Draft of an idea for a change to sensing

[nakorkren (OP)]

Currently, sensing works on individual ships/missiles. As a result, big ships get seen first, small ships get seen last, and how many ships are grouped together in one area is irrelevant.

I'd like to float the idea of changing to a system where sensors detect the COMBINED emissions of whatever is at one location in space (be it fleet or giant alpha strike composed of multiple salvos), rather than the individual emissions. Sensor strength (for active) or sensitivity (for passive) would determine how far out you can see the blob, and sensor resolution (currently exists for active and could be added for passive) would determines when you start to resolve different contacts. Make this the basis of the tracking time boost to PD targeting, which can be arbitrarily adjusted to tune the effectiveness of PD up or down as needed for balancing purposes. Massive alpha strikes get detected earlier than smaller periodic salvos, and hence get hit more easily by PD. By reducing the effectiveness of alpha strikes, you make the ability to reload and fire again more valuable. The sensing changes would also have implications for detection of fleets, encouraging less blobbing if you're trying to be sneaky, and penalizing large ships less (from a detection standpoint) vs a comparable tonnage of smaller ships. It would also make a single ship operating alone harder to spot (at least relative to a giant fleet), which could open up interesting scouting opportunities.

Also, depending on how current sensing is coded and how it was changed to implement this, this approach might reduce game lag due to calculation of sensor detections. Without seeing the code I'm only theorizing, but if you performed sensor checking on a fleet vs fleet basis first, and only started checking fleet vs individual ships once fleets can see there's a blob out there, it may reduce processing time.

Thoughts?

[nuclearslurpee]

I think the one big problem with this is that it promotes very micro-heavy gameplay to avoid detection. A big fleet which would be easily detected would rather spread out into separate fleets which are harder to detect, but this clear optimum is micro-heavy as you have to separate the entire fleet and then give the correct orders to every single ship. Further, since having a bunch of single-ship fleets piled up at the same XYZ-coordinate in space should not change detection behavior physically, which implies that ships need to be positioned some distance away from each other, which means even more fiddling with formation orders to set (and calculate, for large fleets!) the correct distances and angles. Have fun doing that every time you fly through a jump point, Lagrange point, enter and leave planetary orbit, etc. not to mention pulling your fleets together in time to defend against a missile attack. You also have the question of what minimum distance is needed to be detected separately which is another can of worms, especially since too large of a distance leaves a fleet vulnerable to missiles that can close in faster than a fleet can re-form. Of course, a player can always choose to keep their fleet in a doomstack and accept detection penalties, but Aurora is not really designed on the principle of forcing players to choose between "more optimal" and "less micro" except where the limits of the game mechanics (e.g. standing orders) unfortunately force this.

Which brings to mind the question of why can't we just "imagine" that things work this way anyways? After all, a fleet of 50 ships is not 50 ships at the exact same coordinate but rather represents a group of 50 ships flying in some formation with each other which is simply not represented in-game. It's not too hard to assume that the fleet commander ordered her ships to fly in whatever formation is optimal for avoiding detection, without having to represent this in-game and thereby add a bunch of extra complexity that doesn't lead to a good gameplay experience. Even a fleet spread out with a 1,000 km formation diameter is not much bigger than a dot in Aurora scaling (see also: radius of Earth in real life: 6,400 km; radius of Earth in Aurora's tactical map: 0 km. Somehow nobody has an issue with this...).

BTW, as far as missile balance goes I think it is best to wait for 2.2 to release and see for ourselves how things work. No point in solving a problem which may or may not exist in the next version anyways.

[Pury]

If one would like to implement it there might be ways to do so without adding a lot of micro, and even allow AI to use this mechanic effectively as well. What comes to my mind is adding 3 "states" to a formation. Grouped, loosly packed and separated. Each one affecting mainly 2 things: Effectiveness of PD fire and group detection signature mechanic proposed by you. If formation is grouped, then all Final defensive fire PD will work with same efficiency for every ship. With the downside of highest signature for enemy to spot. Loosly packed would work with lets say 20% penalty, and separated 50% penalty. Changing betwen states would not be instantaneous. Time needed would depend on number of ships and some formation commander bonus, to initiative perhaps. Ofc this would require some changes to PD system.

[nuclearslurpee]

If one would like to implement it there might be ways to do so without adding a lot of micro, and even allow AI to use this mechanic effectively as well. What comes to my mind is adding 3 "states" to a formation. Grouped, loosly packed and separated. Each one affecting mainly 2 things: Effectiveness of PD fire and group detection signature mechanic proposed by you. If formation is grouped, then all Final defensive fire PD will work with same efficiency for every ship. With the downside of highest signature for enemy to spot. Loosly packed would work with lets say 20% penalty, and separated 50% penalty. Changing betwen states would not be instantaneous. Time needed would depend on number of ships and some formation commander bonus, to initiative perhaps. Ofc this would require some changes to PD system.

This would probably work better and give the commander Reaction bonus + crew training some actual value in missile combat.

[Jorgen_CAB]

I think that detection could be less binary and more uncertain as they are in reality thus if there are more ships you are more likely to detect something and even identify some of them but you will never be able to know for sure how many there are unless you manages to follow them for a longer time. It also means that more sensors means more chances to make positive contact too.

So contact should be in three states, something detected but you are not entirely sure what it is or even exactly where it is. You have a firm sensor indicator but you don't know anything more than that something is there and lastly you have a firm lock and managed to identify the source.

Sensors should need more than just detecting something and not be binary... in my opinion it would be more fun.

I don't think that more ships should be easier to detect but it will be more likely that you detect something and further out the more ships that are there. It also would remove the need to spreading them out and micro things.

I also think the game would benefit with more sensor mechanics in terms of stealth, detection and identification. More electronic warfare and possibility to spoof sensors and trick them. Why not be able to trick them to thinking there is a ship when it is just a decoy for example, sending the enemy out on a wild ghost chase for example.

I do think that the game would be more fun if sensors were not binary and there were ways to mask ships, say around planets and asteroids for example.

Sure... it would make "combat" in general more spread out in a system perhaps, but generally I would have no big problem with that. It also would mean that a better sensor are not guaranteed to find something before a worse one, just more likely to do so.

The calculation does not have to be more complicated than now, just a few more states before something is fully identified. Even a passive sensors detection should be able to eventually determine with a very high certainty what enemy design it is, you can in real life as all emissions usually are very unique for each type of platform. I also think this would be a good way to make "Flag Bridges" more interesting... they could be made to influence the chance to fully identify sensor sources as they combine the information from en entire squadron or small task force of ships.

[Garfunkel]

Flag bridge as the hub for sensor fusion and C4I? Sounds good but complicated to pull off.

[Steve Walmsley]

I think that detection could be less binary and more uncertain as they are in reality thus if there are more ships you are more likely to detect something and even identify some of them but you will never be able to know for sure how many there are unless you manages to follow them for a longer time. It also means that more sensors means more chances to make positive contact too.

So contact should be in three states, something detected but you are not entirely sure what it is or even exactly where it is. You have a firm sensor indicator but you don't know anything more than that something is there and lastly you have a firm lock and managed to identify the source.

Sensors should need more than just detecting something and not be binary... in my opinion it would be more fun.

I don't think that more ships should be easier to detect but it will be more likely that you detect something and further out the more ships that are there. It also would remove the need to spreading them out and micro things.

I also think the game would benefit with more sensor mechanics in terms of stealth, detection and identification. More electronic warfare and possibility to spoof sensors and trick them. Why not be able to trick them to thinking there is a ship when it is just a decoy for example, sending the enemy out on a wild ghost chase for example.

I do think that the game would be more fun if sensors were not binary and there were ways to mask ships, say around planets and asteroids for example.

Sure... it would make "combat" in general more spread out in a system perhaps, but generally I would have no big problem with that. It also would mean that a better sensor are not guaranteed to find something before a worse one, just more likely to do so.

The calculation does not have to be more complicated than now, just a few more states before something is fully identified. Even a passive sensors detection should be able to eventually determine with a very high certainty what enemy design it is, you can in real life as all emissions usually are very unique for each type of platform. I also think this would be a good way to make "Flag Bridges" more interesting... they could be made to influence the chance to fully identify sensor sources as they combine the information from en entire squadron or small task force of ships.

The game used to be like that in the early years. For example, you would have a thermal contact but no ID or even IFF, so you had to get closer and try to identify ship type and race (similar to Harpoon). However, while that is interesting for the first few contacts, it became very tedious after micromanaging investigation of a hundred such contacts. I changed sensor detection to the current model, which assumes target identification, or even target motion analysis, was handled by the your staff behind the scenes so you could get on with deciding how to handle it.

If I ever get around to a more tactical game than Aurora, it would definitely include something on these lines.

[Jorgen_CAB]

I think that detection could be less binary and more uncertain as they are in reality thus if there are more ships you are more likely to detect something and even identify some of them but you will never be able to know for sure how many there are unless you manages to follow them for a longer time. It also means that more sensors means more chances to make positive contact too.

So contact should be in three states, something detected but you are not entirely sure what it is or even exactly where it is. You have a firm sensor indicator but you don't know anything more than that something is there and lastly you have a firm lock and managed to identify the source.

Sensors should need more than just detecting something and not be binary... in my opinion it would be more fun.

I don't think that more ships should be easier to detect but it will be more likely that you detect something and further out the more ships that are there. It also would remove the need to spreading them out and micro things.

I also think the game would benefit with more sensor mechanics in terms of stealth, detection and identification. More electronic warfare and possibility to spoof sensors and trick them. Why not be able to trick them to thinking there is a ship when it is just a decoy for example, sending the enemy out on a wild ghost chase for example.

I do think that the game would be more fun if sensors were not binary and there were ways to mask ships, say around planets and asteroids for example.

Sure... it would make "combat" in general more spread out in a system perhaps, but generally I would have no big problem with that. It also would mean that a better sensor are not guaranteed to find something before a worse one, just more likely to do so.

The calculation does not have to be more complicated than now, just a few more states before something is fully identified. Even a passive sensors detection should be able to eventually determine with a very high certainty what enemy design it is, you can in real life as all emissions usually are very unique for each type of platform. I also think this would be a good way to make "Flag Bridges" more interesting... they could be made to influence the chance to fully identify sensor sources as they combine the information from en entire squadron or small task force of ships.

The game used to be like that in the early years. For example, you would have a thermal contact but no ID or even IFF, so you had to get closer and try to identify ship type and race (similar to Harpoon). However, while that is interesting for the first few contacts, it became very tedious after micromanaging investigation of a hundred such contacts. I changed sensor detection to the current model, which assumes target identification, or even target motion analysis, was handled by the your staff behind the scenes so you could get on with deciding how to handle it.

If I ever get around to a more tactical game than Aurora, it would definitely include something on these lines.

But I would imagine it would be much the same as detecting active sensor strength/sensitivity you currently have to do in the game, so same mechanic. Given enough time you will identify them all and find them. If you get closer your chances just increase.

To be honest it should not be that more micro intensive than the current model. You would not require to get closer, you just would increase the chances to detect something earlier if you do.

[Steve Walmsley]

But I would imagine it would be much the same as detecting active sensor strength/sensitivity you currently have to do in the game, so same mechanic. Given enough time you will identify them all and find them. If you get closer your chances just increase.

To be honest it should not be that more micro intensive than the current model. You would not require to get closer, you just would increase the chances to detect something earlier if you do.

Yes, but everyone would have to close on every contact to ensure there were not more ships to be discovered before committing to action. Its the same principle. A lot more scouting needed, which is fun at first but gets a lot more tedious over time.

[kilo]

If there was one thing about sensor that I would change it is not the detection rules themselves, but introduce some sort of reaction time when it comes to activating them, spotting targets and allowing to shoot weapons at such a contact.

[Bremen]

If there was one thing about sensor that I would change it is not the detection rules themselves, but introduce some sort of reaction time when it comes to activating them, spotting targets and allowing to shoot weapons at such a contact.

I suppose what you could do is have a detection chance at longer ranges, like at extreme range you have a 1% chance to detect the target every 5 minutes, rising to 100% chance (instant detection) at half of the sensor's effective range. That could have a similar effect in that it would make it easier to detect large fleets than single ships because you'd probably hit 10x 4% chances faster than 1 4% chance - at the very least you'd be quicker to discover something was out there. It might also be interesting if at long ranges you were uncertain how many enemy ships were actually looking at.

On the other hand, IIRC detection is already a significant performance drain, and this could potentially make it several times more computationally expensive depending on how it's handled.

[kilo]

It would not have to be complicated at all. I think it is pretty weird that a crew can run the ship with active sensors turned off, turns them on for some reason and gets a true representation of the surrounding without any delay. I mean, there is no speed of light in the game, which is okay as it is, but data collection and processing could take time and calculating trajectories for enemy ships are impossible to do in an instant.

[Andrew]

The delay on activating the radars of an AEGIS ship and tracking all contacts is pretty much zero . Why should hyper advanced computer systems d  o worse, and why would a 5 sec delay actually improve the game?
Many of those 'new' contacts are already being tracked on passive so that further reduces the complexity for the computers.


This is a strategic game, when Aurora tactical combat simulator or Aurora ship commander comes out , then indeed much more detail orientated combat and sensor models will be needed as that will be the core of the game. Adding lots and lots of detail at the moment means that the player needs to take direct control of events every time a ship detects a new contact , which makes running 30 day or even 30 minute turns hard and moves the focus from the empire to commanding every single patrol ship in excruciating detail

[Steve Walmsley]

It would not have to be complicated at all. I think it is pretty weird that a crew can run the ship with active sensors turned off, turns them on for some reason and gets a true representation of the surrounding without any delay. I mean, there is no speed of light in the game, which is okay as it is, but data collection and processing could take time and calculating trajectories for enemy ships are impossible to do in an instant.

You could have target motion analysis. In effect, you take a bearing on the active contact, then move and take another bearing, then move again, etc. By triangulating all the bearings taken, you can calculate distance and speed of the target. However, once again this will be fun the first few times, but the 100th time would be tedious.

The aim of the sensor model isn't to reflect light speed concerns or processing time, because none of those add any real decision-making on the part of the player, other than to add additional micromanagement to get the same eventual information. Aurora is an operational/strategic game, rather than tactical, so its assumed your staff run TMA, etc. and present you with the information required to make decisions.

[Jorgen_CAB]

But I would imagine it would be much the same as detecting active sensor strength/sensitivity you currently have to do in the game, so same mechanic. Given enough time you will identify them all and find them. If you get closer your chances just increase.

To be honest it should not be that more micro intensive than the current model. You would not require to get closer, you just would increase the chances to detect something earlier if you do.

Yes, but everyone would have to close on every contact to ensure there were not more ships to be discovered before committing to action. Its the same principle. A lot more scouting needed, which is fun at first but gets a lot more tedious over time.

But they would not have to close, just follow the contacts and everything will be detected eventually. Would it not be practically the same as detecting sensor sensitivity/range is now?

Sure you will not know exactly when everything is detected, but that would be the point though... uncertainty.