FreeOrion

Forums for the FreeOrion project
It is currently Thu Dec 14, 2017 10:51 pm

All times are UTC


Forum rules


Always mention the exact version of FreeOrion you are testing.

When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.



Post new topic Reply to topic  [ 19 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: Tue Oct 11, 2016 4:42 am 
Offline
Krill Swarm

Joined: Wed Jan 01, 2014 10:02 pm
Posts: 12
Hi all,

I downloaded and played this game last maybe 2 or 3 years ago, I just did again, amazing game, amazing art/sound! Also I noticed a lot of improvements in balance, it's much more fun now.

I thought I'd provide my thoughts/feedback insofar as it might be helpful.

1. Currently winning at combat seems to be a matter of spamming the ship with the most buffed stats, as has been mentioned apparently elsewhere on the forum there's no "rock/paper/scissors" aspect to it, or more specifically, there's no counterplay to a player spamming the same ship. It looks like there's some talk of fighters, this would be good as it could penalize too much reliance on capital ships due to overkill.

2. The organic line of ships are cool, but they still seem slightly underpowered, if you invest in the tech/production to build them then your empire will be at a disadvantage militarily, it would be nice if there were some way to offset this. I noticed there's a solar web ship part for the organic hull/laser combination, I found the description of how it works confusing, does this mean that each ship in the fleet gains attack related to the size of the fleet? If so, then this could make fleets of organic ships stronger, but why not generalize to other organic-type hulls and non-laser weapons?

3. I feel like in relation to military strength, tech is frequently too strong. For example, in the early game, the player has the "destroyer" ship design available, with stats 6/12 for 78PP. However for 48 research points you can unlock the robotic hull, and get a ship with stats 9/31 for 109PP. One metric to measure value would be to multiply attack x defense and divide by cost, but this metric the robotic is about 60% more resource efficient than the destroyer, so there's almost no conceivable situation in the game where building a destroyer would be a good idea. Some other examples of this exist. An observation that can be made about this is that due to the combat model, when two fleets have the same stats, the fleet with fewer, stronger ships is currently stronger. Therefore I think there's no reason for tech to provide such massive gains in efficiency, if the previously mentioned robotic ship were 400 PP it would still be worth researching that tech.

4. Following the previous vein, in Free Orion weapons are expensive, armor is cheap. I'm not sure, but it's possible that if armor and weapons cost the same per point then it would help fix the previous dynamic. Also, it's generally not worth it to make small corvette/fighter type ships, because again weapons being so expensive vs. cheap armor, if you're going to buy the weapon, why not load up on cheap armor to protect it? And because armor is so cheap, this means that the only limit on the amount is slots on the hull, making unlocking new hulls very strong as in the example in the previous point. I think it's possible that models where a point of armor = a point of damage in PP might open up more interesting types of play.

5. As the game goes to the 150-or-so turn mark, empire management can become taxing, this is true I think of all 4x-style games. The feature allowing the user to rally outpost/colony ships is huge, without it it was impossible to remember where your colonies were supposed to go. Idk if this is very difficult to add, but more features like this would be great, for example, what if when you clicked on a planet you wanted to colonize, you could select the species you want and it would cue a colony ship at the nearest planet with a shipyard with that species? Or what if you could cue "create outpost base/colonize gas giant/build gas giant generator", or cue a basic shipyard and an orbital drydock immediately after? Depending on how the code is structured I realize that these might be difficult but any of these or similar things would be great.

6. Would it be possible to allow the computer to concede earlier after it's in a clearly unwinnable position? Some games I've clearly "won", but will have to micromanage empire/invasions etc. for another 200 turns or so to get the victory. This isn't really a big deal, though, and probably would be better if options were better to cue actions and require less turn-based micromanagement.

Anyways these are my thoughts after playing pretty intensely all weekend and loving it. Thank for making this!

EDIT: There's one more thing I forgot to mention

7. I think that stealth/detection has a lot of potential, but I'm not a huge fan of how it's currently implemented as a global mechanic, i.e., currently if one empire wants to invade another empire's planet but is behind on detection, there's nothing they can do but wait 10 or 20 turns to get a higher detection level. For a different approach, what if there were stealth ship components and detector ship components that "stacked"--i.e., a ship with a large number of detector modules would have more detection than one with fewer modules? Idk, there's probably more to think through here it terms of design but I feel it's more interesting than the global effect as it is now.

Again, thanks for making the game :D


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 8:46 am 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3291
ccla wrote:
I downloaded and played this game last maybe 2 or 3 years ago, I just did again, amazing game, amazing art/sound! Also I noticed a lot of improvements in balance, it's much more fun now.
Cool, getting things balanced has been my main objective since I joined the team and while we're nowhere near there, that it is a noticeable improvement is a good thing, so thanks.
Quote:
1. Currently winning at combat seems to be a matter of spamming the ship with the most buffed stats, as has been mentioned apparently elsewhere on the forum there's no "rock/paper/scissors" aspect to it, or more specifically, there's no counterplay to a player spamming the same ship. It looks like there's some talk of fighters, this would be good as it could penalize too much reliance on capital ships due to overkill.
Yup, I agree with this and hopefully we can get Fighters done soon(ish), then we can move forward to balance stuff more—because they're a fairly big change in the balance of combat, doing further balance work before they come in was fairly pointless, I'm fairly sure more changes will happen post their implementation (you can, in theory, test them from the Fighters branch but the work isn't finished yet).
Quote:
2. The organic line of ships are cool, but they still seem slightly underpowered, if you invest in the tech/production to build them then your empire will be at a disadvantage militarily, it would be nice if there were some way to offset this. I noticed there's a solar web ship part for the organic hull/laser combination, I found the description of how it works confusing, does this mean that each ship in the fleet gains attack related to the size of the fleet? If so, then this could make fleets of organic ships stronger, but why not generalize to other organic-type hulls and non-laser weapons?

Badly explained, the Solarweb concept works with all Organic Line hulls, not just the basic organic hull, we're aware this isn't explained at all well in game and it's open as an issue tog et a rewording/expanded explanation. There aren't, currently, any plans to expand this further as the current work confuses the AI and they need updating to take into account variable effects, so the Web while playable is currently overcosted, it will be reduced in PP cost when the AI can take it into account/use it.

Also, Organic line is supposed to work well with Stealth and that currently needs work, it's not my immediate priority but hopefully when we've time to rework/expand Stealth this'll also strengthen the line a lot. Having said that, I find ORganic stuff, while harder to use, a lot more fun: think of everything that requires the Cellular Growth Chamber as a sort of submarine, load it up on stealth parts and don't bother with armour, can be very effective and very fun, as long as you work fast, the AI has increased detection tech priority recently and you need to keep ahead of their detection for the ships to not die horribly.
Quote:
3. I feel like in relation to military strength, tech is frequently too strong. For example, in the early game, the player has the "destroyer" ship design available, with stats 6/12 for 78PP. However for 48 research points you can unlock the robotic hull, and get a ship with stats 9/31 for 109PP. One metric to measure value would be to multiply attack x defense and divide by cost, but this metric the robotic is about 60% more resource efficient than the destroyer, so there's almost no conceivable situation in the game where building a destroyer would be a good idea. Some other examples of this exist. An observation that can be made about this is that due to the combat model, when two fleets have the same stats, the fleet with fewer, stronger ships is currently stronger. Therefore I think there's no reason for tech to provide such massive gains in efficiency, if the previously mentioned robotic ship were 400 PP it would still be worth researching that tech.
The Destroyer/basic line of hulls are, deliberately, overcosted—you can build them with just a shipyard and they require no research, I'm not sold I have the balance right yet, but I plan to do another pass once Fighters are in the game and might expand some of their stuff a bit.
Quote:
4. Following the previous vein, in Free Orion weapons are expensive, armor is cheap. I'm not sure, but it's possible that if armor and weapons cost the same per point then it would help fix the previous dynamic. Also, it's generally not worth it to make small corvette/fighter type ships, because again weapons being so expensive vs. cheap armor, if you're going to buy the weapon, why not load up on cheap armor to protect it? And because armor is so cheap, this means that the only limit on the amount is slots on the hull, making unlocking new hulls very strong as in the example in the previous point. I think it's possible that models where a point of armor = a point of damage in PP might open up more interesting types of play.
Armour is definitely undercosted currently, that will change. Personally I tend to build glass cannon designs for the most part, but I'm fairly sure Fighters will force me out of that as well.
Quote:
5. As the game goes to the 150-or-so turn mark, empire management can become taxing, this is true I think of all 4x-style games. The feature allowing the user to rally outpost/colony ships is huge, without it it was impossible to remember where your colonies were supposed to go. Idk if this is very difficult to add, but more features like this would be great, for example, what if when you clicked on a planet you wanted to colonize, you could select the species you want and it would cue a colony ship at the nearest planet with a shipyard with that species? Or what if you could cue "create outpost base/colonize gas giant/build gas giant generator", or cue a basic shipyard and an orbital drydock immediately after? Depending on how the code is structured I realize that these might be difficult but any of these or similar things would be great.
Don't colonise with ships, use Outposts and then use the colony building mechanic, this was explicitly introduced to solve this problem, we haven't deprecated colony ships enough yet but plan to do so, hopefully 90%+ of colonisation should be by outpost/building which makes things a lot easier.

Also, the Objects window has had a LOT of love last few cycles, if you filter it to just show your stuff and add extra columns you can do pretty much anything from it, including queuing buildings and ships—not finished yet, and the UI needs some work to make what you can do with it clearer, but it really speeds things up.

We did recently allow expansion shipyards to be queued if you've got the prerequisite, but that's only for stuff beyond the basic shipyard, further work is probably going to happen there but again, requires an AI update and it's a lower priority.
Quote:
6. Would it be possible to allow the computer to concede earlier after it's in a clearly unwinnable position? Some games I've clearly "won", but will have to micromanage empire/invasions etc. for another 200 turns or so to get the victory. This isn't really a big deal, though, and probably would be better if options were better to cue actions and require less turn-based micromanagement.

We are, definitely, hoping to massively expand diplomacy in general. Well, when I say we, I mean the people actually capable of coding this stuff, it's beyond me. But yes, an alliance and/or surrender mechanic definitely needed, mopping up the last few outposts is a pain and the end game is reduced as a result.
Quote:
7. I think that stealth/detection has a lot of potential, but I'm not a huge fan of how it's currently implemented as a global mechanic, i.e., currently if one empire wants to invade another empire's planet but is behind on detection, there's nothing they can do but wait 10 or 20 turns to get a higher detection level. For a different approach, what if there were stealth ship components and detector ship components that "stacked"--i.e., a ship with a large number of detector modules would have more detection than one with fewer modules? Idk, there's probably more to think through here it terms of design but I feel it's more interesting than the global effect as it is now.

Agree with this: what I'm hoping for is refinements for stealth parts so that stealthed ships aren't obsoleted on a single enemy research point, and also to have detection parts also reduce the stealth of stuff in system over time. Exact details nowhere near ready yet, but stealth has so much potential, it's just not a top priority compared to Fighters and general balance stuff.

_________________
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 1:57 pm 
Offline
Content Scripter
User avatar

Joined: Sat Mar 17, 2007 12:28 am
Posts: 685
ccla wrote:
2. The organic line of ships are cool, but they still seem slightly underpowered, if you invest in the tech/production to build them then your empire will be at a disadvantage militarily, it would be nice if there were some way to offset this. I noticed there's a solar web ship part for the organic hull/laser combination, I found the description of how it works confusing, does this mean that each ship in the fleet gains attack related to the size of the fleet? If so, then this could make fleets of organic ships stronger, but why not generalize to other organic-type hulls and non-laser weapons?
Did you take into account that the structure of organic ships grows? After 25 turns they will max out at double their initial structure.

ccla wrote:
4. Following the previous vein, in Free Orion weapons are expensive, armor is cheap. I'm not sure, but it's possible that if armor and weapons cost the same per point then it would help fix the previous dynamic. Also, it's generally not worth it to make small corvette/fighter type ships, because again weapons being so expensive vs. cheap armor, if you're going to buy the weapon, why not load up on cheap armor to protect it? And because armor is so cheap, this means that the only limit on the amount is slots on the hull, making unlocking new hulls very strong as in the example in the previous point. I think it's possible that models where a point of armor = a point of damage in PP might open up more interesting types of play.
Did you take into account that all weapons shoot three times each combat round? The effective damage per weapon is much higher than the attack stat.

_________________
All released under the GNU GPL 2.0 and Creative Commons Attribution-ShareAlike 3.0 licences.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 2:44 pm 
Offline
Krill Swarm

Joined: Wed Jan 01, 2014 10:02 pm
Posts: 12
Quote:
Did you take into account that the structure of organic ships grows? After 25 turns they will max out at double their initial structure.


I haven't done the exact math at this but my sense is that one empire is spamming organic ships and the other is spamming, say, asteroids, then the asteroid empire will win.

Quote:
Did you take into account that all weapons shoot three times each combat round? The effective damage per weapon is much higher than the attack stat.


On this one I think that the number of rounds of combat doesn't matter, my reasoning is that the effectiveness of the ship in combat is proportional to both its damage output and its survivability, the latter is proportional to the armor points, so the ship strength is approximately armor times damage, and this would be true even if the number of turns of combat were infinity, i.e., combat just ran until every side was eliminated except one.

But actually I think that even this doesn't tell the whole story because if a ship survives combat then it can be withdrawn and repaired, so in some ways armor is even better than I'm describing it to be.

I'd have to do more detailed analysis, but basically if armor were more expensive I suspect the game could become more interesting, and you might end up with more of a rock-paper-scissors dynamic. Again, I'd have to do the math, but I could imagine that small fleets of well-armored capital ships would be better countered by large fleets of glass cannons (due to overkill and "first strike" effect--larger fleet will probably do more damage first), but the large fleet of glass cannons could be countered by an even larger fleet of glass cannons with practically no damage per ship, which would be countered by the well armored capital ships.

Also, on the point of strategy diversity, this dovetails with my earlier point about tech being too good--one area is weapons. If you look at mass driver vs. laser, for instance, even at level 1 laser has 40% higher damage than mass driver, but is only about 33% more expensive. This doesn't seem so bad, but when you get to level 4 laser is almost twice as good. The reason I think this might matter is that even if hypothetically the cost scaled with the damage, or maybe even the damage was slightly behind, the tech would still be worth researching because it would save slots on the hull, which in the context of the armor discussion earlier are extremely good. And in the context of opening up lines of play, there might be situations later where you would use, say, mass driver instead of death ray, because you want more damage output per unit production cost and are willing to sacrifice hull slots to do it. But in the current game this kind of tradeoff doesn't exist.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 4:59 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4296
Location: Sol III
ccla wrote:
6. Would it be possible to allow the computer to concede earlier after it's in a clearly unwinnable position?
That belongs to the whole diplomacy part, which hasn't even entered early design stage yet. There is a very rudimentary framework implemented which allows for making peace/declaring war, and IIRC Geoff has already done some preliminary work wrt empire-species/species-species relations, but that more or less is it.

But of course, once we get to that part of the game, things like surrender will definitely be included.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 5:10 pm 
Offline
Content Scripter
User avatar

Joined: Sat Mar 17, 2007 12:28 am
Posts: 685
ccla wrote:
Quote:
Did you take into account that all weapons shoot three times each combat round? The effective damage per weapon is much higher than the attack stat.
On this one I think that the number of rounds of combat doesn't matter, my reasoning is that the effectiveness of the ship in combat is proportional to both its damage output and its survivability, the latter is proportional to the armor points, so the ship strength is approximately armor times damage, and this would be true even if the number of turns of combat were infinity, i.e., combat just ran until every side was eliminated except one.
Proportional is just not enough to describe the interaction.

For example a 10/4 ship will always destroy a 1/30 ship in single combat and survive.

(Not to mention that the 1/30 ship will be useless against a ship with shields.)

_________________
All released under the GNU GPL 2.0 and Creative Commons Attribution-ShareAlike 3.0 licences.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 5:21 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4296
Location: Sol III
ccla wrote:
On this one I think that the number of rounds of combat doesn't matter, my reasoning is that the effectiveness of the ship in combat is proportional to both its damage output and its survivability, the latter is proportional to the armor points, so the ship strength is approximately armor times damage, and this would be true even if the number of turns of combat were infinity, i.e., combat just ran until every side was eliminated except one.
I don't think this is exactly accurate. First, number of rounds certainly do matter, because the more rounds combat has, the more rounds a ship needs to be able to survive (especially against a superior force) before being able to retreat for potential repairs. Which in turn means the number of combat rounds influence how valuable armor is.

But even if not taking that into account, I still think one point of firepower is substantially more valuable than one point of armor. Simply because a ships full firepower can be "reused" each combat round until the ship is completely destroyed, whereas a point of armor is only "usable" once. If it has been "consumed" to absorb damage dealt by a hit, it's gone (until repaired of course). Firepower points aren't "consumed".

Or, to put it a bit differently, firepower is "regenerated" each combat turn to max. Armor is not. That's a big difference.

And finally, weapon and armor parts both require the same slot type, so they compete for hull space (so to speak). Armor not only costs PP, it also costs firepower. For each armor part you place on your ship, you have to give up a weapon part.

TBH, I don't think that even considering the current armor costs, that armor is too cheap. When I design combat ships, I find myself considering each time if I should put one more piece of armor on a ship, or go with more weapons. Armor is cheaper, but sacrificing firepower for it is a high price for the extra structure IMO.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 5:56 pm 
Offline
AI Contributor

Joined: Tue Feb 17, 2015 11:54 am
Posts: 224
Sloth wrote:
ccla wrote:
Quote:
Did you take into account that all weapons shoot three times each combat round? The effective damage per weapon is much higher than the attack stat.
On this one I think that the number of rounds of combat doesn't matter, my reasoning is that the effectiveness of the ship in combat is proportional to both its damage output and its survivability, the latter is proportional to the armor points, so the ship strength is approximately armor times damage, and this would be true even if the number of turns of combat were infinity, i.e., combat just ran until every side was eliminated except one.
Proportional is just not enough to describe the interaction.

For example a 10/4 ship will always destroy a 1/30 ship in single combat and survive.

(Not to mention that the 1/30 ship will be useless against a ship with shields.)


And 10*4 > 1*30. So the formula damage*structure predicts the correct outcome.


If you look at a combat-do-death situation in first order, damage*structure*enemy_dmg/(enemy_dmg-shields) is the combat strength of the ship. Second order effects come from the discretization effects and stochastic uncertainties when dealing with multiple, possibly different ships in a fleet vs fleet scenario.

Being able to retreat from a combat or minimizing damage taken or focussing on destroying enemy ships in the first 3 combat rounds are different additional considerations you may have once you are confident enough to win/survive the battle.
If you only take a look at first order effects, you want to maximize damage*effective_structure/cost for a single ship or damage*effective_structure/cost^2 for ships in a fleet.


@Vezzra: That regeneration argument doesn't hold. It breaks with the fact that you can only fire as long as you have armor - so a single point of firepower is worth more on a 100 structure ship than it is on a 10 structure ship because you will be able to shoot more often (again not taking into account second order effects).

_________________
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 6:06 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
MatGB wrote:
Armour is definitely undercosted currently, that will change. Personally I tend to build glass cannon designs for the most part, but I'm fairly sure Fighters will force me out of that as well.

There was a fairly long thread a while back that sort of convinced me the other way, that it was worth having more, cheaper, ships by ramping up the armor. In early game, that generally means robotic hulls with two weapons/two armor. In some games when my research levels are low I'll cap weapon development at Plasma 4 and Xentronium plate, which maybe Deflector shield. I need the research points for other things. With high levels of industry, you can churn out ships like that. However, I usually play games with 150 systems, so mine don't typically last that long. I have noted that the AIs tend to be more weapons-forward in their ship designs.

Quote:
Don't colonise with ships, use Outposts and then use the colony building mechanic, this was explicitly introduced to solve this problem, we haven't deprecated colony ships enough yet but plan to do so, hopefully 90%+ of colonisation should be by outpost/building which makes things a lot easier.

Why would you want to get rid of colony ships? Yes, in general, outposts are better most of the time, but there are use cases for the ships. I don't see a need to spend time getting rid of something that can be beneficial.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 9:29 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4296
Location: Sol III
Morlic wrote:
That regeneration argument doesn't hold. It breaks with the fact that you can only fire as long as you have armor - so a single point of firepower is worth more on a 100 structure ship than it is on a 10 structure ship because you will be able to shoot more often (again not taking into account second order effects).
To just point out that a point of firepower is more valuable (meaning, should cost more PP) than a point of armor I think the regen argument is sufficient. Of course it does not factor in that the value of your firepower depends (also, among other factors) on your structure, which means more armor makes your weapons more valuable (and the PPs spent for them more effectively invested). But even that does not work linearly, otherwise a design with only one weapon and as much armor as possible would always be superior to a design sacrificing armor for more firepower (provided armor is cheaper than weapons), which isn't the case. Which is why balancing weapon and armor parts on a ship design is not a no-brainer already, even with the current low armor costs.

Determining how much more valuable firepower is than armor is of course very difficult (I'd say even impossible because of the many factors influencing the worth of your firepower). But it is more valuable - if we made armor and weapons cost the same per point of capacity, I wouldn't bother with armor on my ships anymore.


Top
 Profile  
 
PostPosted: Tue Oct 11, 2016 11:26 pm 
Offline
AI Contributor

Joined: Tue Feb 17, 2015 11:54 am
Posts: 224
It is not sufficient.

Mathematically, what you need for your argument to hold is some "worth" of firepower that remains constant regardless of structure or at least scales slower than linear with it. For example, considering the proposed formula damage*structure, your argument would be completely invalid as the worth of fire power decays with the structure itself.

Quote:
Of course it does not factor in that the value of your firepower depends (also, among other factors) on your structure, which means more armor makes your weapons more valuable (and the PPs spent for them more effectively invested). But even that does not work linearly, otherwise a design with only one weapon and as much armor as possible would always be superior to a design sacrificing armor for more firepower (provided armor is cheaper than weapons), which isn't the case.


You are messing up math here. Again, consider
Code:
f(s,d) = s * d
with s being the structure and d the damage is a quadratic function in 2 variables and is maximized for a given sum of damage and structure if the two are equal. Note that the "value" of damage, i.e. the partial derivative of f(s,d) in respect to d equals the structure s (i.e. is linear in respect to the structure). You don't want to maximize the structure with low d however, as the opposite holds as well - the value of armor scales linearly with the existing firepower of the ship as well.

Now, if that formula is a good enough approximation for evaluating a design is another question and depends what other factors you want to consider. As a first order approximation for raw combat strength, I think the formula is fairly accurate especially given its simplicity. Note that shields can be considered as a third scaling factor or by replacing structure with the effective_structure (as shields effectively multiply your structure by some factor given by the enemy weapon strength and your shield value).

Obviously, if you know the exact enemy designs stats, abusing the discreteness of the numbers will almost always yield better results. If you already overpower the enemy, you may want to have more structure so you lose fewer ships. If you play with stealth, armor loses a lot of worth. If you want to use the ship without any fleet support, you want to weight the cost linearly, if you want to use it in a fleet, you whould weight it with the cost squared instead... That's why the formula can only be a first order approximation.

_________________
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0


Top
 Profile  
 
PostPosted: Wed Oct 12, 2016 1:52 am 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4296
Location: Sol III
Of course, if you determine the value of firepower and structure by f(s, d) = s * d, then firepower and structure are equally valuable. But, as you said yourself:
Morlic wrote:
Now, if that formula is a good enough approximation for evaluating a design is another question and depends what other factors you want to consider.
While it certainly is a good first order approximation for a rough, simple combat strength evaluation, I doubt it's sufficient to base cost balancing considerations on that. At this point however I have to admit that I lack the math skills to argue my point on mathematical grounds. I've to go by "gut feeling" and experience (with FO and other 4X space games I played). Even with armor being cheaper than weapons (on a per damage/structure point basis), going full armor and minimal weapons very rarely has been a good idea.

But even leaving the regen argument aside, the real question when it comes to balancing costs of weapons vs armor is how much value adds one weapon part vs. one armor part. And here the fact that each hull type comes with a certain amount of base structure, but no firepower becomes important. Because raising the structure of a hull from 20 to 30 by adding an armor part adds less value to the hull than raising its firepower from 0 to 10 by adding a weapon part. Factoring this in complicates things a lot of course, as the value of armor part vs weapon part heavily depends on the base structure of the hull, the strength of the armor and weapon parts in question, and the ratios of these values. One thing however is certain IMO: armor will never add the same value as a weapon. There might be situations where it comes close, but usually weapons will be clearly more valuable. Which is why I think armor must stay cheaper.


Top
 Profile  
 
PostPosted: Wed Oct 12, 2016 6:48 am 
Offline
AI Contributor

Joined: Tue Feb 17, 2015 11:54 am
Posts: 224
Yes, the base hull structure definitely adds a bias regarding the parts. I would argue (if armor and weapons cost the same) it holds up to the point where you added enough weapons to reach that bias. Then armor would become just as valuable as firepower again.

If weapons and structure cost the same, it is abusable in the sense that I can design ships with 1 structure but the same damage as you have structure and reach a draw in fights as both ships are destroyed and mine is cheaper. With that level of discretization abuse, structure would become worthless. These counter-ships then of course could be countered by adding a ship which just deals enough damage to destroy those and suicide them before the actual fight and because those counter-ships are cheaper and can't easily be countered by an even cheaper ship. First, because they are likely to be the cheapest available damage to do the job(potentially limited by hull choice) and second because they are somewhat protected by the rest of the balanced ships in the fleet that have enough structure to tank a few hits. So even in that edge case, the armor will likely hold value.

Now, having armor and weapons part cost (roughly) the same per point seems not a very fun approach in general (partly due to above issue), so balancing around that is not ideal either way, I perfectly agree with that. As long as both have its uses it does not matter if it is balanced in a mathematical sense as any player can choose to build the same designs. If using somewhat more armor or more weapons part is the best way to go does not really matter as long as you will use both parts in the designs throughout the game.

_________________
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0


Top
 Profile  
 
PostPosted: Wed Oct 12, 2016 8:41 am 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3291
Observation: when I balance stuff, I use the maths as a starting point then tweak based on gameplay experience, that's a lot easier with the AI's dynamic designing paradigm as I can see how they adapt as well. Just the pure maths is never going to be enough because different hulls have different abilities outside of combat (contrast eg: the Robotic Hull self repair against the Energy Frigate speed, in a straight up fight the robo hull will win, but they find it harder to catch the frigates, who can choose their ground more easier, etc).

I never got around to balancing the costs of armour vs weapons, especially not with refinements, it needs doing even if it's only minor tweaks or just staying the same. I did shields and I did hulls, but other parts no. And doing them right now is impossible, Fighters changes the paradigm a lot and makes high power high damage weapons less effective overall, etc.

_________________
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.


Top
 Profile  
 
PostPosted: Wed Oct 12, 2016 8:56 am 
Offline
Krill Swarm

Joined: Wed Jan 01, 2014 10:02 pm
Posts: 12
Wow, so many responses :D

Before going through specifics, I'll provides some basic justification for my armor times attack metric of ship strength...

Suppose ship 1 has damage d1 and armor a1, and ship 2 has damage d2 and armor a2. Let t1 be the time that it takes ship 1 to destroy ship 2, and t2 be the time it takes ship 2 to destroy ship 1.

Then
t1 * d1 = a2
t2 * d2 = a1
or
t1 = a2/d1
t2 = a1/d2

ship 1 wins if:
a2/d1 < a1/d2
or:
a1 * d1 > a2 * d2

This means that generally when two ships fight, the one with the armor*damage advantage will probably win.

Quote:
Proportional is just not enough to describe the interaction.

For example a 10/4 ship will always destroy a 1/30 ship in single combat and survive.

(Not to mention that the 1/30 ship will be useless against a ship with shields.)


When I said proportional to both armor and damage I meant the formula above, in this case 10 * 4 = 40, which is greater than 1 * 30 = 30. I'm ignoring shields right now.

Quote:
I don't think this is exactly accurate. First, number of rounds certainly do matter, because the more rounds combat has, the more rounds a ship needs to be able to survive (especially against a superior force) before being able to retreat for potential repairs. Which in turn means the number of combat rounds influence how valuable armor is.


Idk, I still think that ignoring the number of combat rounds makes sense. Look at it this way, consider the meeting of two fleets, f1 and f2, suppose that under the infinite round combat model (i.e., fleets fight until one is competely destroyed) f1 would win. If the number of combat rounds is finite, then the only effect this has is to halt the combat prematurely, if f1 would have won under an infinite round system, it will always be ahead under a finite round system for any nonzero finite number of rounds.

Quote:
And finally, weapon and armor parts both require the same slot type, so they compete for hull space (so to speak). Armor not only costs PP, it also costs firepower. For each armor part you place on your ship, you have to give up a weapon part.


We can actually quantify this problem, at least to a first approximation. When deciding how to allocate slots between armor and weapons, the following are factors:
*damage per weapons slot
*PP per weapons slot
*armor per armor slot
*PP per armor slot
(actually as it turns out damage per weapon slot and armor per armor slot don't matter under the assumptions used here, but this will be a result)

Let x be the number of slots used for weapons, y be the number of slots used for armor s be the total number of slots, then we have:

x + y = s

Let pw be the price per weapon slot, pa be the price per armor slot, and p be the total price, then:

x * pw + y * pa = p

Let dw be the damage per weapon slot, aa be the armor provided by an armor slot, and e be (slightly simplified) ship effectiveness as I defended in the earlier metric, then

dw * x * aa * y = e

If we're just trying to optimize ship effectiveness e, then we seek to optimize x * y under x + y = s where s is constant, the optimum solution is x = y, i.e., split slots evenly between armor and weapons. This all assumes that the base armor for the hull is zero, if it's substantially higher then this pushes the balance towards using more slots for weapons and fewer for armor.

However, usually you're not trying to optimize effectiveness e, but effectiveness per unit cost e/p, using y = s - x we get:

e/p = (dw * aa) * (s * x - x^2) / ((pw - pa) * x + pa * s)

Taking the derivate and solving for zero yields:
pw * x^2 = pa * y^2

Solving:
x = s / (1 + sqrt(pw/pa))

So according to the formula, if the ratio of cost for weapons to armor is 10:1, then the slots used for armor to slots used for weapons should ideally be around 3:1, and the damage per weapon or armor per armor (module) does not impact it. This ignores factors like shields and base hull armor, both of which push slightly towards more weapons. Another factor that pushes towards optimal e rather than optimal e/p (in this case more weapons) is that in free orion's combat model it's better to have fewer stronger ships than in a larger number of weaker ships. However, I suspect this formula is a reasonable first approximation.

Quote:
If you look at a combat-do-death situation in first order, damage*structure*enemy_dmg/(enemy_dmg-shields) is the combat strength of the ship. Second order effects come from the discretization effects and stochastic uncertainties when dealing with multiple, possibly different ships in a fleet vs fleet scenario.


Exactly what I'm saying, my thinking is that maybe if the cost per armor/damage were adjusted, then using these effects as well as others might play more of a strategic role in fleet design.

Quote:
You are messing up math here. Again, consider
Code:
f(s,d) = s * d

with s being the structure and d the damage is a quadratic function in 2 variables and is maximized for a given sum of damage and structure if the two are equal. Note that the "value" of damage, i.e. the partial derivative of f(s,d) in respect to d equals the structure s (i.e. is linear in respect to the structure). You don't want to maximize the structure with low d however, as the opposite holds as well - the value of armor scales linearly with the existing firepower of the ship as well.


I'm not sure why this shows my math is messed up, but your point is valid, the marginal cost of additional weapons is a function of current armor and cost per weapon, and true vice versa, so the comparison of marginal benefits of weapons vs marginal benefits of armor is nontrivial. However I think I just solved the optimization problem above, which I think addresses this and pretty much tells the whole story on the simplified "effectiveness = damage * structure" model


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 19 posts ]  Go to page 1, 2  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group