Further Refinement of Stealth & Detection

This is for directed discussions on immediate questions of game design. Only moderators can create new threads.
Message
Author
User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Further Refinement of Stealth & Detection

#1 Post by eleazar »

There have been significant issues raised about stealth and detection, but they tend to be scattered through a lot of threads, and get forgotten. This topic is designed to consolidate that discussion.


Here's one of the simpler issues to start. The detection range circles usually turn into a un-informative mess with more than a few empires once long-range techs are researched. For this reason, presumably, they are now turned off by default. I don't think there is any great solution to displaying overlapping detection ranges.
The situation could be improved by only displaying the player's detection range circles. The player isn't given as much information, but all of what he's given will be always easily comprehensible, and this useful feature could be turned on by default.

At a later time we could consider showing some of the other empire's circles based on espionage or alliances, and since that would generally be a smaller subset of the empires overlapping would be less problematic.

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

Re: Further Refinement of Stealth & Detection

#2 Post by Dilvish »

eleazar wrote:The detection range circles usually turn into a un-informative mess with more than a few empires once long-range techs are researched. I don't think there is any great solution to displaying overlapping detection ranges.
I think the basic blended color approach is just fine. The darker artifacts that wind up appearing with longer range detection are ugly, but happen even with just a single empire. At one point I thought a key aspect of the circle shading was handled by GiGi & there was nothing to do about it, but now I'm pretty sure it is fixable. I'll poke around at it a bit more.
The situation could be improved by only displaying the player's detection range circles. The player isn't given as much information, but all of what he's given will be always easily comprehensible, and this useful feature could be turned on by default. At a later time we could consider showing some of the other empire's circles based on espionage or alliances, and since that would generally be a smaller subset of the empires overlapping would be less problematic.
I hope you leave in an option for seeing other empire's detection circles even before such time as you have that information gated by espionage. I would sorely miss it if I didn't even have a tech to research to get access to it again.
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
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Further Refinement of Stealth & Detection

#3 Post by Dilvish »

Ok, I fixed the detection circle overlap artifacts that appeared with high detection ranges. I couldn't attach any screenshot here so I made a new thread in the Programming forum.
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: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: More feedback thoughts (8050)

#4 Post by Vezzra »

EDIT: Moved this and follow-up posts from here. /EDIT
Magnate wrote:I still haven't been able to use stealth properly, for two reasons. The main one is that it's all or nothing - either I have better stealth tech than my opponent has detection, in which case my ships can evade his all the time, or I don't in which case I have wasted the research for nothing. I don't really see how this model can be a satisfactory part of the game - surely stealth needs to be fallible even at higher tech.
This is exactly my gripe with the current stealth mechanics too. Mat intends to do something about this problem, by means of scripting - but I don't think that will work. IMO the mechanic is fundamentally flawed.

We previously had a different mechanic, where we had only one detection meter - detection range, no separate detection strength. The stealth stat of an object was simply deducted from the detection range stat of the detector, meaning that the higher the object's stealth, the closer you had to be to be able to detect it. Example:

Ship A has detection range 100, ship B has stealth 40. A can detect B at a range of 100 - 40 = 60. If A had detection range 200, it coud see B at a range of 160. If B had stealth 100, A would only see it if B was in the same system.

IMO that system has been simple, elegant, and far superior to the revised mechanic. The reason for the revision was a micromanagement problem: If you get better scanners, you need to send scouts with the new scanners to all regions you want to monitor, because only with these new scouts you're able to detect things you haven't been able to see before. With the new mechanic, where detection strength is a global meter independent of the scanner parts (only detection range depends on the parts), all scanners get instantly upgraded to the new detection strength once you research the next detection tech, so you don't need to send new scouts.

IMO that approach falls short, because you still need to send new scouts if you want to take advantage of the enhanced detection range also (as this stat is still dependent on the ship parts), so you only save yourself some of the perceived micromanagement. And that's bought at the cost of having a more or less crippled mechanic. Avoiding micromanagement is a very important design goal for us, but I think in this particular case we went a bit overboard with the micromanagement avoiding policy.

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

Re: More feedback thoughts (8050)

#5 Post by Dilvish »

Vezzra wrote:IMO that approach falls short, because you still need to send new scouts if you want to take advantage of the enhanced detection range also (as this stat is still dependent on the ship parts), so you only save yourself some of the perceived micromanagement. And that's bought at the cost of having a more or less crippled mechanic. Avoiding micromanagement is a very important design goal for us, but I think in this particular case we went a bit overboard with the micromanagement avoiding policy.
I find your argument persuasive.
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
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: More feedback thoughts (8050)

#6 Post by Geoff the Medio »

The extra micromanagement issue was that you'd have to send your ships back to every system, every time you got a new ship part with better detection range-strength. After researching / unlocking the part, you'd have to design some new ships, then produce them, then send them to *every system*. They'd have to go right into every system, not just close by / within some (ever increasing) distance, since being at distance 0 was the only way to get your full potential detection strength in that system.

With decoupled range and strength, increasing range to ship detection doesn't require any additional micromanagement, as any system will have been already scanned by your max potential detection strength, as long as it's been in range of any detector. Getting updating ship parts doesn't require re-scanning anything, and actually makes it easier to do any re-scanning in future if/when the empire detection strength increases.
...you still need to send new scouts if you want to take advantage of the enhanced detection range also (as this stat is still dependent on the ship parts)...
This argument doesn't make much sense to me. Do you mean you still need to send new scouts if you want to take advantage of the new higher detection strength? Why would you need to send scouts after just a range increase? It's not going to reveal anything new in a system you could already reach / have scanned recently. Maybe you'd be able to see into a remote system you can't get a scout into, but that seems like a rarer concern...

That all said, I'm not opposed to switching back to something like "if (range - stealth > 0)" if that's the general preference.

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

Re: More feedback thoughts (8050)

#7 Post by Dilvish »

Geoff the Medio wrote:The extra micromanagement issue was that you'd have to send your ships back to every system, every time you got a new ship part with better detection range-strength.
I think we may not have had the hash-lined display of 'basic visibility' objects back then, which essentially makes this issue about the same (at least fairly similar) under both approaches, it seems to me. Particularly if we add some little UI marker like Mat and I have discussed for indicating whether your empire has ever 'explored' (i.e. had a ship present in) a given system, then this is a fairly modest hassle in both approaches.
That all said, I'm not opposed to switching back to something like "if (range - stealth > 0)" if that's the general preference.
I think it would be good to test this out again. It strikes me that it might even be worthwhile to (at least temporarily) make stealth mechanic be a galaxy setup option, so that people could go back and forth a bit to get a better feel for the similarities and differences.
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: More feedback thoughts (8050)

#8 Post by MatGB »

I've been working through the stealth specials and bonuses to try to standardise them but it is, frankly, a mindnumbing job and the more I do it the more I realise how much I dislike the overall setup.

It is, at a basic level, simple, but it doesn't really allow for much flexibility and the only way I can see to drastically improve the paradigm is to split range/detection strength techs and introduce a lot more detection strength settings, plus a lot of parts/buildings that affect stealth overall.

It should be possible to play a stealthy/sneaky race like Laenfa or Trith and have a radically different strategy available to you, but at the moment it's basically not there at all, and even with a lot of work I'm not sure that a) it'll be worth the effort or b) we'll ever be able to get an AI to take it into account either to use or defend against.

And frankly the few times I have gone down the stealth route it made defeating the AI so ridiculously easy that I gave up—I can imagine it would be fun to materialise a previously completely hidden fleet around a human players homeworld, but there's no fun in simply doing it to an AI that had no way of seeing it coming.

One thing, although we've discussed the idea of marking systems on the map that either have or haven't been visited, I intensely dislike the idea, at all, that an object the size of a planet can be hidden from a telescope without incredibly powerful tech. So while I'm happy the Experimentors hide their planet, I'd much rather that, barring incredible late game transcendence level tech, planets are always visible.

Something hidden in/on the planet is a different thing entirely (the Great Machine in Babylon 5 springs to mind).

What I'm thinking of doing, once I've figured out how, is to create a branch for where-I've-got on stealth and leaving it until such time that either I'm in the mood or someone else wants to pick it up and run with the idea. But if we do want to rethink the entire paradigm, I'm definitely up for that.

But can we leave it until after 0.4.5, I suspect it'll take some time if it's a complete rewrite.
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: More feedback thoughts (8050)

#9 Post by Geoff the Medio »

Geoff the Medio wrote:That all said, I'm not opposed to switching back to something like "if (range - stealth > 0)" if that's the general preference.
Something else to consider is range-dependent stealth modifiers for mobile things (ships). The idea would be to have buildings or ship parts that modify all, an empire's own, allied, or just enemy-owned ships' stealth by an amount dependent on distance. This would not cause one to need to move ships around to re-scan planets, but would make distance from the "active radar broadcast" building or the "stealth vortex" field a factor in how much stealth is needed to actually be stealthy, or how close ships can get and remain stealthy.

Having stealth-boosting or stealth-countering stuff could also require some resource to activate, such as influence (for some varieties of this content).

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: More feedback thoughts (8050)

#10 Post by Vezzra »

Dilvish wrote:I think we may not have had the hash-lined display of 'basic visibility' objects back then, which essentially makes this issue about the same (at least fairly similar) under both approaches, it seems to me. Particularly if we add some little UI marker like Mat and I have discussed for indicating whether your empire has ever 'explored' (i.e. had a ship present in) a given system, then this is a fairly modest hassle in both approaches.
If stealthy planets were the only concern here, you might be right, but, as much as I like the old mechanic, I have to admit that Geoff really has a point here:
Geoff the Medio wrote:The extra micromanagement issue was that you'd have to send your ships back to every system, every time you got a new ship part with better detection range-strength. After researching / unlocking the part, you'd have to design some new ships, then produce them, then send them to *every system*. They'd have to go right into every system, not just close by / within some (ever increasing) distance, since being at distance 0 was the only way to get your full potential detection strength in that system.
That applies to everything stealthy that could lurk in your systems, and because of that it's even worse than what Geoff describes here: You not only have to go to and rescan every single system each time you upgrade your scanner tech, if you want to be really thorough, you'll have to constantly patrol all of your systems, because there could be moving stealthy stuff (lets call it MSS) that might have just evaded your scouts when they were at system A, because it went to system B. When your scout finally got to B, MSS has left B too and returned to A. The only way to counter something like that would be to a) have a bunch of scouts constantly patrolling through all your systems (ugh), or b) station a scout in all your systems. In which case you of course need to replace all these scouts each time you research a new detection tech - ugh, ugh.

I didn't realize the full extend of the micromanagement created by that approach. As simple and elegant as it is, and as much as I like it, it very unfortunately creates exactly the thing we are determined to avoid: A mechanic that gives you a clear advantage if you're willing to micromanage it. It's even kind of proportional: the more micromanagement you're willing to invest, the more advantage you get.

I really appreciate that both of you, Geoff and Dilvish, would be willing to give the former approach another try:
Dilvish wrote:
Geoff the Medio wrote:That all said, I'm not opposed to switching back to something like "if (range - stealth > 0)" if that's the general preference.
I think it would be good to test this out again. It strikes me that it might even be worthwhile to (at least temporarily) make stealth mechanic be a galaxy setup option, so that people could go back and forth a bit to get a better feel for the similarities and differences.
But now I'm the one who would be very reluctant to do that (I know, kind of ironic considering how strongly I've been advocating the old mechanic). Geoffs point is a strong one, and the micromanagement involved will be considerable I think.

That all said, this still doesn't make to current mechanic a viable alternative. The stealth stuff is even more tricky than I thought, and I already thought it to be a tricky beast to tackle to begin with. I've spent several hours mulling over ideas, only to dismiss them as unusable as what we've already tried. However, finally I came up with something I want to propose, I'll put that in a separate post. EDIT: Just discovered a major flaw with my most recent idea. Need to think over things again. :( /EDIT
Geoff the Medio wrote:
Vezzra wrote:...you still need to send new scouts if you want to take advantage of the enhanced detection range also (as this stat is still dependent on the ship parts)...
This argument doesn't make much sense to me. Do you mean you still need to send new scouts if you want to take advantage of the new higher detection strength? Why would you need to send scouts after just a range increase? It's not going to reveal anything new in a system you could already reach / have scanned recently. Maybe you'd be able to see into a remote system you can't get a scout into, but that seems like a rarer concern...
Well, judging by my own test game experiences, gaining additional scanner range can be crucial, especially if you have longer/multiple front lines (depending of course on your situation). Being able to see farther into enemy territory, or being able to spot incoming enemy forces at longer range is of utmost importance to be able to deploy defense forces in time. So I frequently found myself in the situation where it would have been the clearly advantageous thing to replace scouts with outdated scanner parts ASAP. Which often enough (when you have several of these scouts scattered around your front lines) turned out to be a bit of a chore.

But of course that's far from having to re-visit all systems with new scouts.

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

Re: More feedback thoughts (8050)

#11 Post by MatGB »

Vezzra wrote:That applies to everything stealthy that could lurk in your systems, and because of that it's even worse than what Geoff describes here: You not only have to go to and rescan every single system each time you upgrade your scanner tech, if you want to be really thorough, you'll have to constantly patrol all of your systems, because there could be moving stealthy stuff (lets call it MSS) that might have just evaded your scouts when they were at system A, because it went to system B. When your scout finally got to B, MSS has left B too and returned to A. The only way to counter something like that would be to a) have a bunch of scouts constantly patrolling through all your systems (ugh), or b) station a scout in all your systems. In which case you of course need to replace all these scouts each time you research a new detection tech - ugh, ugh.
Except...

It wouldn't matter for your systems, each of your planets will have their own detection abilities (and I'm thinking of resurrecting the scanning ministry building I scripted up as one of my very first attempts at contributing anyway), so if something stealthy arrives in or close to one of your systems you'll see it, and will know that you need to send a scout out to track it down. Part of me says having patrol ships within "safe" territory is something that ought to be semi needed anyway, but nowhere near at the level you suggest might be needed.

Are we coming to the conclusion that none of us like any of the potential ways of dealing with stealth?
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
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: More feedback thoughts (8050)

#12 Post by Vezzra »

MatGB wrote:It wouldn't matter for your systems, each of your planets will have their own detection abilities (and I'm thinking of resurrecting the scanning ministry building I scripted up as one of my very first attempts at contributing anyway), so if something stealthy arrives in or close to one of your systems you'll see it, and will know that you need to send a scout out to track it down. Part of me says having patrol ships within "safe" territory is something that ought to be semi needed anyway, but nowhere near at the level you suggest might be needed.
Yes and no. Of course all the systems where you have colonies might be exempt from the need to patrol them frequently, but even that only works provided the colonies detection range is at least the same as the detection range of your scouts. Otherwise a scout ship might be able to detect something in a system you've a colony in the colony is not able to see. And the systems you've no colony in (and even within your territory that's usually the majority of the systems), you'll need to patrol with scouts, as they will have better detection at distance 0 that even the nearest colony.

The scenario where something is hiding within what you deem "safe" territory might not be an particularly common cause, but definitely a possible one. Your enemy could have sneaked stealthy scouts past your defence lines and now they camp somewhere in your territory, providing your enemy with very precious intelligence, without you knowing it. Now you research a new detection tech, that might be able to reveal these suckers. A player willing to build a sufficient number of scouts with the new scanner parts and thoroughly combing through his territory will have a far better chance at discovering and neutralizing the enemy scouts than a player who doesn't want to put up with the micromanagement involved.

I think that's a realistic scenario where the willingness or unwillingness to micromanage things can make a considerable difference. Too much for my taste TBH.
Are we coming to the conclusion that none of us like any of the potential ways of dealing with stealth?
Not yet. I'm in the process of writing a proposal on a revision of the detection/stealth mechanic that I could live with. Stay tuned ;) EDIT: Correction: I'm in the process of trying to come up with something workable. No success yet. :( /EDIT

That all said, if someone has an idea how to make the old mechanic work while avoiding the micromanagement issue at least to a tolerable degree, I'm all ears. I like the old mechanic and would love to be able to reintroduce it.

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

Re: More feedback thoughts (8050)

#13 Post by MatGB »

Vezzra wrote: The scenario where something is hiding within what you deem "safe" territory might not be an particularly common cause, but definitely a possible one. Your enemy could have sneaked stealthy scouts past your defence lines and now they camp somewhere in your territory, providing your enemy with very precious intelligence, without you knowing it. Now you research a new detection tech, that might be able to reveal these suckers. A player willing to build a sufficient number of scouts with the new scanner parts and thoroughly combing through his territory will have a far better chance at discovering and neutralizing the enemy scouts than a player who doesn't want to put up with the micromanagement involved.
But there is, and should be, a balance to be found. I'd love to introduce, say, a "pirate base" special that hides or gets created in asteroid fields or nebulae that you'd need to actively seek out, and stealth-as-a-strategy would basically need some way of being able to sneak a fleet past unpatrolled areas.

Which is the problem, some degree of strategic empire management (which should involve anti-piracy patrols in my little ideal game ;-) ) is needed, too much is micromanagement.

Regarding planetary detection, currently planets get all the bonus of a detection tech immediately, scouts with that part need to be built, if I reintroduce the scanning ministry, what I'm thinking is planets get a bonus to detection equal to their infrastructure as long as they don't have a dedicated scanning facility, which would bring it up even more, but still not as good as the best scout hulls (which currently equal scanning facilities).

Something worth considering in all this is to make planetary detection a meter, I like the idea of having it slowly expand.
Not yet. I'm in the process of writing a proposal on a revision of the detection/stealth mechanic that I could live with. Stay tuned ;)
Look forward to 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
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: More feedback thoughts (8050)

#14 Post by Geoff the Medio »

MatGB wrote:Are we coming to the conclusion that none of us like any of the potential ways of dealing with stealth?
Nobody ever seems to respond to or consider my suggestions along these lines:
Geoff the Medio wrote:Something else to consider is range-dependent stealth modifiers for mobile things (ships). The idea would be to have buildings or ship parts that modify all, an empire's own, allied, or just enemy-owned ships' stealth by an amount dependent on distance. This would not cause one to need to move ships around to re-scan planets, but would make distance from the "active radar broadcast" building or the "stealth vortex" field a factor in how much stealth is needed to actually be stealthy, or how close ships can get and remain stealthy.

Having stealth-boosting or stealth-countering stuff could also require some resource to activate, such as influence (for some varieties of this content).
Notably, I think, having range-dependent stealth-reducing buildings would make it necessary for the player moving stealthy ships to be careful where they're positioned, but wouldn't give the player trying to detect the ships any extra micromanagement to optimize their detection network. Particularly if the buildings are limited in number and perhaps require a special resource or other limiting factor for their production, there would be no way to micromanage them, but there would be need to carefully select where to put the few a player has available.

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

Re: More feedback thoughts (8050)

#15 Post by Dilvish »

Geoff the Medio wrote:Nobody ever seems to respond to or consider my suggestions along these lines:
Sorry to have left you feeling the idea was ignored. As for me personally, I found it hard to quickly form an opinion about, but I've thought about it more now. It does sound interesting, but I'm concerned that perhaps the considerations you mention for avoiding micromanagement might not really be enough-- and if the whole idea only really matters with respect to a very small handful of buildings at rare resource locations, then it seems like it will mostly be leaving the stealth situation unchanged.

Here's another idea-- make stealthy ships be more akin to a stealthy planet, in that if you have any physical presence in the system then you get basic visibility of the stealthed object. If we also made a rule that a stealthed ship had to either 'slow down' and pause for a turn upon entering a system, or else entirely lose its stealth for a turn, then I think the micromanagement would largely disappear. You wouldn't have to maintain physical presence in each and every system; it would generally be enough to just have a buffer zone of presence that an enemy would have to travel through. If their stealth is good enough they could be confident of getting through without combat, but they would still have some kind of faint warp signature or whatever that was detected (basic vis). If the enemy did pass through the buffer in either fashion, then a cat-and-mouse game might start up, but that is hardly micromanagement, that sounds like a fun chase. Upgrading detection tech would not trigger micromanagement, because your old scouts are still good enough to get basic visibility in their same system, and then you would know to send in the good detectors-- again a chase, not micromanagement. I like the idea pretty well, if I do say so myself :D

Also, regarding Mat's objection to stealthy planets being undetected from outside the system-- here is a little brainstorming. What if from outside the system a stealthed planet appeared as an asteroid belt-- If your empire had never explored the system (and so you have never had even basic visibility of the actual planet), then the server tells your client that the planet is an empty asteroid belt. Until you have actually physically explored the system, or have ever gotten actual partial visibility of the real planet, then the server continues to feed info for the planet as an asteroid belt, and if you have partial visibility of the system if feeds your client visibility info saying you have partial visibility of the 'asteroid belt' (although the server is actually tracking that you do not have any visibility of the planet). The more I think about it, the more I like it :D
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