Testing Economic Balance Changes

A forum for chatting about in-development game features.
User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Testing Economic Balance Changes

Postby sven » Sun Jun 30, 2019 4:22 pm

zolobolo wrote:Update: They don't attack when they arrive :(
They have also closed peace with my ally and offered peace (no compensation or demand) - is this why they dont attack?


Could you upload this game (if you still have it)? I've made a bunch of changes to the logic around triggering attacks recently, and it's possible you've just found a bug.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Testing Economic Balance Changes

Postby sven » Sun Jun 30, 2019 4:32 pm

Zoolimar wrote:Something like
Base Upkeep*sqrt((Part1^2+Part2^2+...)/Total Number of Parts)


This is quite clever. A bit obscure, but, I don't think the ship upkeep formula is necessarily something that needs to fully grasped by the player. As long as we show them the upkeep impact of changing around whatever parts in real time inside the designer, it's probably ok if the formulas are a bit complex. I'll need to give something like this a shot.

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Sun Jun 30, 2019 4:55 pm

Zoolimar wrote:This part is literally impossible to have unless the resources that ships use for upkeep are not coins. No matter how high or low are the upkeep costs you always could use them all.

Sure its possible :)

It is even currently like this on stable version and was even more true prior to that: sure ships had a ridiculously low upkeep of 1 coin increasing by +1 per hull size tier but it was and still is impossible on stable to run out of coin for upkeep when normal precautions are met meaning a minimum of one market per planet and fully utilized trade pool

Like I said: fully agree on giving higher upkeep in general to ships so we cannot keep buying out every building, and especially for the largest ship hulls so that the AI does not keep mass-producing these (as it is both inefficient and not fun)

So here is my current example:
+ 269 Taxes
+ 154 Trade
- 340 Ships
- 92 Buildings
- 5 Ground Units
- 18 Hostile Environments
- 10 Agreements
+ 11 Food surplus
Total: -21
Ship count: 51 (with the above mentioned distribution). 7 of these are transports so combat ship count is actually: 44
Population 152 B
24 Stars
28 Planets
Metal: 865 (312)

As visible, I have just stopped producing ships not because of Metal shortage but due to upkeep costs (boarded a lot of medium ships from Marauders and Yoral that stacked up the costs)

The upkeep I think is a bit too much here: At turn 129 and for the above empire size, the player should probably be able to field twice as many ships (currently this is around 1.5 ships per planet). Sure its not a good comperison, but think abouth this: we currently have a single low-tech planet with 7.5 B pop and 3 fully operational spacecraft not counting Dragon and Starliner ;)

Solution: Halve the upkeep costs from 340 to 170:
Destroyer/Loght Cruiser: 3
Cruiser: 5
Battlecruiser: 11
Carrier: 17

This would mean that trade would still fully need to be utilized to ensure upkeep cost are covered, but ships could be produced as usuall and the surplus would not lead to regular buying out of improvements as even a mine/farm costs 150 and we havent even done any refit yet :)

Lets formulate this another way:
If we try to limit the number of ships via Coin, Metal is overriden as the defining resource for ship production and that is its only purpose

Production cannot effectively be stacked unless metal production can keep up with ship production but if coin limits production overall, then the equation is:
1. Coin limits max amount of ships
2. Metal limits the speed at which ships can produce (limits the amount of factories that can be built)
3. If coin limit is reached, metal starts to pile up, but it does not make sense to increase production = metal production is much less important while Coin is much more important as well as keeping all of its other functions (buyout, refit, bribe, purschase of slaves, mercs)
4. If player chooses to produce only cheap units to fully utilize ship "capacity", those also cost less metal, so again metal does not play much of a role

Also adjusted the graph to represent the situation if militarisation limit is reached in 15 turns (just an example, lets say this is when upkeep consumes all the available income when producing EC only)

Lets see what we get when only producing Light Carriers under such specific circumstances:
- Metal Consumption is higher by 66%: this seems considerable (and a bit unlogical) untill we check out the below benefits:
+ Upkeep is identical but consdiering the savings up untill the point where LC reaches EC upkeep we have surplus coin here in the end
+ 10 Light Carriers are produced compared to 3 Escort Cruisers
+ 66% more Defensive power overall
+ 250% more offensive power overall!
So we have to ask: what sense does it make to produce anything else if coin is the defining resource of ships?

According to the below, empires with more metal resource should build light cruisers and those with less metal, probably also LC and maybe Escort carrier if they are really lacking metal and still need/have storng production short term (an odd combination :)).
Ideally I would say you should build both but this secondary limitation (besides metal) can easily cause strange behaviour because of the below

I would suggest intead so to leave metal as the single defining resource of ship production speed AND capacity and balance coin instead to refit costs and maybe to panetary buyout so the later cannot occur frequently
Attachments
Limit Reached.GIF
Limit Reached.GIF (24.06 KiB) Viewed 12141 times
Last edited by zolobolo on Sun Jun 30, 2019 5:38 pm, edited 4 times in total.

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Sun Jun 30, 2019 5:15 pm

sven wrote:Could you upload this game (if you still have it)? I've made a bunch of changes to the logic around triggering attacks recently, and it's possible you've just found a bug.

Sure can: game_8302

User avatar
Zoolimar
Posts: 16
Joined: Sat Jun 29, 2019 11:42 am

Re: Testing Economic Balance Changes

Postby Zoolimar » Sun Jun 30, 2019 6:16 pm

The upkeep I think is a bit too much here: At turn 129 and for the above empire size, the player should probably be able to field twice as many ships (currently this is around 1.5 ships per planet). Sure its not a good comperison, but think abouth this: we currently have a single low-tech planet with 7.5 B pop and 3 fully operational spacecraft not counting Dragon and Starliner

1.5-2 Ships per planet seems like more than enough. Make a little more incentives to use mixed fleets and a number drops to a more comfortable ~30 with cruisers/carriers replacing some of the destroyers/light cruisers.

3-4 full fleets or 2-3 full fleets plus a few fast reacting detachments. For an empire that owns/controls around 1/3 of the galaxy that seems to be more than enough to be able to wage war on 2 fronts if the need arises.

It is even currently like this on stable version and was even more true prior to that: sure ships had a ridiculously low upkeep of 1 coin increasing by +1 per hull size tier but it was and still is impossible on stable to run out of coin for upkeep when normal precautions are met meaning a minimum of one market per planet and fully utilized trade pool

Technically you still can exhaust it. You just need certain dedication :D

If coin limit is reached, metal starts to pile up, but it does not make sense to increase production = metal production is much less important while Coin is much more important as well as keeping all of its other functions (buyout, refit, bribe, purschase of slaves, mercs)

That's actually not that bad, unless your production capabilities are explicitly tailored to your mining capabilities. On the other hand if you have more production than mining and allow the metal to pile up you would be able to replace losses during war faster than your opponent. I prefer to play like this as it allows more space for dealing with mistakes or surprises.

I would suggest intead so to leave metal as the single defining resource of ship production speed AND capacity and balance coin instead to refit costs and maybe to panetary buyout so the later cannot occur frequently

So the more ships you have the more time you need to build new ones? That seems like a recipe for having almost no markets at all. At least as long as metal needs to be amassed before you can build a ship and each built ship also needs metal upkeep.

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Sun Jun 30, 2019 7:41 pm

Zoolimar wrote:1.5-2 Ships per planet seems like more than enough. Make a little more incentives to use mixed fleets and a number drops to a more comfortable ~30 with cruisers/carriers replacing some of the destroyers/light cruisers.

Already have the carrier accounted for, and have mixed fleets. On these map settings I rarely get by with less then 3 fully combat ready fleets - I can currently afford these but not the rapid response reserves I normally use against Raiders (and pirates and harpies in earlier builds). Its a quetion of preference of course how much is enough. The US currently has around 10 fully operational carreir strike groups and it "only" needs to content with Earth the better part of which is allied with it.

2-3 carrier groups is not a great proposition for the largest empire in the galaxy ;)

Zoolimar wrote:Technically you still can exhaust it. You just need certain dedication :D

Belive me there was no lack of dedication from my side :) I always put destroyer production to repeat build and mostly also a couple of medium hulls as well. I have never exhausted the coin production even so which is fine though I also never could get really beyond around turn 250

Zoolimar wrote:That's actually not that bad, unless your production capabilities are explicitly tailored to your mining capabilities. On the other hand if you have more production than mining and allow the metal to pile up you would be able to replace losses during war faster than your opponent. I prefer to play like this as it allows more space for dealing with mistakes or surprises.

Yes, that is the most efficient way of producing: your production was wasted if it consumed more metal then your metal income and could grow exactly till that point till your metal intake is nullified: if there was surplus you could build more and/or bigger with more factories. Simple and makes sense

If coin can become a limiting factor, then factories should only be built until coin surplus is eliminated in which case both metal and the factories become irrelevant. They would only matter until lthe limit is reached and then metla just piles up and factories produce trade or science (wasting potential) untill something is destroyed and they need to/can produce again (micro)

Zoolimar wrote:So the more ships you have the more time you need to build new ones? That seems like a recipe for having almost no markets at all. At least as long as metal needs to be amassed before you can build a ship and each built ship also needs metal upkeep.

Building speed is only indirectly affected by already cosntructed ships: we are not talking abouth a theoetical concept here: this is stable build and up till then
Max building speed is achieved when metal income = production capacity (labor). Surplus labor is mostly wasted as any overconsumption of metal will quickly deplete reserves.
It still makes sense to build up reserves if you want to: it is possible and not discouraged altogether if someone wants to do this for flexiblity (a good mechanic that leaves the decision up to the player what to prioratize)
But overall fleet "capacity" is determined by overall metal production: the more metal you dig out from the soil, the more ships you can build. There is no metal upkeep for existing ships

This is a simple system, and if one wants to build a lot of ships, they need to make sure they produce enough metal and have enough factories to produce them quickly enough but withouth exceeding /Turn yield

Add coin limitation into the equation and metal will become irrelevant: we can see from the graphs that metal consumption should even decrease so it will be less of a relevant resource if at all: if we produce les ships/or less metal intensive ships, metal consumption will reduce that is guaranteed.

It might also hit the AI of course: in my current game, the second largest empire owning 24 planets, can currently only maintain 25 ships: granted: this is due to its production concentrated on building Light Carriers but in this case they currently only field 1 proper fleet and one dead-in-the water carrier with a single escort ship - for their size this is not a guarante for longevity ;)

Maintenance costs would solve the exploits involving boarding mechanics and would definitively reduce the number of large ships which is cool
But they might also make medium ships not viable, limit overall number of ships to a point where planetary defenses are OP (and annoying), player cannot build a large number of fleets even if they want to sacrificy food and research production, make trade less relevant, make metal irrelevant and reduce the competitivenes of the AI. The script of course can be given bonuses to not be crushed by the upkeep costs (have you seen the rediculous bonuse the AI gets ?) but this might feel odd when playing as while the player is thoroughly dependent on finance hub-worlds to maintain a small amount of fleets, the AI would not be

I think it might be worth seeing where the problematic amount of ships come from that need to be adressed (under what settings) as I have never seen them. On my current stable-build game I have 35 planets and around 60 combat ships at turn 216 and have been-still producing ships constantly on 3 production-focused worlds (no boarding though this time)

User avatar
Zoolimar
Posts: 16
Joined: Sat Jun 29, 2019 11:42 am

Re: Testing Economic Balance Changes

Postby Zoolimar » Sun Jun 30, 2019 8:03 pm

But overall fleet "capacity" is determined by overall metal production: the more metal you dig out from the soil, the more ships you can build. There is no metal upkeep for existing ships

This is a simple system, and if one wants to build a lot of ships, they need to make sure they produce enough metal and have enough factories to produce them quickly enough but withouth exceeding /Turn yield

So you propose to literally have unlimited ship production that grows in speed as your empire grows with no upper limit? Well ,besides how much your computer can take. That seems like a recipe for disaster. With upkeep you have a clear cycle which drives you forward - develop planets -> build ships up to income -> conquer -> develop planets. Repeat.

I think it might be worth seeing where the problematic amount of ships come from that need to be adressed (under what settings) as I have never seen them. On my current stable-build game I have 35 planets and around 60 combat ships at turn 216 and have been-still producing ships constantly on 3 production-focused worlds (no boarding though this time)

That right there seems like too many. What are you doing with them? 30 should be more than enough to crush almost anyone and conquer the galaxy. As long as they are not pure destroyers/light cruisers.

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Sun Jun 30, 2019 8:34 pm

Zoolimar wrote:That right there seems like too many. What are you doing with them? 30 should be more than enough to crush almost anyone and conquer the galaxy. As long as they are not pure destroyers/light cruisers.

Its not a problematic number especially when you are fighting on 4 fronts all at once + putting down rebellions due to our contual efforts to harmonise the masses: I was testing if the Tinkers are OP with their refilling missiles (spoiler: they are but was fun :))

At this stage we have already won the game (the red area were enemy territory for the most part of the game), so the number of ships will not be going down anymore and they also had around 10 turns to pile up a bit

Technically the AI alliance has more planets, pop, coin, metal and research but since they are trying to rebuild their fleets with medium ships, they cannot get a critical mass together in time to put up a fight so yeah its done.

The reason why income is so "low" is that we are not trading with anyone due to beign at war with every everyone and all that - but you can see that there is no real danger of running out of coin, metal on the other hand already does the job for you
Attachments
Situation.png
Situation.png (1.01 MiB) Viewed 12123 times

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Testing Economic Balance Changes

Postby sven » Mon Jul 01, 2019 12:40 am

zolobolo wrote:Note: back when AI was pumping out capital ships only I have already crunched the numbers and found that small ships are already more cost efficient. This did not change but is made worse with upkeep cost increase.


So, the tactical game is complex enough that what exactly "cost efficient" means is hard to pin down. But, when harpyeagle did his AI mods, he put together a fairly carefully hand-tuned ad-hoc function for estimating the "combat power" of any ship design. Harpy's formula is based on combining an "attack" number with a "defense" number, using the formula "power=attack*(1+defense)".

Using harpyeagle's evaluation functions, I've pulled some numbers for calculating cost-efficiencies from zolobolo's recent game 8302 upload.

Here's a link.

A few notes: at this point in zolobolo's game, he's fighting Tinkers, so his ship designs are largely aggressively specialized anti-missile builds. These are probably not ideal representations of what a mid game Ashdar fleet will typically look like. Similarly, for the AI data tables, the AI's are using autodesigned ship loadouts, a min/maxing player could probably do better.

All that said, the trends are still pretty clear. The way the game is balanced right now, if we use a harpyeagle style attack x defense evaluation function, bigger ships are reliably "more efficient" than smaller ones. But, if we follow zolobolo's example, and think exclusively in terms of attack power, then the "best" designs are actually the destroyers / light cruisers, as they give you more bang for your buck.

I'm not certain those trends are totally wrong. Particularly in the mid/late game, destroyers don't have much staying power if they come under fire, so making them relatively efficient in a pure attack/upkeep sense is maybe not wrong. What I think perhaps is wrong inside the current numbers is the degree by which destroyers are outperforming late game ships in an attack/upkeep metric. Zolobolo's "Marines carrier" (an escort carrier with assault shuttles), scores a massive 9.2 for attack/upkeep, while a traditional fleet carrier autodesign build is sitting at 2.6 in the same metric, and zolobolo's light cruiser scores a 6.2.

I'm not quite sure what the "fix" is here. But, I think you can make the argument that 1) destroyers, light cruisers, and escort carriers are all a bit "under priced" in terms of their upkeep costs. 2) We do perhaps need the parts that are slotted into a hull to have a larger impact on it's upkeep; a torpedo destroyer outfitted with antimatter missiles is an offensive beast, while the same ship with nuclear missiles is much less so, and it would probably make sense, both in game balance and world building terms, if the high tech version cost significantly more to maintain than the low tech build.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Testing Economic Balance Changes

Postby sven » Mon Jul 01, 2019 1:52 am

sven wrote:The way the game is balanced right now, if we use a harpyeagle style attack x defense evaluation function, bigger ships are reliably "more efficient" than smaller ones. But, if we follow zolobolo's example, and think exclusively in terms of attack power, then the "best" designs are actually the destroyers / light cruisers, as they give you more bang for your buck.


Another interesting trend in the data is that while the attack/upkeep "efficiency advantage" is pretty extreme, if you look at the other costs for producing ships, it's much less pronounced. In particular, attack/labor cost is almost flat across most designs, while attack/metal cost does show an advantage for small craft, but, it's relatively modest compared to the attack/upkeep advantage. Bottom line, the attack/upkeep numbers are outliers compared to the efficiency scores for other resources, which is probably a sign that the current balance is off.

User avatar
Zoolimar
Posts: 16
Joined: Sat Jun 29, 2019 11:42 am

Re: Testing Economic Balance Changes

Postby Zoolimar » Mon Jul 01, 2019 4:41 am

All that said, the trends are still pretty clear. The way the game is balanced right now, if we use a harpyeagle style attack x defense evaluation function, bigger ships are reliably "more efficient" than smaller ones. But, if we follow zolobolo's example, and think exclusively in terms of attack power, then the "best" designs are actually the destroyers / light cruisers, as they give you more bang for your buck.

Ideally you'd want something like what Space Empires had - a "combat simulator" where you can pit any ships against each other, either with both sides being AI controlled or with one being controlled by the human. For fast tests.

Zolobolo's analysis is pretty on point and probably stems from a simple fact that SiS uses the I-go-You-go model. You could see similar behaviour and results in say tabletop Warhammer. If you can ensure that you attack first than survivability of your pieces matters much less than having overwhelming firepower. Tabletop wargames normally use alternating activations to deal with this, though it is not a guarantee when there is no cover on the field and one side uses long range weapons.

Zolobolo uses high mobility to ensure that he would attack first and where he wants to do it. Yoral use unlimited range of missiles to strike before the enemy can get close and so on. Actually would be interesting to see how his fleet would do against a full cloaked fleet of cruisers/command cruisers.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Testing Economic Balance Changes

Postby sven » Mon Jul 01, 2019 6:03 am

Zoolimar wrote:Ideally you'd want something like what Space Empires had - a "combat simulator" where you can pit any ships against each other, either with both sides being AI controlled or with one being controlled by the human. For fast tests.


Yeah, it's a feature that's been on my TODO list for years now. One of these days, I may actually find the time to code it up ;)

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Mon Jul 01, 2019 4:12 pm

sven wrote:Yeah, it's a feature that's been on my TODO list for years now. One of these days, I may actually find the time to code it up ;)

Wouldn't recommend it: if you have a low entry-level option to simulate battles, that reduces the value of galaxy battles. I have noticed this on myself in Total War games: by having instant access to all the factions and all of their units right from the main menu the allure of discovering and trying these units out was blunted. Simply put: a super dred has much more of an impact if it is an occurance which is hard to come by and only available in limited numbers

Plus the coding involved would need to be considerable as new GUI and selection mechanic would need to be implemented and any development time not adding to the main galaxy map gameplay reduces the effort going into that version (also consider bug-fixing effort thereon for the function)

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Mon Jul 01, 2019 4:37 pm

sven wrote:Bottom line, the attack/upkeep numbers are outliers compared to the efficiency scores for other resources, which is probably a sign that the current balance is off.

I would rather say: the current blaance is very delicate and allows for production of both small, medium and large craft.
What I am trying to say is that an invasive change like trying to controlling number of ships via coin will mess up this balance considerably.

Whatever formula is used for coin maintenance cost calculation, if it limits an empire with a typcial balanced economy (at leat one market per planet and trade pool full), in how many ships it can upkeep that will break this balance and lead to anomalies I was trying to describe and showcase in the second set of graphs. Please note: amount of ships is already being limited by metal, I think that a more safe way for duing such rebalancing would be that resource already dedicated for this function
sven wrote:So, the tactical game is complex enough that what exactly "cost efficient" means is hard to pin down.

Agreed.
I am making a lot of simplifications but none would read these posts if we really get into the details. e.g.: in case of large ships we would also need to consider shield recharge rate, ship size and accuracy reduction thereof, critical mass etc... Yes, a fleet of 3 battleships and 5 Escort Cruisers will beat 20 Light Cruisers in most cases, but they will not be able to engage them as the LCs will be spread out and blockade all of their systems untill the doomstack is borken up and destroyed peacmeal - this I try to represent via the unit count attribute advantage

There are also mutiple examples where LCs can simply outperform any amount of large ship hulls as they are bar far the most cost AND time effective way of producing PD capacity which is optimal to counter:
- Rockets,torps
- Small crafts: fighters, bombers and even Shuttles (making them the best anti-boarding option as well)
In the above cases, the whol Defense Power advantage of larger ship hulls can dissapate as the hard-counter PD simply nullifies enemy attacks
That is also why ligth cruisers or destroyers are a common staple of my fleets (besides beign realisitc and me enjoying the designs :)): you always need PD capacity as you will encouter at least one of the above

The later is mostly an issue though of Escort Cruiser designs: they are less efficient PD vessels then smaller crafts and since large vessels always provide PD capcity, they can simply replace them in this regard

Like you said: it is complex enough that we cannot get into all the details but all things considered you see my point: an additional limitation focusing on number of ships will tilt all of these exising mechanics and you would likely need to rebalance a considerable number of attributes if both small, medium and large hulls are to be expected to remain viable throughout the game (its like the introduction of derivatives to the finance market)

Unless you only want players to produce the smallest or the largest vessels available: these two extreemes are simple to achieve but again can be done via Metal better
Last edited by zolobolo on Mon Jul 01, 2019 5:17 pm, edited 1 time in total.

zolobolo
Posts: 1544
Joined: Fri Nov 25, 2016 3:49 pm

Re: Testing Economic Balance Changes

Postby zolobolo » Mon Jul 01, 2019 4:46 pm

Zoolimar wrote:would be interesting to see how his fleet would do against a full cloaked fleet of cruisers/command cruisers.

Light Cruisers must stay in distance: the cloacked ships will exit cloacking even when they are not in optimal range and thus will waste their turn when firing first

In case of Marine Cruisers, there is nothing they can do: boarding ignores cloacked ships


Return to “Testing”

Who is online

Users browsing this forum: No registered users and 59 guests

cron