Patch: Robotic hulls rebalancing

Creation, discussion, and balancing of game content such as techs, buildings, ship parts.

Moderators: Oberlus, Committer

Message
Author
User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Patch: Robotic hulls rebalancing

#1 Post by MatGB »

Attached, first patch in the ongoing Ship and ship parts balancing project.

Patch does a number of things.

1) Updated en.txt to reflect current state of play, nothing controversial here it's stuff I missed on the last pass to en.txt and reflects already committed patches. It includes a minor clarification to the wording for the robotic interface shields that hadn't been clear to me during discussions as well.

2) Rearranges positioning within techs.txt putting all hulls of the same line together and adding headers for all hull lines making it easier to read and find entries, this in itself has no effect on gameplay but makes finding scripting a lot easier

3) Reorganises the Robotics hull tech line to split off the geo-integration facility hulls into a different line from the nanorobotics line, this was inspired by the way Organics currently handles things and was partially to allow access to the nanorobotic hull without researching the completely different flavour wise self gravitating hull. The research time and cost of the nanorobotic hull tech is slightly increased to reflect the lack of interstitial tech.

4) Reduces the costs of various robotic hull lines as well as the medium and large hull, specifically:

a) Medium hull is reduced to 20 pp, Large to 30 pp, making it 10/20/30 for small/medium/large, this keeps them as expensive hulls but slightly reduces the need to build an orbital incubator just to have cheap colony ships. In addition, speed of Large hull is increased to 60 as it's otherwise a complete waste of time and space, it was either that or remove it in my opinion. This puts the base cost of all three hulls at 10 pts per external slot, which is expensive but not excessively so.

b) reduced cost of Robotic Hull to 40, Self Gravitating Hull to 80, Logistics Facilitator to 100 and Titanic Hull to 180. This keeps each of these hulls as more expensive per slot than pretty much all other hull types, but not to the extent that they're unfieldable. Note the cost of the nanorobotic hull is unchanged, that was already a perfectly viable hull but other changes now make it even moreso.

I've played with this patch extensively over the last few weeks tweaking the numbers as I went. It is, with the numbers as is, possible to hold your own in the early game, build up and then overwhelm AI players that are using the still much cheaper asteroid and organic line hulls, but is hard work and requires a lot of thought, on one playthrough I used the exact same seed and played with each different line starting with organics then trying with asteroids, energy and then robotics. Note that apart from random specials the map was the same each time, and I played robotics last, but it still took me nearly twice as long as with organics, energy line (with organic hull for early game) was still the easiest by far.

Robotic hulls remain balanced and a little under powered with or without the robotic interface shields, I tested that extensively as well, I suspect that really needs its own tech to unlock it tied to robotic hulls and force field harmonics but that's not a priority for me personally, I definitely loved the flavour it adds.

I hope to give the Organics line a pass for cost balancing soon, that'll be a complete reverse as most hulls will need their costs increased to some extent—at which point it might also be needed to slightly tweak up the cost of the basic robotic hull, but I don't expect this and hope to avoid it, I think it's about right at the point I propose.

The most controversial thing I've done in here is, I suspect, changing the costs of the basic medium and large hulls, this was basically due to the observation that they're not used, Dilvish has the AI set to put incubators on all colonising worlds by default, and as it cuurently is that's the right decision, even though they're unlocked at the beginning they're simply too expensive to consider even as basic colony ships as is, the minor tweaking brings them into "expensive but usable" instead of "so expensive you want to avoid them" and I really do think it's necessary given the other options available early in the tech tree. It should be viable to use basic hulls in the early game and for later game colonising, currently it's foolish to do so, but it'll need rebalancing organics to fully acheive this.

Dilvish, regarding the AI and this patch, very little needs to be changed immediately as the only hulls they normally use are Medium hulls, that becomes more viable so it could be if you wished that Incubators become unnecessary on non-warship building planets, but submitting this as is will break nothing—hopefully this will let you think about having an AI go for robotics, but that isn't an urgent need, it's simply now a viable, if still underpowered, strategy.

This took more work than I was expecting, the hardest thing was finding the right cost for the basic robotic hull assuming it should be viable to compete in the early game with an AI that's using organic hulls without using them at all yourself, anythign above 40 simply meant they overwhelm with numbers in virtually every test I tried. with this applied, playing with robotic line ships is a fun challenge, and playing robotic species/robotic ships/robtic interface shields is good fun and requires interesting decisions regaring keeping fleets together or splitting up.

I'm fairly sure the patch is formatted correctly, LMK if there are any issues with it.
Attachments

[The extension patch has been deactivated and can no longer be displayed.]

Mat Bowles

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

User avatar
Ta'Lon
Space Squid
Posts: 66
Joined: Sat Feb 01, 2014 5:03 pm

Re: Patch: Robotic hulls rebalancing

#2 Post by Ta'Lon »

Thanks for your hard work. Downloaded, and I will look it over after Big Bang...
Conquering the galaxy, one planet at a time...

Any artwork that I submit for use in the graphics forum is submitted under the CC-BY-SA 3.0 License.

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Patch: Robotic hulls rebalancing

#3 Post by MatGB »

Bump.

I'd like to start makign a pass on part two of the project, sorting out organic & asteroid costs and doing somethign with energy line hulls, does anyone have any feedback on this or a reason to not commit it?
Mat Bowles

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

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Patch: Robotic hulls rebalancing

#4 Post by MatGB »

Geoff, this is the one that's pending, appears I forgot to put the I release this code and text under the GPL and CC licences FO uses but beyond that it should be good to go, doesn't hurt the AI in any way but does open up a new line of ships for competetive play (although we need to do something with the Trans Spatial Hull, as it stands it's pointless).

All came from discussion in FreeOrion • View topic - Ship and ship parts balancing and flavor ideas.
Mat Bowles

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

Vegavis
Space Squid
Posts: 71
Joined: Sun Jan 12, 2014 7:04 am

Re: Patch: Robotic hulls rebalancing

#5 Post by Vegavis »

LGTM. This is obviously better than the status quo, so any imperfections shouldn't stop it being committed. I look forward to the other rebalancings.

Yay, the Basic Large Hull is no longer useless! With two mass drivers and one Standard Armor, its sqrt(weapons*armor)/cost is slightly worse than the Basic Medium Hull, but its size makes up for that.

In the first game I won, I used mostly Robotic Hulls (because I didn't know better), so it's not impossible to win with them, even for a newbie. It was a small, crowded galaxy, though, so once I got a good start, I won economically regardless of ship choices.

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Patch: Robotic hulls rebalancing

#6 Post by MatGB »

Having played with this patch a lot, I think the costs from this are roughly right and the other hull lines should have prices increased to be on a par with Robotics in this patch, the game is a challenge, I sometimes find I'm overwhelmed by a single AI opponent building cheaper organics or asteroids, but if I can get past the initial rush the game is over at roughly turn 250 on a 350 system map, whereas Organics tend to be over way before that (150 isn't uncommon, 200 the norm) and energy even faster once they're built (fractals are seriously overpowered in so many ways, balancing them is going to be a PITA).

But the real challenge is going to be figuring out what to do with Organics, as they're not balanced within themselves, there are some hull types that you simply don't bother with, etc. Plus, the good Org hulls are so cheap ATM that anything done to nerf them needs to be done in line with Asteroids and, well, it's a much bigger job that I'll possibly need help with.
Mat Bowles

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

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: Patch: Robotic hulls rebalancing

#7 Post by Geoff the Medio »

If Dilvish doesn't object on AI grounds, then it sounds worth applying...

User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Patch: Robotic hulls rebalancing

#8 Post by Dilvish »

I don't have time to test this out now, but from Matt's description of the changes it sounds like I'd agree with him that it shouldn't actually hurt the AI. Though the AI wouldn't know (for now) to take advantage of some of these changes, it doesn't sound like it should be a very noticeable lack. So it sounds ok to me to go ahead and commit this.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Patch: Robotic hulls rebalancing

#9 Post by MatGB »

Thinking about it, there might be a slight, small, advantage for the AI in the early game, they sometimes build a lot of the Lynx line, these get slightly better with this. Beyond that, at some point you might want to give a Robotic line option to the current tech choice mix but it's still weaker than the others currently so it's not in any way needed.

But yeah, I deliberately went for the stuff the AI doesn't use first, and of course the reason the AI doesn't use it is because,w ell, it's terrible as is ;-)
Mat Bowles

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

Vegavis
Space Squid
Posts: 71
Joined: Sun Jan 12, 2014 7:04 am

Re: Patch: Robotic hulls rebalancing

#10 Post by Vegavis »

MatGB wrote:Having played with this patch a lot, I think the costs from this are roughly right and the other hull lines should have prices increased to be on a par with Robotics in this patch
Sounds good to me.
MatGB wrote:fractals are seriously overpowered in so many ways, balancing them is going to be a PITA
Presumably the number of slots should be reduced a lot. Even the Quantum Hull's ten external slots are overpowered. The big energy hulls can be good, distinctive ships without being ridiculously large.
MatGB wrote:But the real challenge is going to be figuring out what to do with Organics, as they're not balanced within themselves, there are some hull types that you simply don't bother with, etc. Plus, the good Org hulls are so cheap ATM that anything done to nerf them needs to be done in line with Asteroids and, well, it's a much bigger job that I'll possibly need help with.
It might be a big job to balance them exactly, but it needn't be done in one step (and can't, because future changes will change the balance). Organics are so misbalanced that even wild guesses will probably be better than the status quo. (The Organic Hull is a faster, more fragile equivalent to the Basic Large Hull, so they should cost about the same: 30, not the current 10. Right?) The sooner the initial changes are committed, the sooner other players can test them and improve them. So it makes no sense to wait until you have the balance exactly right.

AFAICT the controversy over organics is simply over how to nerf them — especially over whether to remove a slot. But I see no reason they shouldn't have their price corrected to begin with.

Balancing should not be postponed for the sake of the AI. If the AI depends on an imbalance, it's broken, because the game can't be expected to stay imbalanced. Instead, the AI should play as though the game is balanced. If this produces suboptimal play, that's not the AI's fault — it's a balance problem, and it's not the AI's job to work around it.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: Patch: Robotic hulls rebalancing

#11 Post by Geoff the Medio »


User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Patch: Robotic hulls rebalancing

#12 Post by Dilvish »

Vegavis wrote:The Organic Hull is a faster, more fragile equivalent to the Basic Large Hull, so they should cost about the same: 30, not the current 10. Right?
Not necessarily about 30, though probably above 10, sure. The Organic Hull requires some research and a special shipyard, which had been considered to make it reasonable for the hull cost to be cheaper than the 'freebie' initial hulls along with the speed/structure tradeoff, and one of the more recent piece-wise balancing steps had been (to my recollection) to significantly increase the cost of the organic shipyard along with its tech. Having the hull cost be the same or about the same as a Basic Large Hull could also be a reasonable approach (among the many reasonable approaches), but then I would very much think that the tech cost and organic shipyard cost should go back down.
So it makes no sense to wait until you have the balance exactly right.
If you truly have any concern that anyone is waiting to get balance exactly right before making any changes, please just spend some more time reviewing the discussion in this board and the commit history :D
Balancing should not be postponed for the sake of the AI.
Patches for balance and other content changes are nearly always posted here for use and testing by our active playtesters to get some consensus they seem like a good idea & are not broken in some way, for a while before being committed. Since these content changes can generally be installed by anyone without needing to actually build FO, this hasn't been deemed to be a significant hindrance to testing by anyone involved enough to contribute to the balance discussion. I think the general feeling has been that it takes a great many players downloading and trying the prebuilt packages for each player that winds up being interested and motivated enough to get involved in contributing here. Since most of our players play against the AI rather than against other humans in multiplayer, and we've liked the additional momentum and player attention the game has gotten with the (at least somewhat) competitive AI, we've generally planned to delay committing changes that would cripple the competitiveness of the AI, until such time as the AI can be adjusted for it (like with the Supply changes that just went in).
If the AI depends on an imbalance, it's broken, because the game can't be expected to stay imbalanced. Instead, the AI should play as though the game is balanced. If this produces suboptimal play, that's not the AI's fault — it's a balance problem, and it's not the AI's job to work around it.
I'm having a hard time understanding what you're really trying to say here; the sentences seem a little bit in conflict. The AI takes the game as it is (except to the extent it is lagging behind changes and taking the game the way it was a while ago :lol: ) I keep being reminded that there are very few details about gameplay that I should expect to necessarily stay the same, whether they seem well balanced currently or not. And even for the things which we have currently identified as not having satisfactory balance, it's not as if any of us can somehow know what that future balance decision will be and design the AI to play according to that anticipated future balance point rather than what the game is at the time (and it doesn't seem like it would be a really good idea to try that even if it were possible).

The AI does now have some dynamic adaptability to certain content changes, specifically ship hull and parts cost and primary stats. So the AI will automatically adjust its ship design choices in response to content changes increasing or decreasing a ship hull cost or structure, or a weapon or armor value (for currently known hulls and parts). Its building construction and tech research priorities are more statically constrained, though, so if the strategic cost/value balance for one of those significantly changed the AI would be at a disadvantage until it could be manually adjusted for that. Putting that in the context of the current thread, if the cost of various hulls are increased or decreased, with AI will mostly be automatically responsive to that, but with the changes discussed above it would be over-prioritizing the organic hull tech and shipyard if those costs aren't also adjusted. So I would recommend including some reductions to the organic tech and shipyard cost at least initially, and then these changes could probably be made without much of an adverse effect on AI competitiveness. If you wind up deciding that the organic tech cost and shipyard cost should still be high, then hopefully that's something that could wait for me (or someone else) to adjust the AI for it.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Patch: Robotic hulls rebalancing

#13 Post by MatGB »

Aye, worth restating but I agree with all of that, if the AI starts being stupid for some reason I get bored and play less, so I never want to do anything that stops the AI being at least a moderate challenge. There are aspects of the game that're unbalanced, the AI doesn't rely on them, it does use the currently best available hull types (for example, I've never seen an AI build a Cellular Growth Chamber, does anyone use that line apart from maybe Sentient Hulls in the end game?).

If something we do makes the currently most power hulls, like Endomorphic, Heavy Asteroid, etc a lot less powerful in relation to other hull types then it would hurt the overall gameplay significantly.

So if I do put forward proposals that rework them significantly, I'd know they couldn't be put up without working with Dilvish to adapt the AI for them, etc. The way we've discussed before is a series of small incremental changes that aim towards a certain goal, if for example the cost of the Xenocoordination Centre is changed as a one off effect it'd make the AI less competetive for a bit but it wouldn't kill them, whereas tripling the cost of Organic Hulls and reducing their effectiveness would hurt it more, etc.

(I'm looking at a ballpark of 15 per, possibly 20 per, for the current ORganic Hull ship, but I now need to look at it in the round, the other thread on this has a lot more of the discussion on this).

There are also a few things that could do with a bit of consensus on before further movement, I'll start threads for them when I've got them clear in my own head.
Mat Bowles

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

Vegavis
Space Squid
Posts: 71
Joined: Sun Jan 12, 2014 7:04 am

Re: Patch: Robotic hulls rebalancing

#14 Post by Vegavis »

Dilvish wrote:If you truly have any concern that anyone is waiting to get balance exactly right before making any changes, please just spend some more time reviewing the discussion in this board and the commit history :D
ISTM the community does hesitate to make balance tweaks — that is, it treats even the simplest cost adjustments as potentially dangerous, even though they can't possibly break anything. It's reasonably willing to change mechanics in the interest of balance, but it's no more willing to make simple cost adjustments The cost tweaks in this very thread were presented as controversial: “The most controversial thing I've done in here is, I suspect, changing the costs of the basic medium and large hulls”. I don't understand how this could be controversial, since it was well known that the Large Hull was useless and the Medium Hull very poor.
Dilvish wrote:The AI takes the game as it is (except to the extent it is lagging behind changes and taking the game the way it was a while ago :lol: ) I keep being reminded that there are very few details about gameplay that I should expect to necessarily stay the same, whether they seem well balanced currently or not. And even for the things which we have currently identified as not having satisfactory balance, it's not as if any of us can somehow know what that future balance decision will be and design the AI to play according to that anticipated future balance point rather than what the game is at the time (and it doesn't seem like it would be a really good idea to try that even if it were possible).
I agree that the AI can't predict rule changes and shouldn't try. However, there is one balance property that's very predictable: prices should be accurate. Whatever the rules are in the future, we know the price of each technology or buildings or hull will roughly match its value.

So if something is underpriced (such as the Organic and Heavy Asteroid hulls currently), you can confidently predict that the price will rise. If something is overpriced (such as the Basic Large Hull, until today), you can confidently predict that the price will fall. If you tune the AI to exploit inaccuracies in the current prices, then you can predict that it will play worse when the imbalances are fixed. If you tune it to play as though prices were accurate, then you can predict it will play better when the imbalances are fixed.
Dilvish wrote:The AI does now have some dynamic adaptability to certain content changes, specifically ship hull and parts cost and primary stats. So the AI will automatically adjust its ship design choices in response to content changes increasing or decreasing a ship hull cost or structure, or a weapon or armor value (for currently known hulls and parts). Its building construction and tech research priorities are more statically constrained, though, so if the strategic cost/value balance for one of those significantly changed the AI would be at a disadvantage until it could be manually adjusted for that.
When the strategic choices are imbalanced, isn't it easier (and more robust) to make the AI expect balance, and change them to fit that, than to change the AI to fit each transient imbalance?

User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Patch: Robotic hulls rebalancing

#15 Post by Dilvish »

Vegavis wrote:The cost tweaks in this very thread were presented as controversial: “The most controversial thing I've done in here is..."
FYI, the English phrase "the most controversial thing here" doesn't actually mean that anything is controversial, any more than "the most expensive component is a five cent capacitor" means that the five cent capacitor is deemed actually expensive. These phrases are a type of reassurance that something is either not bad at all, or at least not worse than X.
However, there is one balance property that's very predictable: prices should be accurate. Whatever the rules are in the future, we know the price of each technology or buildings or hull will roughly match its value.
I have to disagree -- price/cost is intended to be related to value, but it is far far from an actual measure of value. There are many factors that go into setting costs besides value (tech taken to enable, restricted build locations, etc.), and it can even be variable within a game -- is the value of Psionics to an empire having a telepathic species only 1/6 of the value of the same tech to an empire without a telepathic species? Value is nevertheless highly situational as well -- even if costs were intended to be purely a value measure their generally fixed nature would make them a poor strategic value assessment.
So if something is underpriced (such as the Organic and Heavy Asteroid hulls currently), you can confidently predict that the price will rise.... If you tune it to play as though prices were accurate, then you can predict it will play better when the imbalances are fixed.
I've already made the point that future balance decisions can't really be predicted-- having an opinion that one thing is currently more cost advantageous over another doesn't really say anything about whether the first is likely to become more expensive versus the second becoming less expensive versus either or both simply having their stats changed and costs remaining the same. And that's simply with an A vs B comparison, whereas the actual game balance issues are much more complex. You seem have a rather astounding degree of confidence in your ability to know the future; I certainly don't possess that kind of ability.

Furthermore, as Matt and I both pointed out, even if a future change were already 100% decided upon, we would not want to change the AI to be dependent on that change in advance of the change actually being made -- that would in our minds make the AI a perversely crippled player in the meantime.
When the strategic choices are imbalanced, isn't it easier (and more robust) to make the AI expect balance, and change them to fit that, than to change the AI to fit each transient imbalance?
Are you are saying that the AI should assume that all strategic choices are always of equal value or that the strategic value should be assumed to be equal to the cost? Or that I should code up the AI with some algorithm to try predicting what the eventual balance point of all content is going to be, and have the AI start playing according to its prediction of the future game, rather than according to the actual content it is presented with? That's probably not really what you mean, since it doesn't seem very 'robust', but I can't really figure out any interpretation that actually would make sense to me.

Keep in mind that it is a key part of FO that content be highly customizable by end users; it can only be expected that they will sometimes implement things that seem fun to them but that are not really balanced. The AI will be expected to play as good a game as possible in that circumstance (though it will probably always have rather constrained ability to determine the significance of novel user content). I will be trying over time to increase the extent of AI dynamic assessment of strategic value of various choices, but it is far more difficult than specific assessments of specific content and progress on that front is likely to be slow.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

Post Reply