and i have no problems upgrading all my 6 arches while being 2nd weakest civ on Deity fought by 2 neighbours. I noticed AIs with low / struggling to upgrade their units though.

 

[paul_soryu]

I noticed AIs with low / struggling to upgrade their units though.

In early game there's a huge difference of how much GPT you get depending on the start (and Pantheon chosen). Resources like Truffles, Cotton, Perfume, Silk and Tobacco can make you a millionaire, while with some others you'll barely come out even (like +3 GPT). As I said before, I think there should be a somewhat alternative way to upgrade units, like spending hammers on "upgrades". If disbanding units was made stronger, upgrading by building or buying new units would become a viable method. Gold upgrades could be viewed as a luxury for the best units.

 

[Alpakinator]

You can trade luxuries, strategics, tribute CS, win luxury quests, build markets, connect cities. If you have 10 /t with this brutal version you won't have problems with ancient/classical upgrades provided you use the gold mostly for upgrades of your large army.