Comsat/planetary defenses re-design

For what's not in 'Top Priority Game Design'. Post your ideas, visions, suggestions for the game, rules, modifications, etc.
Message
Author
User avatar
Dilvish
AI Lead, Programmer
Posts: 4712
Joined: Sat Sep 22, 2012 6:25 pm

Re: Comsat/planetary defenses re-design

#31 Post by Dilvish » Sat Sep 01, 2018 6:16 pm

Geoff the Medio wrote:
Sat Sep 01, 2018 10:33 am
...make combat target selection ignore unarmed ships (ie. those without direct weapons or fighter parts) unless there are no targetable armed ships present. Combat could still be used to take out troop ships (but only after all armed ships defending them are gone).
I'd be fine with most of that proposal, but I am a bit concerned about treating troop ships the same as decoys or scouts. In the large-grain treatment I would lean towards thinking that troop ships should probably count like armed ships, at least with respect to attacks by planetary defenses. If an attacker could count on troop ships not getting attacked until a round after which all armed ships had been destroyed, it would be easier to plan an attack that includes enough troopships to capture a planet, with just enough armed ships to take down shields and act as fodder to shield the troopships, which could then capture the planet. I'll acknowledge that the issue does kind of go both ways, troop ships could act as fodder to protect armed ships, and just like now, it is still possible to plan that kind of attack even if the troop ships are valid targets. But I have a different proposal to at least partially help with that--

Right now, when planets attack, they get just a single generally-very-strong attack. I'd propose that planet defense attacks be considered like an amount of energy per round, that can get divided between multiple targets-- if a planet with 40 defense targets a ship with just 10 structure, it would still have 30 more points of energy damage to inflict that round on other targets.

Another possible tweak could be that planets would prefer to only target armed ships so long as the planet has shields, but as soon as they lose shields they also start targetting troop ships even if armed ships are present (and that would be especially effective if the planet defenses could attack multiple targets like I propose).
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

Ophiuchus
Vacuum Dragon
Posts: 542
Joined: Tue Sep 30, 2014 10:01 am

Re: Comsat/planetary defenses re-design

#32 Post by Ophiuchus » Tue Sep 04, 2018 10:30 am

Dilvish wrote:
Sat Sep 01, 2018 6:16 pm
Right now, when planets attack, they get just a single generally-very-strong attack. I'd propose that planet defense attacks be considered like an amount of energy per round, that can get divided between multiple targets-- if a planet with 40 defense targets a ship with just 10 structure, it would still have 30 more points of energy damage to inflict that round on other targets.

Another possible tweak could be that planets would prefer to only target armed ships so long as the planet has shields, but as soon as they lose shields they also start targetting troop ships even if armed ships are present (and that would be especially effective if the planet defenses could attack multiple targets like I propose).
I also think that the single attack coming from the planet is not handled right at the moment. I think its intended to be the big cannon which is a danger to the ships you can't efficiently destroy with your smaller guns.
Maybe it should target armed ships it can destroy with a single shot first, then those it can destroy with two shots.
I also like the idea that the targetting changes with the shield. But i rather suggest that the planet prefers shooting troop ships as soon as the shields are less than half.
Also maybe we could add a flak or launch some interceptors from the planet to help getting system superiority.

The energy-distribution suggested by dilvish is interesting in that its different from ship weapons. It would be hacky to code though.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Morlic
AI Contributor
Posts: 248
Joined: Tue Feb 17, 2015 11:54 am

Re: Comsat/planetary defenses re-design

#33 Post by Morlic » Tue Sep 04, 2018 6:20 pm

Dilvish wrote:
Sat Sep 01, 2018 6:16 pm
I'll acknowledge that the issue does kind of go both ways, troop ships could act as fodder to protect armed ships
Which would at least be reduced if we increased troop pod price to be close to weapon cost and increase the capacity in return. This would also open up more interesting troop designs than spamming pods on every slot for maximized efficiency...
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

Ophiuchus
Vacuum Dragon
Posts: 542
Joined: Tue Sep 30, 2014 10:01 am

Re: Comsat/planetary defenses re-design

#34 Post by Ophiuchus » Tue Sep 04, 2018 9:17 pm

Morlic wrote:
Tue Sep 04, 2018 6:20 pm
Dilvish wrote:
Sat Sep 01, 2018 6:16 pm
I'll acknowledge that the issue does kind of go both ways, troop ships could act as fodder to protect armed ships
Which would at least be reduced if we increased troop pod price to be close to weapon cost and increase the capacity in return. This would also open up more interesting troop designs than spamming pods on every slot for maximized efficiency...
i actually sometimes add one occasional stealth or armor or choose a heavier hull for front line troops to soak some shots, especially if i cant risk leaving the troops outside the planet i am going to take over.

but +1 for at least closing the gap a bit more. but i think this needs proper balancing with invasion cost and the default troops so the differences between good_ average_ bad_ troops doesnt completely vanish.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

User avatar
em3
Vacuum Dragon
Posts: 508
Joined: Sun Sep 25, 2011 2:51 pm

Re: Comsat/planetary defenses re-design

#35 Post by em3 » Thu Sep 06, 2018 11:46 am

I have another suggestion regarding decoys:

Maybe have loosing ships en masse incur a cost in happiness, for example on the planet they were built on and/or at the allied sytem they were destroyed at (for human shield scouts)?

This would add another trade-off to consider when employing zerg tactics. Additionally, there could be species resistant to this cost (exobots, obviously, maybe some zealots or kamikaze-like species).
[...] for Man has earned his right to hold this planet against all comers, by virtue of occasionally producing someone totally batshit insane. - Randall Munroe, title text to xkcd #556

Ophiuchus
Vacuum Dragon
Posts: 542
Joined: Tue Sep 30, 2014 10:01 am

Re: Comsat/planetary defenses re-design

#36 Post by Ophiuchus » Thu Sep 13, 2018 2:17 pm

I just wanted to say i do not think that there should be no decoys. Actually i think having decoys should be a viable tactic. But it should not lead to a sitation you can "overpower" the enemy using decoys. Also i think the targetting system could lead to people building specifically designed decoys (e.g. to lure antimatter shots) which i think is actually a good thing. And maybe the counter to that could be research new battle scanner technology/policies or find new leaders.

To elaborate on the comsats. The base hull is the most powerful decoy base because of multiple reasons.
  • base hull does not even need a shipyard. so if you see enemy advancing, you can start build where you need. and also if you just captured a planet you can immediately start producing
  • base hull takes only 3 build turns, so you can produce them even as response
  • base hull was (is?) so cheap that you can just mass produce them. that worked quite well in the past even on blockaded planets.
  • while the flak at the beginning can clean up base hulls very efficiently; with the +5 hp bonus which is necessary for the hull to work against mines the flak is does not work so well anymore
  • in our combat system having many vessels is/was a very good distraction
  • the main drawback with having many "useless" ships is maintenance cost, but actually you can misuse the production system to produce vessels in advance without having them count for maintenance (->which leads to micromanagement)
From my playing experience.. since we have stealth bombers i mostly stopped using comsats ;)

regarding the happiness cost - while that would work against comsats, it would give an incentive to spread and manage ship production between multiple planets --> micromanagement, so i'd rather not have this.
instead could cost happiness or make other troubles for leaders though (as geoff envisions those to be few)
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

User avatar
Dilvish
AI Lead, Programmer
Posts: 4712
Joined: Sat Sep 22, 2012 6:25 pm

Re: Comsat/planetary defenses re-design

#37 Post by Dilvish » Thu Sep 13, 2018 4:28 pm

A couple responses, including a new idea, actually two new related ideas--
Ophiuchus wrote:
Thu Sep 13, 2018 2:17 pm
[*]base hull takes only 3 build turns, so you can produce them even as response
One thing we could look at changing to incrementally help is to make the build time for base hull be 4 turns rather than 3, to further reduce its suitability for last-minute deployment in the face of an incoming attack.
[*]base hull was (is?) so cheap that you can just mass produce them. that worked quite well in the past even on blockaded planets.
I think it is counterproductive to continue talking about how cheap they used to be, they have been triple that cost for a fair while now. Most blockaded planets (at least after Sent Auto) can make a few of these, but 'mass production' by a blockaded planet is less commonly feasible (depending on what level you mean by 'mass production'). And it sounds to me like we should try a further incremental change of increasing the base cost to 4 or 5.
[*]the main drawback with having many "useless" ships is maintenance cost, but actually you can misuse the production system to produce vessels in advance without having them count for maintenance (->which leads to micromanagement)
I am trying to figure out what 'misuse' you mean here-- 66%+ producing a batch of comsats and then pausing production, so they could be rapidly finished off when an attack threatens, but not adding to maintenance in the meantime? That may have been a very viable approach when they were really cheap, but it does carry a bit more noticeable now, since it ties up a chunk of production that might wind up being actually wasted.

Partly related to that, though. and just to the general tactic of deploying a bunch of comsats right before a battle so that they are immediately destroyed and only add to maintenance for a single turn, would be a change to tally up not just completed ships, but also include in the tally under-construction ships. That way a set of comsats would at a minimum be adding to maintenance costs for a period of 4+ turns, and it would entirely do away with the micromanagement issue just noted.

Furthermore, this discussion is also making me think about the possibility of not just doing a straight tally, but introduce the element that having a large number of ships at one location increases the logistics problems of trying to service them all at once, so we would generate the overall tally by first making a system-by-system tally, and at each individual system the local tally would be raised to a power, perhaps in the order of 1.3 - 1.5. At 1.3 a stack of 10 ships together would carry a maintenance cost of about 20 individual ships, at 1.5 it would correspond to about 32. Particularly in combination with the above rule about counting under-construction ships, I think that would really drastically limit the use of base hulls for decoys. It would also substantially penalize their use for troop drops, to the point where they would be seldom deployed but I think that would be acceptable (and in general, nearly anything we do to make large deployments of comsats less viable will have a similar effect on troop drops)..
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

Ophiuchus
Vacuum Dragon
Posts: 542
Joined: Tue Sep 30, 2014 10:01 am

Re: Comsat/planetary defenses re-design

#38 Post by Ophiuchus » Thu Sep 13, 2018 9:04 pm

Dilvish wrote:
Thu Sep 13, 2018 4:28 pm
Partly related to that, though. and just to the general tactic of deploying a bunch of comsats right before a battle so that they are immediately destroyed and only add to maintenance for a single turn, would be a change to tally up not just completed ships, but also include in the tally under-construction ships. That way a set of comsats would at a minimum be adding to maintenance costs for a period of 4+ turns, and it would entirely do away with the micromanagement issue just noted.
To be completely fair one should maybe take project progress into account (so a 30% finished ship takes 30%maintenance cost of a finished ship). Then on the other hand that would complicate maintenance cost calculation and prediction a lot and also maybe its good to have an incentive to have a short production queue (so sometimes drop a project rather than pausing it for a long time).
So id rather say: full maintenance cost for hulls and parts in the production queue.
This could even make production queue prediction easier as the maintenance cost does not change when ships are finished.
On the first hand again one who does long term planning (adding ships which are not going to be build in would get penalized)
So maybe: full maintenance cost for hulls and parts of already started projects(!?)
Dilvish wrote:
Thu Sep 13, 2018 4:28 pm
Furthermore, this discussion is also making me think about the possibility of not just doing a straight tally, but introduce the element that having a large number of ships at one location increases the logistics problems of trying to service them all at once, so we would generate the overall tally by first making a system-by-system tally, and at each individual system the local tally would be raised to a power, perhaps in the order of 1.3 - 1.5. At 1.3 a stack of 10 ships together would carry a maintenance cost of about 20 individual ships, at 1.5 it would correspond to about 32. Particularly in combination with the above rule about counting under-construction ships, I think that would really drastically limit the use of base hulls for decoys. It would also substantially penalize their use for troop drops, to the point where they would be seldom deployed but I think that would be acceptable (and in general, nearly anything we do to make large deployments of comsats less viable will have a similar effect on troop drops)..
Hm, so you need to spread out your fleets to keep maintenance cost low? Sounds ugly.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

User avatar
Dilvish
AI Lead, Programmer
Posts: 4712
Joined: Sat Sep 22, 2012 6:25 pm

Re: Comsat/planetary defenses re-design

#39 Post by Dilvish » Thu Sep 13, 2018 9:24 pm

Ophiuchus wrote:
Thu Sep 13, 2018 9:04 pm
On the first hand again one who does long term planning (adding ships which are not going to be build in would get penalized) So maybe: full maintenance cost for hulls and parts of already started projects(!?)
Good point, and that sounds to me like a decent tweak to the idea.
Hm, so you need to spread out your fleets to keep maintenance cost low? Sounds ugly.
That aspect is certainly the main drawback to the idea, I think. But it seems that we've seen a fair bit of unhappiness with the tendency towards doomstacks, and this would certainly create a disincentive from deploying like that. Perhaps it's too much of a headache-type disincentive though. And it's one of those many things that theoretically the AI could handle better than a human, yet I would not relish having to actually code up even a merely reasonable AI treatment for it. But perhaps it could just be focused more on decoys, or large stacks (say 5+ in size) of non-military/troop ships in general.
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
Vezzra
Release Manager, Design
Posts: 4657
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Comsat/planetary defenses re-design

#40 Post by Vezzra » Fri Sep 14, 2018 10:57 am

I feel the urge to toss a potentially bizarre idea into the mix: what about just discarding the base hull entirely? Personally I never liked it anyway, do we really need it? In what way would gameplay really suffer if we don't have that hull? Sure, you'd have to change a couple of strategies, or better, some strategies wouldn't be possible anymore, but in most of these cases I wonder if that would really be a bad thing, considering all the issues having that hull available causes...

Morlic
AI Contributor
Posts: 248
Joined: Tue Feb 17, 2015 11:54 am

Re: Comsat/planetary defenses re-design

#41 Post by Morlic » Fri Sep 14, 2018 5:45 pm

Then people would use a small hull or similar instead for the same purpose which costs trivially more.
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
Vezzra
Release Manager, Design
Posts: 4657
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Comsat/planetary defenses re-design

#42 Post by Vezzra » Sun Sep 16, 2018 1:32 pm

Morlic wrote:
Fri Sep 14, 2018 5:45 pm
Then people would use a small hull or similar instead for the same purpose which costs trivially more.
But can't be produced without a shipyard (IIRC), which significantly impedes the ability to just spam them quickly practically everywhere where you might need them. That alone would go a long way to alleviate the issues we're discussing here (IMO).

User avatar
Oberlus
Small Juggernaut
Posts: 703
Joined: Mon Apr 10, 2017 4:25 pm

Re: Comsat/planetary defenses re-design

#43 Post by Oberlus » Sun Sep 16, 2018 7:59 pm

Vezzra wrote:
Sun Sep 16, 2018 1:32 pm
Morlic wrote:
Fri Sep 14, 2018 5:45 pm
Then people would use a small hull or similar instead for the same purpose which costs trivially more.
But can't be produced without a shipyard (IIRC), which significantly impedes the ability to just spam them quickly practically everywhere where you might need them. That alone would go a long way to alleviate the issues we're discussing here (IMO).
It does not alleviate the situation where you build them to defend your older systems (instead of just recently conquered/colonised). And that is the most comon situation. Plus there is usually little need to do the cannon fodder trick on recently conquered systems because your doomstack is already there, at least playing against AI.

User avatar
em3
Vacuum Dragon
Posts: 508
Joined: Sun Sep 25, 2011 2:51 pm

Re: Comsat/planetary defenses re-design

#44 Post by em3 » Mon Sep 17, 2018 10:29 am

How about, instead of making comsats harder to use, create several viable strategies that counter comsats?
[...] for Man has earned his right to hold this planet against all comers, by virtue of occasionally producing someone totally batshit insane. - Randall Munroe, title text to xkcd #556

Jaumito
Space Kraken
Posts: 137
Joined: Tue May 16, 2017 3:42 am
Location: Catalonia, France, Europe, Earth, Sol, Orion Arm, Milky Way, Virgo Cluster

Re: Comsat/planetary defenses re-design

#45 Post by Jaumito » Mon Sep 17, 2018 11:20 am

em3 wrote:
Mon Sep 17, 2018 10:29 am
How about, instead of making comsats harder to use, create several viable strategies that counter comsats?
Some you can only use situationally, but between flak, interceptors, lots of cheap external slots (mini-asteroid swarm hull is perfect for this midgame, but there are others), minefields and stealth bombers, we're not lacking counters already.

Post Reply