Fractal Softworks Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Starsector 0.97a is out! (02/02/24); New blog post: Simulator Enhancements (03/13/24)

Pages: [1] 2

Author Topic: sensor diversification  (Read 7213 times)

Cik

  • Admiral
  • *****
  • Posts: 607
    • View Profile
sensor diversification
« on: December 20, 2015, 09:38:17 AM »

FIRST OF ALL I THINK IT SHOULD BE REMOVED BECAUSE IT RUINED THE GAME RABBLE RABBLE ::)

Spoiler
he he
[close]


this'll probably be a megathread but it needs to be to cover the complexities.

i probably don't need to outline how the current sensor system works. i think it's a good baseline.

first a note on philosophy: this game isn't a simulation. what does that mean? it means that you shouldn't add extra depth and confusion to it for realism's sake. that does not mean that you cannot add realism, just that it must serve a gameplay purpose. my suggestions are taken from how some sensor systems behave in real life as a baseline, however in the end they don't really need to compare that closely; as long as they add depth and interesting gameplay options, they can behave however.

now: what are the problems with the current sensor system?

1. in the end-game it pretty much stops mattering. once fleets blow up to a reasonable size they can see each other from halfway across the sector even if they are perfectly stationary.

2. sensors are not very granular. there's a huge mishmash of sensor types that are all rolled into one "sensor strength" which allows you to see, and a "sensor signature" which is really simple and allows you to be seen.

3. the omnidirectional and "invincible" nature of the sensors of all fleets mean that there are very few strategies that can be employed in trying to lose or gain a sensor advantage over an enemy fleet. this is mostly a result of problem #2

4. sensors are too binary. it's almost impossible to gain a sensor advantage with the current mechanics. you are nearly always detected at almost exactly the same range they detect you. this means that it's almost impossible to sneak up on fleets, or to hide effectively while surveying an area.

so, my suggestion is diversify sensors a little bit into distinct types. there are a few types of sensors we can reasonably extrapolate will be used in pretty much any space setting, as they are relatively simple, robust, and can be produced by pretty much anyone. all sensors should have tradeoffs, strengths and weaknesses and i'll do my best to point out what these are and why.

first, sensor types: for our purposes we will divide sensors into active and passive.

active sensors emit detectable radiation or 'noise' that means that anyone who is listening for an active system and is in it's area of effect will 'hear' it with whatever system they are using. an example of this is active sonar. you emit a pulse. the pulse hits the enemy. the enemy hears the pulse hit them. the pulse comes back. you can detect the enemy. active systems necessarily reveal at least your rough position. that means they should allow you to be detected better by the enemy (if they have some means of detecting you)

passive sensors don't emit anything. the eyeball is a passive sensor; the light hits the object, the light enters your eye, you can see it. you can of course look at someone who isn't looking in your direction and they won't notice. infrared systems are also passive, you can see the heat coming off something without emitting anything. passive systems are necessarily more stealthy because there isn't a way to detect that something is looking at you, however they usually have other downsides such as lower detection range, uncertainty about detected object's bearing, speed, range etc. or possibility of false-positives in detection.

now, onto specific sensors.

infrared: simple, lowtech and in use in pretty much every combat vehicle since the late 40s, infrared is a sensor that sees heat, or more accurately heat radiated off of surfaces. infrared leakage is pretty much impossible to stop (in fact, totally impossible) and because of the extremely low temperature in space it is easy to detect over vast distances. infrared sensors should detect larger heat sources better, and anything that has it's engines engaged much better. e-burn should light up like a christmas tree in infrared. infrared should probably have a 360 degree detection ability, however it would not detect as well on your tail because of your own drive plume blinding the sensors. crazy ivans may be required. infrared is a passive sensor, meaning you do not have to emit anything to see hot objects. because of this, it is far stealthier than an active system and thus should not increase your sensor signature. since infrared is technologically fairly simple, nearly every ship should have an infrared detection system.

RADAR: basically a radio echolocation system. radar is an active system (should increase your signature if you turn it on) but provides a superior detection against targets at long range, especially if they are large. detection arcs can be anything from a forward fixed beam to a 360 degree system, but a smaller frontal arc might be better for gameplay and simplicity's sake. note that unlike infrared, RADAR only cares about how big and how far away you are. it doesn't matter if you are standing still or what your engines are doing, it can detect you anyway. the downside is that anyone caught in the radar beam knows they are being looked at. radar systems should be standard on most ships, perhaps barring civilian vessels. I'd expect pretty much any warship to have one.

RW (radar warning) not so much a detection system as a defensive measure, the RW system tells you if a radar is looking at you. useful to know if someone is trying to find you, it's range should be almost unlimited. it is a passive sensor as it does not emit anything. most military-grade ships should have a comparable system.

exotic sensors: these should be rarer, and the amount of them is almost unlimited. but for starsector i'd suggest two.

gravimetric: gravimetric sensors detect massive objects, and the more massive the easier and the further away it can be detected. for our purposes we can assume these are relatively coarse, however they should be very effective at detecting large concentrations of cruiser+ ships as the entire point should be hunter-killer type actions. passive, as you are not really emitting anything, only reading the gravity of an object from range. should be standard on most phase-ships, as they are the u-boats of space. dedicated scout or surveying ships may also have them.

phase detection (cool name needed): detects phase fields at long range. effectively cuts through phase ships' sensor signature reduction. active(?) should be present on most expansion epoch+ or tri-tachyon ships. domain prototypes or very expensive scout ships may also have them.

other things:

sensor fuzz

Q: what's the difference between a hound and a cerberus at millions of kilometers?

A: uh...

some 'fuzziness' to sensors at long ranges might not be amiss. you should nearly always be able to tell with any sensor what class the ship is, but the difference between the aforementioned hound and cerberus might be impossible to tell. some sensors should be better at determining the exact ship, while others (infrared, probably) just tells you the rough direction it's moving and it's approximate size.

likewise, not being able to detect the whole fleet, but only the less stealthy parts of it would be pretty cool. for example, you turn on your IR system and hide in a belt. you're watching the jump point and you see an onslaught. it gets closer. then you see it's medium escort of destroyers, then it gets closer, you start to see it's less stealthy frigates. it gets closer. then you see it's phase frigates. the sky's the limit with this sort of thing.

more granularity to move speed: would be nice to have a few more buttons that limit your fleet's speed but at which you gain bonuses to your signature, but which aren't as crazy as go dark is.

deployables, more specifically sensor buoys: these are of course attendant on a system that isn't there (yet?) but i mean come on that's the rest of the post too so whatever.

deployable sensor buoys would make a ton of sense in-setting (in the tactical map already!) and add a far larger non-personal aspect to sensors. these should be anchorable to objects or able to be put into orbit around planets, asteroids, or the sun; depending on the sensor type they are using they should have their own signature. they should of course be able to be destroyed, perhaps more interesting though is the opportunities to tie these to factions and allow the AI to use them as more limited objectives in their constant wars. not only that, but they would be of great help to the player in allowing them to scope out jump points or ring systems or what-have-you in order to get a much better picture of local areas without directly approaching and risk being blown up.

the uses for these are of course endless, but they have downsides; they would be vulnerable to destruction by your's or your factions enemies, perhaps only have a limited duration, etc. if or when these are added the sector factions should likely have them deployed around their space already, as that would make sense. a deliberate campaign to blind a specific faction seems like it would be great fun.

so, if we implement any of this what will we have to modify about the sensor system as it is right now?

well, a bunch of new hotkeys for the bar will need to be added to turn off and on various systems. several more 'layers' will have to be added to the detection mechanic; each fleet will need a signature for each type of sensor in the game, and each fleet will need as many sensors as that fleet will have. how difficult is that? i'm honestly not sure. it could be anything from easy to impossible. the hardest part seems to me like it would be teaching the AI to use new systems, though maybe the backbone of that is already there. it seems pretty smart about the current system, for one.

so what's the payoff?

the payoff i believe will be relatively awesome. the sensor mechanics of cat & mouse will get more interesting. in particular, ambushes should be easier to pull off as a stationary target will be comparably much harder to see (due to infrared having trouble seeing stationary ships) breaking up the omnidirectional nature of current sensors means that how you approach the enemy will matter. ships can be easily diversified within the system via sensor types and strengths. factions, too.



this of course skips the effects entirely of electronic warfare, jamming etc. perhaps i will go on about that later.

Spoiler
nobody's going to read this are they?
[close]


anyway those are my thoughts. thoughts?
Logged

Megas

  • Admiral
  • *****
  • Posts: 12118
    • View Profile
Re: sensor diversification
« Reply #1 on: December 20, 2015, 09:49:27 AM »

Sensors are most relevant when attempting to sneak into markets.  Otherwise, sensors cease to matter once highly visible fleets of major factions become the only remaining threat in the game.

One way to make sensors a bit more useful in the endgame is to add a boss faction (that invade through the deactivated gates or otherwise mysteriously appear) that does not use transponders.  Boss faction could be aliens, Knights Templar faction knockoff, the Domain itself, something.  In other words, an endgame threat that can sneak up on players like pirates do in the early game.
Logged

Cik

  • Admiral
  • *****
  • Posts: 607
    • View Profile
Re: sensor diversification
« Reply #2 on: December 20, 2015, 10:10:41 AM »

Sensors are most relevant when attempting to sneak into markets.  Otherwise, sensors cease to matter once highly visible fleets of major factions become the only remaining threat in the game.

sure but should that be the case? seeing the enemy is the most important step of warfare pretty consistently through every era.
Logged

Dabor

  • Lieutenant
  • **
  • Posts: 92
    • View Profile
Re: sensor diversification
« Reply #3 on: December 20, 2015, 10:27:56 AM »

While I think sensors are mostly fine as they are, the passive/active distinction is an idea I'm all for trying.

Having a choice between high quality and long range sensors is also a simple but nice distinction. So a civilian fleet would have great detection range, but only be able to figure out the rough number and size of ships. A military fleet of the same class could do both, whereas something like pirates or a patrol might not have the greatest range but would obviously want to get details about what it's about to attack.
Logged

Gothars

  • Global Moderator
  • Admiral
  • *****
  • Posts: 4403
  • Eschewing obfuscatory verbosity.
    • View Profile
Re: sensor diversification
« Reply #4 on: December 20, 2015, 11:23:17 AM »

in the end-game it pretty much stops mattering. once fleets blow up to a reasonable size they can see each other from halfway across the sector system even if they are perfectly stationary.

I agree with this. To me it seems more like an endgame and AI problem, though. E.g. if bounty fleets would use the terrain to hide that would give the mechanic more late game relevancy. Or if enemy patrols would coordinate to stay hidden and then strike at you together. Specfic skills and abilities would also be nice.

The proposed changes seem far to complex to me. Complexity should arise from interesting terrain and ability interactions, and not from the base mechanic.
Logged
The game was completed 8 years ago and we get a free expansion every year.

Arranging holidays in an embrace with the Starsector is priceless.

Pushover

  • Captain
  • ****
  • Posts: 292
    • View Profile
Re: sensor diversification
« Reply #5 on: December 20, 2015, 11:52:19 AM »

in the end-game it pretty much stops mattering. once fleets blow up to a reasonable size they can see each other from halfway across the sector system even if they are perfectly stationary.

I agree with this. To me it seems more like an endgame and AI problem, though. E.g. if bounty fleets would use the terrain to hide that would give the mechanic more late game relevancy. Or if enemy patrols would coordinate to stay hidden and then strike at you together. Specfic skills and abilities would also be nice.

The proposed changes seem far to complex to me. Complexity should arise from interesting terrain and ability interactions, and not from the base mechanic.

I've still been able to hide effectively by Going Dark in a ring system, despite 2 Onslaughts, an Atlas, 2 Phaetons, 2 Dominators, and some other stuff. Took my sensor profile to under 150.

That said, I agree that the AI would be awesome if it used the terrain better, such as pirates staying stationary in a ring system, or pirates hovering at the edges of nebulae.
Logged

Cik

  • Admiral
  • *****
  • Posts: 607
    • View Profile
Re: sensor diversification
« Reply #6 on: December 20, 2015, 11:56:55 AM »

in the end-game it pretty much stops mattering. once fleets blow up to a reasonable size they can see each other from halfway across the sector system even if they are perfectly stationary.

I agree with this. To me it seems more like an endgame and AI problem, though. E.g. if bounty fleets would use the terrain to hide that would give the mechanic more late game relevancy. Or if enemy patrols would coordinate to stay hidden and then strike at you together. Specfic skills and abilities would also be nice.

The proposed changes seem far to complex to me. Complexity should arise from interesting terrain and ability interactions, and not from the base mechanic.

I've still been able to hide effectively by Going Dark in a ring system, despite 2 Onslaughts, an Atlas, 2 Phaetons, 2 Dominators, and some other stuff. Took my sensor profile to under 150.

That said, I agree that the AI would be awesome if it used the terrain better, such as pirates staying stationary in a ring system, or pirates hovering at the edges of nebulae.

going dark in a ring system is really effective, though it usually won't prevent them from finding you if your last-known was somewhere near where you're hiding. that isn't really a flaw though. the only other problem is that the ring moves relative to you but afaik that was on the to-do. hiding will get a lot easier relatively soon which is very cool.

making most jump points lagrange points instead of now where they are sort of free-floating would be very nice to have. as it is you can't really hide in open space with anything bigger than a frigate. i say lagrange because it would probably result in a cluster of nearby objects you could hide in, which would be a boon for the player and the pirates (who really need the help because it feels like most of the time i just see them getting ruthlessly slaughtered even next to their own bases)

totally offtopic but whatevs
Logged

Cik

  • Admiral
  • *****
  • Posts: 607
    • View Profile
Re: sensor diversification
« Reply #7 on: December 20, 2015, 12:06:11 PM »


I agree with this. To me it seems more like an endgame and AI problem, though. E.g. if bounty fleets would use the terrain to hide that would give the mechanic more late game relevancy. Or if enemy patrols would coordinate to stay hidden and then strike at you together. Specfic skills and abilities would also be nice.

The proposed changes seem far to complex to me. Complexity should arise from interesting terrain and ability interactions, and not from the base mechanic.

no argument with your first point

obviously i disagree with the second one considering i proposed the idea in the first place  :P

of course it can scale anywhere from the full depth of individual sensor modelling to just a very basic passive/active, depending on how deep you want to go. the real test is if it adds complexity and depth or just complexity. my guess is that it would be both, but opinions may differ. IMO the sensor system is a pillar of this type of game, just as important as the shooting part. being able to design a sensor system deep enough that it can support ambushes, surprise attacks, cat & mouse and electronic warfare is very important to me. the sensor system provides not only a fun game to play on the strategic layer but provides all sorts of ways to set up interesting tactical battles, so making it very fleshed out can only be a good thing.
« Last Edit: December 20, 2015, 02:06:20 PM by Cik »
Logged

Weltall

  • Admiral
  • *****
  • Posts: 774
    • View Profile
Re: sensor diversification
« Reply #8 on: December 20, 2015, 12:10:28 PM »

I really had to read it twice to understand some things. Sadly the part of the different types of sensors did not register in my little brain XD

What I wanted  to say though is that if ALL that will be added, one of the main and I meant MAIN features of the games, will be stealthiness. Especially if like you said a handful of shortcut buttons for stealthiness will appear, players will be able to use stealthiness as means to "win". Well when I say win, I say it will be a really powerful tool that will help to boost people faster. I do not think it would be bad to have some stealth options, but wouldn't it be better if for example there were stealth points, that would be used for 1 skill to lower it's detectability very much, without losing speed, but after the points end the system will not be able to cloak the fleet anymore.

IF, I said IF, I read this right you suggest many different systems and quite some detail. Sure I would like to be able to sneak up on fleets, so they will not run away or not chase me, but definitely I would prefer something simple rather than complex.
Logged
Ignorance is bliss..

nomadic_leader

  • Admiral
  • *****
  • Posts: 725
    • View Profile
Re: sensor diversification
« Reply #9 on: December 20, 2015, 07:28:12 PM »

Summary of OP for confused people:

Basically he's saying he wants active/passive sensors like in naval warfare. Play the naval wargame simulation Harpoon to see what he means. Passive sensors that let you hear a little and not be seen much, and active sensors that let you hear a lot but be seen a lot. Right now the going dark is a little too simple for his taste, basically.

Science:
All radio, radar, microwave, cel phones, wifi, infared, etc and even visible light  uses electromagnetic waves, the only difference is in the frequency. https://en.wikipedia.org/wiki/Electromagnetic_spectrum

Passive sensor example: a home radio. It doesn't transmit* any electromagnetic waves, so it is passive. it hears other radio transmissions, but it doesn't transmit any of its own. As a result, it can only tell you the direction the signal is coming from. But in larger fleets with more radio receivers, you could triangulate and get distance as well.


Active sensor: Example: RADAR in planes today. RADAR just a radio transmitter- it spits out signals, then waits to see how long it takes the signal to bounce off something and come back to the airplane. So you get more details like composition, size, speed, etc. But because it is spitting our radio, anyone with a receiver can hear it.

My opinion
The sensors are so much fun in this version, we should add more. The seperation of the campaign and the combat is the clubfoot for starsector, so you must keep making campaign richer and offering fun non-combat gameplay.

So sensors should get more nuanced. There should be special hullmods or cargos or skills (though I hate skills) that grant you this

Right now whatever sensor mode you have, going dark, normal, or sensor burst, your sensor range and your detection range are always about equal. That's boring.

A quick proposal based on OP:

Passive: you have low visibility, but you can see 3 times as far as you can be seen. However you ONLY see those little blue triangle "unidentified contact" symbols, and if you have a small fleet the distances aren't so reliable since you can't triangulate. See mockup below:


Active:
More or less like we have now, where you detection and your own sensor ranges are about the same. But with a few different ranges created by various different hullmods, skills etc.

Stealth should also be something you can do a little more.

Plus a few other special types or modes for abilities, like a low visibility, but high range sensor that drained a lot of CR or something like this, or some kind of X-ray that would actually fry you hulls a little bit. Directional stuff might be interesting too. who knows.


THe sensors are so much fun! Thank you for adding them to the game! Make them even more interesting and fun and with more options!



*lets forget local oscillators etc for the purposes of the game.
Logged

TrashMan

  • Admiral
  • *****
  • Posts: 1325
    • View Profile
Re: sensor diversification
« Reply #10 on: December 21, 2015, 02:47:04 AM »

Speaking realistically, stealth is for smaller fleets and/or smugglers.

When you have fleet of giant capital ships, stealth is pretty much impossible - AND IT SHOULD BE.
Logged

Cik

  • Admiral
  • *****
  • Posts: 607
    • View Profile
Re: sensor diversification
« Reply #11 on: December 21, 2015, 04:28:25 AM »

Speaking realistically, stealth is for smaller fleets and/or smugglers.

When you have fleet of giant capital ships, stealth is pretty much impossible - AND IT SHOULD BE.

not even the case. go dark in a ring system and sit still.

what you have to realize is that even a large fleet is pathetically tiny compared to even the smallest of stellar objects. hiding an eagle in plain sight should be hard, but powering down to basic essentials, turning the engines off and then floating in a sea of mass that encompasses a planetary system and weights in at hundreds(?) thousands, millions of magnitudes heavier than your spaceship is not something that should be impossible.

similar size fleets should not always detect each other, that's really silly. at that point you're just removing the whole point of the sensor system. you might as well just remove terrain entirely if that's what you think and go back to .65xs total omniscience.
Logged

TrashMan

  • Admiral
  • *****
  • Posts: 1325
    • View Profile
Re: sensor diversification
« Reply #12 on: December 21, 2015, 07:58:59 AM »

Speaking realistically, stealth is for smaller fleets and/or smugglers.

When you have fleet of giant capital ships, stealth is pretty much impossible - AND IT SHOULD BE.

not even the case. go dark in a ring system and sit still.

what you have to realize is that even a large fleet is pathetically tiny compared to even the smallest of stellar objects. hiding an eagle in plain sight should be hard, but powering down to basic essentials, turning the engines off and then floating in a sea of mass that encompasses a planetary system and weights in at hundreds(?) thousands, millions of magnitudes heavier than your spaceship is not something that should be impossible.

similar size fleets should not always detect each other, that's really silly. at that point you're just removing the whole point of the sensor system. you might as well just remove terrain entirely if that's what you think and go back to .65xs total omniscience.

http://www.projectrho.com/public_html/rocket/spacewardetect.php
Logged

Weltall

  • Admiral
  • *****
  • Posts: 774
    • View Profile
Re: sensor diversification
« Reply #13 on: December 21, 2015, 08:25:47 AM »

Dat moment that logic and facts collide XD

Even though I do not care that much for a hide and seek game, although at times I would love to have the option, I want to point out this is a game. It was mentioned in the opening post that this is not a simulation game. It should not matter if ships should be or should not be detectable. Realism should not take away the fun of the game, but add to the game. Of course a lot of games that decide to do something that is not logical, gain a lot of people being like; THIS MAKES NO SENSE! IT SHOULD BE CHANGED OR REMOVED! FIX IT OR WE WILL CRUCIFY YOU!.
« Last Edit: December 21, 2015, 08:27:50 AM by weltall »
Logged
Ignorance is bliss..

Cik

  • Admiral
  • *****
  • Posts: 607
    • View Profile
Re: sensor diversification
« Reply #14 on: December 21, 2015, 08:30:58 AM »

Speaking realistically, stealth is for smaller fleets and/or smugglers.

When you have fleet of giant capital ships, stealth is pretty much impossible - AND IT SHOULD BE.

not even the case. go dark in a ring system and sit still.

what you have to realize is that even a large fleet is pathetically tiny compared to even the smallest of stellar objects. hiding an eagle in plain sight should be hard, but powering down to basic essentials, turning the engines off and then floating in a sea of mass that encompasses a planetary system and weights in at hundreds(?) thousands, millions of magnitudes heavier than your spaceship is not something that should be impossible.

similar size fleets should not always detect each other, that's really silly. at that point you're just removing the whole point of the sensor system. you might as well just remove terrain entirely if that's what you think and go back to .65xs total omniscience.

http://www.projectrho.com/public_html/rocket/spacewardetect.php

i know all about rho, and i also know that this isn't a simulator. there are gameplay reasons that trump realism.

unfortunately there really aren't any hard sci-fi space warfare games, as i would play them in a heartbeat of course.

in the game ring systems hide you. since they can hide you the amount they can hide is open to interpretation which is what i'm doing.

realism is it's own field which i am perfectly happy to discuss, but realism wise the sensor system in the game should simply be removed because "realistically" you can detect the signature of a rocket engine running from an extrasolar point of view.

Quote
Of course a lot of games that decide to do something that is not logical, gains a lot of people being like; THIS MAKES NO SENSE! IT SHOULD BE CHANGED OR REMOVED! FIX IT OR WE WILL CRUCIFY YOU!


even putting aside realism, there are compelling gameplay reasons to change up the sensor system. i only suggest what i suggest because it would modify gameplay in interesting ways.

Logged
Pages: [1] 2