Design: Detection and Visibility

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

Re: Design: Detection and Visibility

#31 Post by eleazar » Fri Aug 01, 2008 1:31 am

Geoff the Medio wrote:
eleazar wrote:With this implementation you can still conceal your stuff, but it doesn't require the careful player to do try to keep track of the number of planets in enemy systems to he knows when one is turned totally invisible.
Why is disappearing planets so much worse than disappearing ships or fleets?
Because it's relatively simple and straightforward to answer the question, "Am i now, scanning this system with a super-scanning-ship?"...
while it's much more annoying to have to answer the question, "Have i ever scanned this system with a super-scanning-ship?"
To find hidden planets you only have to search each system once (per tech level of scanning gear), which requires you to keep track of where you have scanned. This IMHO is anti-fun. When searching for stealthy ships you only need to consider the present.


The rest later.

User avatar
tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

Re: Design: Detection and Visibility

#32 Post by tzlaine » Sun Aug 03, 2008 8:35 pm

Geoff the Medio wrote:Some questions we still need to address include:
* Assuming the stealth / detection mechanics are the same for the galaxy map and the battle map, do we need entirely separate stealth and detection meters (and techs and ship parts and buildings, etc.) to set / modify / boost them? Either way, we can have effects that only alter the meters in one context or the other, but should the values be presented to players and stored internally as a single or multiple meters of each type?
I say yes. I'd like people to be able to concentrate one branch of the tech tree, and use that branch to exercise a strategic advantage. The two strategies here would be 1) You can see me coming, but you can't stop me from staying around/passing through your systems, and 2) You can't see my assault fleet coming until its too late, even though you can see all the ships once they arrive. I also think that omni-stealth will be too powerful -- I worry that it will allow a player to cruise around at will, undetected on either the galaxy or tactical maps.
* How far away, measured in units of the typical distance between systems on the map, should it be possible to see the following types of objects? This is about typical objects of their type, and doesn't imply that special cases couldn't be harder or easier to see.
** Systems (stars) - Currently all these can be seen by all players. However, this makes exploration of the galaxy less interesting and could potentially give advantages to human players over AIs in that humans are better able to infer what direction to scout based on the direction towards the centre of the galaxy.
I'd like to hide stars before they're explored. When faced with a branching path, the AI will have a hard time figuring out that taking the left fork is more likely to allow the exploration of all those stars on the left, assuming there's only one or two on the right. This is trivial for a human. AIs are hard to get right, so I don't want to handicap ours in the first "X".
** Planets - Currently all are visible after a system is explored, but if systems can be seen before reaching them, should planets as well? Being able to see into other systems without sending ships could be a valuable ability for certain strategies that don't use a lot of ships.
I vote for planets being visible only when a star is visited. They should be visible thereafter, but changes to their contents, like buildings, etc., should not change until they are revisited.
** Buildings - Currently buildings are visible when the planet they are on is visible, but it might be interesting to make it generally harder to see buildings... A much more refined or advanced long-range detection system would be useful if it allowed a player to see buildings in other systems that would otherwise be undetectable.
Same as for planets. I think all of a star system's contents should be visible to anyone currently in that system. This applies to defenses, starbases, buildings, etc. This does not apply to settings, like focus, etc.
** Other Empires' Fleets - Currently these are visible on starlanes adjacent to systems the player can see or in systems the player can see.
*** Should it matter if a fleet is in a system or travelling on a starlane?
Yes. This can be used to confer an advantage to the defender. Moving fleets should be more detectable.
** Starlanes - Currently these are visible after one of the systems they are connected to is explored (ie. the planets in the system are visible).
If stars are only revealed when explored, I'd also like the starlanes leading to them not to be fully revealed either. For example, if you reach star A and it has a starlane to star B, which is not yet visible, you should know that there's a starlane there, but not how long it is, until you actually follow it and reach B.
*** Could you see a fleet travelling on a starlane even if the lane itself isn't yet known / visible?
I say no. KISS.
* Are there levels of visibility? Fleets could be detected just as "a fleet", with no further information about contents, or the approximate number of ships could be known, or the exact number but no details about their designs, or rough classifications of designs could be known, or exact design information could be known. Similarly for planets, the environment type, whether it is populated, approximately how populated and how developed it is could be known at with sufficient levels of detection.
** What determines how much detail is available? One option is to provide more information depending on how much the detection meter exceeds an object's stealth meter and distance from the detector.
*** Detection >= Stealth + Distance ... Can see object and its basic type (ship, system, etc.)
*** Detection >= Stealth + Distance + 10 ... Can see some additional information
*** Detection >= Stealth + Distance + 20 ... Can see more additional information, etc.
** Even if the special case for detection, where objects with stealth 0 are always visible, is in effect, seeing details about the object could require meeting the above more-restrictive conditions.
This all sounds good to me. I don't have a strong opinion about what information the different detection levels should tell you though. I only want to apply stealth to ships, though (and to fleets, where a fleet's stealth is min(ship stealth meters)).
* Should the special case for detection, where objects with stealth 0 are always visible, that is mentioned in the battle map detection vs. stealth discussion in the v0.4 Design Pad, be applicable for galaxy map detection as well? If systems are always visible, this would make sense to do, but if systems are invisible unless a player gets a detecting object close enough to see them, this is perhaps not necessary.
** Then again, we might want a way to reliably make an object visible to the whole universe on the map without requiring another special case... Perhaps a star goes nova; the whole galaxy could be expected to see this.
** Even if systems aren't always visible, we could just give systems a base stealth of 0.01 (or other suitably small but non-zero number) so that they fall under standard detection rules, and not the stealth 0 special case.
I think we should mirror the tactical stealth mechanics in the galaxy map as closely as possible, unless there is good reason not to.
* Is it important or beneficial for players to get bonuses to detecting ships if their empire knows something about the ships that are being detected? In particular, if the player has knowledge of the ship's design, should that give a bonus, or is this too complicated to worry about? Assuming appropriate espionage functionality exists, having bonuses to detect known ship designs could make spying to steal ship design more useful, even if an empire has no plans to build such designs for its own use.
I think this would be a great addition to the spying minigame, but it will probably adding too much detail to incorporate scans from tactical battles, etc.
* How should we deal with cases where a more-visible object is contained within a less-visible object? For example, a planet might have stealth 5 and the system it is in might have stealth 20. A ship would see the planet before it could see the system. We can't (or won't) have any way to show an isolated planet, apparently without a system, on the map, so either the planet will remain invisible until the system can be detected, or the visible planet will force the system to be visible as well.
** If there are levels of visibility, say 1 and 2 where level 2 gives more detailed info than level 1, should a level 2 planet make its system visible at level 2 as well?
** What if the system is at level 1, which doesn't reveal the system's contents (planets) but the planet is a level 2? Revealing just the level 2 planet, but not any other planets in the system (since the system's visibility isn't high enough to reveal its contents), could be complicated and prone to UI issues and would likely be best avoided.
If we apply stealth only to ships/fleets, these problems go away. If not, I think a contained object's detection should be impossible if its containing object is not already detected.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Re: Design: Detection and Visibility

#33 Post by eleazar » Sun Aug 03, 2008 9:37 pm

tzlaine wrote:I say yes. I'd like people to be able to concentrate one branch of the tech tree, and use that branch to exercise a strategic advantage. The two strategies here would be 1) You can see me coming, but you can't stop me from staying around/passing through your systems, and 2) You can't see my assault fleet coming until its too late, even though you can see all the ships once they arrive.
So you want a detection tech that can see and identify enemy ships as long as they are in another star system, but is utterly worthless in your own system? That seems confusing. It's going to be clumsy to tell the player that his ship has Stealth(A) (prevents detection when far away) bonuses, instead of Stealth(B) (prevents detection from nearby stuff) bonuses. Besides why can't you scan ships using Stealth(B) from another system with Detection(A) and communicate their position to those ships fighting them?

tzlaine wrote:
** Starlanes - Currently these are visible after one of the systems they are connected to is explored (ie. the planets in the system are visible).
If stars are only revealed when explored, I'd also like the starlanes leading to them not to be fully revealed either. For example, if you reach star A and it has a starlane to star B, which is not yet visible, you should know that there's a starlane there, but not how long it is, until you actually follow it and reach B.
I don't see how that is at all practical. Starlanes go in a straight line. A human could easily trace the starlane's vector and figure out where it will probably end up. Even if unexplored starlanes pointed in a random direction, humans could guess where it will lead with reasonable accuracy if they have about half the lanes explored in the area.

I doubt this is your intention, but this would really add a lot of "busywork" to exploration relative to what we do now. I.E. you don't have to travel every starlane to have a section of space thoroughly explored. Following starlanes that are very likely to lead to stars you've already explored is not very fun. By some quick calculations, if every lane needs to be traveled to be explored, that could easily require twice as much travel time to explore the same area, or to look at it another way, half the amount of interesting things you'll find per turn of exploring.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Re: Design: Detection and Visibility

#34 Post by eleazar » Mon Aug 04, 2008 2:57 am

Geoff the Medio wrote:
eleazar wrote:I believe the feeling of pseudo-exploration (finding stars, but no info about them) will cease to be interesting rather quickly. It's just an intermediate time-burner before the player can find out something useful and interesting about a system. Basically it's fluff that makes part of the game take longer without adding much to it.
I don't think we're talking about the same thing here... Finding out that a system exists before you find out what's in it doesn't slow down anything, any more than knowing the location of all systems on the map before you've explored those systems would "slow down" exploration.
I was assuming you intended that each "discovered" star would get a sitrep message, which would indeed take time, as would messing with special tech to see further stars, but maybe you don't intend to emphasize star discovery with sitrep messages. Additionally with "low visibility" in the galaxy it may be the best thing to simply let your (traditionally slow) first colony ships sit around while the scouts explore or the super-telescope is built for reasons explained below, and this too slows things down.
Geoff the Medio wrote:Not knowing where systems are before you've gotten a ship to adjacent systems also doesn't slow you down, as you couldn't have sent a ship to the (newly-discovered) system anyway, since you wouldn't yet have found the starlanes that connect it to the surrounding systems, and you'd have to get a ship next to the system before it can go into the system in either case. Being able to know that something (a system) exists before getting details about it gives the player more information, not less. The difference is, I suppose, that instead of revealing the location and existence of all systems to players at the start of the game, we delay giving this information until they player can do something about it (ie. they have a nearby ship) or until another player gives them the information, in which case they probably can't immediately do much with the information anyway.
You aren't taking into account the strategic difference between a Space 4X (such as FO currently is) and Civ in exploration:

In Civ you are exploring nearly blindly. You can make uncertain guesses about how the coastline will go, and some things give you hints about latitude, but beyond that there's seldom any indication weather exploring North, S, E, or W is more likely to provide you with better land.

In many 4X space games the type of star indicates different probabilities of finding planets of different quality. So instead of exploring (nearly) blind, the player in the early game is making strategic choices: do i send my colony ship East towards one very promising star, or West towards three closely spaced stars, which are each less likely of having great planets. Does the player expand aggressively towards the galactic center, or does he focus on the uncontested rim, and risk becoming hemmed in.

These kinds of decisions IMHO improve the depth of early-game exploration. It allows the player to exercise foresight rather than just plunge (more or less) blindly into the void.

I grant that you don't need to see the entire 500 star galaxy to exercise this kind of strategic foresight, but you do need to see more than the immediately adjacent stars. So i wouldn't object to a gradual revelation of the stars per se, as long as the range of vision for revealing stars was long enough that the player can consistently make exploration plans that go further than the very next jump.

Geoff the Medio wrote:Also, regarding delayed exploration. A way to implement this would be to give increasing bonus to detection with time spent by a ship in same system. If planets (or buildings on them) have various stealth ratings, they would become (more) visible over time as the ship's detection meter increases due to is stationary time-dependent bonus.
I have strong reservations about rewarding prolonged detection times any greater than a full turn (it does seem reasonable that a ship entering a system, and, immediately leaving shouldn't be able to accomplish the same quality of detection as one stationed there it's whole turn) especially along with game concepts like totally invisible stars/planets. If it takes a long time to find something, the player is going to pushed into having detector ships sitting all over the place, and how is he going to know if he's scanned long enough? How is he going to remember which systems he's scanned for X turns with X tech?
...
Well, i guess i could see an argument for having the stealth value of a stationary ship degrade over time. This would make sneaking through a system easier than gradually building up an invisible invasion fleet in the same place.

Geoff the Medio wrote:Also also, regarding having multiple ways to do things like detection (espionage, ships exploring, long range sensor techs). Consider the Terrans in Starcraft. They can detect using comsat, missile turrets or science vessels. Each has pros and cons and picking which to use, or whether to use any, has important strategic consequences and might depend on whether the opponent is going for fast mobile stealthy units (ie. dark templar rush) or not. Having these options can be a good thing, if they are strategically distinct.
I'm only slightly familiar with Starcraft (i'm not fond of RTS), however it's not important. We're not in disagreement about the value of having options with deeply strategic consequences. The disagreement is because i don't see much strategic interest/value in some of the ideas proposed in this thread, or in some cases, i see some value, but anticipate some associated problems which negate the value. Also i try to aim for the sweet-spot between minimum complexity and maximum strategic gameplay.

User avatar
tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

Re: Design: Detection and Visibility

#35 Post by tzlaine » Mon Aug 04, 2008 1:28 pm

eleazar wrote:
tzlaine wrote:I say yes. I'd like people to be able to concentrate one branch of the tech tree, and use that branch to exercise a strategic advantage. The two strategies here would be 1) You can see me coming, but you can't stop me from staying around/passing through your systems, and 2) You can't see my assault fleet coming until its too late, even though you can see all the ships once they arrive.
So you want a detection tech that can see and identify enemy ships as long as they are in another star system, but is utterly worthless in your own system?
No. I'm not sure where you got that idea. A fleet that is detectable in another star system will also be detectable in yours, assuming the same detection equipment (i.e. meter) is being used on both systems.
That seems confusing. It's going to be clumsy to tell the player that his ship has Stealth(A) (prevents detection when far away) bonuses, instead of Stealth(B) (prevents detection from nearby stuff) bonuses. Besides why can't you scan ships using Stealth(B) from another system with Detection(A) and communicate their position to those ships fighting them?
Because they are different kinds of detection. Galaxy map (strategic) detection tells you that a fleet is in a location, such as a system. Tactical map detection tells you exactly where the fleet is within the system. This is analogous to knowing that a fleet is in the Mediterranean Sea because you saw it pass Gibraltar, but now you don't know exactly where it is within the Sea. To actually attack it, you need more precise info than "it's around here somewhere". I don't see a contradiction here.
eleazar wrote:
tzlaine wrote:
** Starlanes - Currently these are visible after one of the systems they are connected to is explored (ie. the planets in the system are visible).
If stars are only revealed when explored, I'd also like the starlanes leading to them not to be fully revealed either. For example, if you reach star A and it has a starlane to star B, which is not yet visible, you should know that there's a starlane there, but not how long it is, until you actually follow it and reach B.
I don't see how that is at all practical. Starlanes go in a straight line. A human could easily trace the starlane's vector and figure out where it will probably end up. Even if unexplored starlanes pointed in a random direction, humans could guess where it will lead with reasonable accuracy if they have about half the lanes explored in the area.
It may not be perfect, but it is better at hiding information about other star systems' positions than the current system. I'm certainly open to alternatives.
I doubt this is your intention, but this would really add a lot of "busywork" to exploration relative to what we do now. I.E. you don't have to travel every starlane to have a section of space thoroughly explored. Following starlanes that are very likely to lead to stars you've already explored is not very fun. By some quick calculations, if every lane needs to be traveled to be explored, that could easily require twice as much travel time to explore the same area, or to look at it another way, half the amount of interesting things you'll find per turn of exploring.
I think you're overstating the amount of busywork that will happen. Part of the problem is that I overstated the rule I intended: "until you actually follow it and reach B" should have read "until you actually follow it and reach B, or unless you have already explored B". You shouldn't need to travel every starlane, but you should need to actually visit a star to know what starlanes lead there.

User avatar
tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

Re: Design: Detection and Visibility

#36 Post by tzlaine » Mon Aug 04, 2008 1:40 pm

eleazar wrote:
Geoff the Medio wrote:Not knowing where systems are before you've gotten a ship to adjacent systems also doesn't slow you down, as you couldn't have sent a ship to the (newly-discovered) system anyway, since you wouldn't yet have found the starlanes that connect it to the surrounding systems, and you'd have to get a ship next to the system before it can go into the system in either case. Being able to know that something (a system) exists before getting details about it gives the player more information, not less. The difference is, I suppose, that instead of revealing the location and existence of all systems to players at the start of the game, we delay giving this information until they player can do something about it (ie. they have a nearby ship) or until another player gives them the information, in which case they probably can't immediately do much with the information anyway.
You aren't taking into account the strategic difference between a Space 4X (such as FO currently is) and Civ in exploration:

In Civ you are exploring nearly blindly. You can make uncertain guesses about how the coastline will go, and some things give you hints about latitude, but beyond that there's seldom any indication weather exploring North, S, E, or W is more likely to provide you with better land.

In many 4X space games the type of star indicates different probabilities of finding planets of different quality. So instead of exploring (nearly) blind, the player in the early game is making strategic choices: do i send my colony ship East towards one very promising star, or West towards three closely spaced stars, which are each less likely of having great planets.
This is exactly the kind of info that a human player should not have, because she is able to take advantage of it easily, whereas an AI is not. I could live with giving the AI the entire map at the outset of the game too. I'm not above cheating. ;) In fact, I agree that the hiding of info from the player makes the game a bit more tedious -- I just don't want to build in a human advantage.
Does the player expand aggressively towards the galactic center, or does he focus on the uncontested rim, and risk becoming hemmed in.
Note that this strategic choice is available to both humans and AIs under all regimes suggested so far.
Geoff the Medio wrote:Also, regarding delayed exploration. A way to implement this would be to give increasing bonus to detection with time spent by a ship in same system. If planets (or buildings on them) have various stealth ratings, they would become (more) visible over time as the ship's detection meter increases due to is stationary time-dependent bonus.
I have strong reservations about rewarding prolonged detection times any greater than a full turn (it does seem reasonable that a ship entering a system, and, immediately leaving shouldn't be able to accomplish the same quality of detection as one stationed there it's whole turn) especially along with game concepts like totally invisible stars/planets. If it takes a long time to find something, the player is going to pushed into having detector ships sitting all over the place, and how is he going to know if he's scanned long enough? How is he going to remember which systems he's scanned for X turns with X tech?
...
Well, i guess i could see an argument for having the stealth value of a stationary ship degrade over time. This would make sneaking through a system easier than gradually building up an invisible invasion fleet in the same place.
I agree that this is a bad idea, and for largely the same reasons. If I have to choose between moving through a system automatically and stopping and lingering for a turn or more, I'm going to have to micromanage the movement of my scouts. Yuck.

User avatar
MikkoM
Space Dragon
Posts: 318
Joined: Fri Mar 10, 2006 12:32 pm
Location: Finland

Re: Design: Detection and Visibility

#37 Post by MikkoM » Mon Aug 04, 2008 2:35 pm

tzlaine wrote:
Geoff the Medio wrote: * How far away, measured in units of the typical distance between systems on the map, should it be possible to see the following types of objects? This is about typical objects of their type, and doesn't imply that special cases couldn't be harder or easier to see.
** Systems (stars) - Currently all these can be seen by all players. However, this makes exploration of the galaxy less interesting and could potentially give advantages to human players over AIs in that humans are better able to infer what direction to scout based on the direction towards the centre of the galaxy.
I'd like to hide stars before they're explored. When faced with a branching path, the AI will have a hard time figuring out that taking the left fork is more likely to allow the exploration of all those stars on the left, assuming there's only one or two on the right. This is trivial for a human. AIs are hard to get right, so I don't want to handicap ours in the first "X".
This would seem like a good in game reason to hide the stars. However even if we do hide the stars from the player and make the gaseous substance appear so that it doesn`t immediately give away the stars positions, the player still chooses what kind of a shape he/she wants for the galaxy and how many stars there will be in that galaxy. (And in my opinion this is a good thing.) So if the player chooses a galaxy with a regular shape can we really hide this kind of a galaxy so well that it isn`t easy for the player to determine at which direction he/she is more likely to find most of the stars? Can we make the starting position so vague that a human player hasn`t got a clue at which direction the galaxy expands. Because if the player can easily determine at which direction the galaxy expands, then he/she will have an advantage compared to the AI even if the stars are hidden. Also seeing all of the stars immediately and knowing the shape of the galaxy might help the player to determine where he/she wants to expand and possibly even what kind of a shape he/she wants his/her empire to have, as eleazar already said.

As for delayed exploration I would consider this to be a too good of an idea to miss, if we truly want to have an unique and interesting exploration system. Since if we will just have scouts that follow preset starlanes, arrive at a star system and discover all of its contents at once and then move on, there is really no interesting decision making or discoveries in our exploration system. Scouts just arrive at star systems, reveal their content and move on, and the player probably checks this new information, but his/her primary interest is in exploring as many systems as possible and sending colony ships to the most juicy ones. In this kind of a system info about individual star systems becomes less important as the player`s focus is in mass exploration, revealing as much useful info as quickly as possible.

Whereas if there is a delayed discovery of in system contents such as specials, space mosters etc. the player will hopefully be faced with a choice to either move on strait away and try to find as many planets as he/she can that look ideal for his/her race, or to stay in a particular system for a while and discover what is the true value of that system and the planets in it, as he/she reveals possible specials, monsters, minor races etc. And if we will make colony ships not very easy to come by in the early game, this will hopefully support detailed exploration of individual systems, since you probably want to make sure that your not so numerous colony ships reach their destinations and colonise planets that offer you the maximum benefit.

With delayed discoveries the player has something to look forward to even when he/she has discovered what kind of planets a star system contains and/or where this star is located in the galaxy and what kind of a star it is. There is the excitement of detailed discoveries if the player decides to stay and explore a particular system more thoroughly and this information also helps him/her to decide where to send his/her colony ships.

Personally I am hoping that our space exploration could be as exciting or at least nearly as exciting as sports game drafts and a delayed discovery system seems like a good way to achieve this goal. That is of course if we can balance it so that searching an individual system doesn’t take too long, but at the same time takes enough time that it can be separated from instantly getting all of the information.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Re: Design: Detection and Visibility

#38 Post by eleazar » Mon Aug 04, 2008 8:25 pm

tzlaine wrote:
eleazar wrote:
tzlaine wrote:I say yes. I'd like people to be able to concentrate one branch of the tech tree, and use that branch to exercise a strategic advantage. The two strategies here would be 1) You can see me coming, but you can't stop me from staying around/passing through your systems, and 2) You can't see my assault fleet coming until its too late, even though you can see all the ships once they arrive.
So you want a detection tech that can see and identify enemy ships as long as they are in another star system, but is utterly worthless in your own system?
No. I'm not sure where you got that idea. A fleet that is detectable in another star system will also be detectable in yours, assuming the same detection equipment (i.e. meter) is being used on both systems.
No, i really don't understand what you mean. I can't reconcile all your statements.

Are you proposing?
* Detector type A could detect the presence of a fleet, but not any specifics about position or composition, it works at long range.
* Detector type B can detect the specific location and composition of a fleet, but it only works locally.
* Both types of detectors are countered by different stealth techs

tzlaine wrote:[To actually attack it, you need more precise info than "it's around here somewhere". I don't see a contradiction here.
To be able to "see" ships in such detail that you know they are 4 Mark II Crusiers (with one tech), but you can't "see" them (with the other tech) well enough to know where they are certainly is very counterintuitive.
eleazar wrote:
tzlaine wrote:
** Starlanes - Currently these are visible after one of the systems they are connected to is explored (ie. the planets in the system are visible).
If stars are only revealed when explored, I'd also like the starlanes leading to them not to be fully revealed either. For example, if you reach star A and it has a starlane to star B, which is not yet visible, you should know that there's a starlane there, but not how long it is, until you actually follow it and reach B.
I don't see how that is at all practical. Starlanes go in a straight line. A human could easily trace the starlane's vector and figure out where it will probably end up. Even if unexplored starlanes pointed in a random direction, humans could guess where it will lead with reasonable accuracy if they have about half the lanes explored in the area.
It may not be perfect, but it is better at hiding information about other star systems' positions than the current system. I'm certainly open to alternatives.[/quote]
So why do we need to hide information about immediately adjacent stars? At minimum exploring a system should reveal all connecting starlanes (as it does now, ignoring the possibility of hidden lanes) and reveal the star types of all stars those lanes lead to. All problems solved.

tzlaine wrote:
eleazar wrote:In many 4X space games the type of star indicates different probabilities of finding planets of different quality. So instead of exploring (nearly) blind, the player in the early game is making strategic choices: do i send my colony ship East towards one very promising star, or West towards three closely spaced stars, which are each less likely of having great planets.
This is exactly the kind of info that a human player should not have, because she is able to take advantage of it easily, whereas an AI is not. I could live with giving the AI the entire map at the outset of the game too. I'm not above cheating. ;) In fact, I agree that the hiding of info from the player makes the game a bit more tedious -- I just don't want to build in a human advantage.
At anything that involves "foresight" humans and AI are going to operate in radically different ways. If we are ruthlessly going to eliminate any gameplay element which humans or AI would tend to have any advantage in, there is not going to be any game left.

I admit i can't code an AI, but i have a hard time believing that this would be a significant advantage to human players. It seems mathematically pretty straight-forward calculate which direction of exploration has the best chance of finding good planets. (Assuming that universe generation is fixed to generate planets at the probabilities it promises). I have never noticed that AI players were at a disadvantage in this area in any of the 4X games, so creating an adequate AI can't be too hard.
There is enough uncertainty in the galaxy that even a completely random explorer (assuming it's just smart enough to avoid re-explore the same stars) would probably not be significantly disadvantaged to a human player who can see all or part of the map.

tzlaine wrote:
eleazar wrote:Does the player expand aggressively towards the galactic center, or does he focus on the uncontested rim, and risk becoming hemmed in.
Note that this strategic choice is available to both humans and AIs under all regimes suggested so far.
Not if the player doesn't know where the center is. The logical outworking of your ideas is that the center should be initially unknown, and after your post, MikkoM suggests outright that it should be unknown.

tzlaine wrote:I agree that this is a bad idea, and for largely the same reasons. If I have to choose between moving through a system automatically and stopping and lingering for a turn or more, I'm going to have to micromanage the movement of my scouts. Yuck.
You see that there are similar issues with having totally invisible planets/stars which can only be revealed by a scanning ship of an arbitrary and unknown level? I.E. the player has to sweep sweep his part of the galaxy each time he researches a new scanning tech.

MikkoM wrote:However even if we do hide the stars from the player and make the gaseous substance appear so that it doesn`t immediately give away the stars positions, the player still chooses what kind of a shape he/she wants for the galaxy and how many stars there will be in that galaxy. (And in my opinion this is a good thing.) So if the player chooses a galaxy with a regular shape can we really hide this kind of a galaxy so well that it isn`t easy for the player to determine at which direction he/she is more likely to find most of the stars? Can we make the starting position so vague that a human player hasn`t got a clue at which direction the galaxy expands.
I doubt we can hide it from the player for long. The human eye/brain is brilliant at pattern recognition, even compared to the best AI. It seems just as likely that the most level human/AI playing field would be in a totally revealed galaxy.

User avatar
Krikkitone
Creative Contributor
Posts: 1504
Joined: Sat Sep 13, 2003 6:52 pm

Re: Design: Detection and Visibility

#39 Post by Krikkitone » Mon Aug 04, 2008 10:23 pm

I'd say

Star=Detected as soon as you see a starlane that connects to it (realism explanation: there are bazillions of stars visible through a telescope, but only starlane ones are accessible/important.. and therefore only those ones show up on the galaxy map)..Basic characters determined (color)

Starlane=detected as soon as you are in a system it connects to, (may also have techs that allow it to be detected, or through contact with another player)

Planet=Detected when in system... or out of system with sufficient tech
Size+env determined with planet
Specials may have greater levels of 'stealth' (will automatically be discovered if they affect colonists you have)

Buildings
Planetary populations
planetary meters
etc.

Have fairly high degrees of stealth.. and that stealth Should increase with tech... "spies" (including passive info gathering) should be the best at getting this data.

example, the Construction meter would make the value of the contruction meter easier to see. (ie if it is 20+ you can tell but 0-19.999 all look the same (unless your detection improves/you get closer then you can tell planets 15+ from uncolonized rocks)... similar with population... and economy meters. (only the owner can see the actual focus)


and Detection+Stealth Tech should be one single meter and the same for both galaxy and tactical maps

However I would say a ship is MUCH easier to detect when travelling along a starlane. (so that you might be able to tell a ship came into your system, but it is hiding in the system now.

User avatar
tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

Re: Design: Detection and Visibility

#40 Post by tzlaine » Wed Aug 06, 2008 1:42 pm

eleazar wrote:
tzlaine wrote:A fleet that is detectable in another star system will also be detectable in yours, assuming the same detection equipment (i.e. meter) is being used on both systems.
No, i really don't understand what you mean. I can't reconcile all your statements.

Are you proposing?
* Detector type A could detect the presence of a fleet, but not any specifics about position or composition, it works at long range.
* Detector type B can detect the specific location and composition of a fleet, but it only works locally.
* Both types of detectors are countered by different stealth techs
Something like that. The idea is to have a strategic detection-and-stealth system, that determines what is visible on the galaxy map, and a separate and independent tactical detection-and-stealth system, that determines what is visible on the battle map. The level of detail available to both systems is not something I have a firm idea about, though presumably more detail would be available on the tactical level.
tzlaine wrote:[To actually attack it, you need more precise info than "it's around here somewhere". I don't see a contradiction here.
To be able to "see" ships in such detail that you know they are 4 Mark II Crusiers (with one tech), but you can't "see" them (with the other tech) well enough to know where they are certainly is very counterintuitive.
Why does that matter? Rules of a made-up game are inherently arbitrary. Is the 1+3 movement pattern of the Knight in chess intuitive? I believe UI should be intuitive to whatever degree we can manage, but how does intuitiveness apply here? I'm giving reasons for why I think the rules should look like this. My reasons are in terms of providing the player interesting strategic and tactical choices based on this mechanic. Do you dispute my reasoning there, or can you site an advantage to making all detection uniform across the tactical and galaxy maps?
eleazar wrote:I don't see how that is at all practical. Starlanes go in a straight line. A human could easily trace the starlane's vector and figure out where it will probably end up. Even if unexplored starlanes pointed in a random direction, humans could guess where it will lead with reasonable accuracy if they have about half the lanes explored in the area.
It may not be perfect, but it is better at hiding information about other star systems' positions than the current system. I'm certainly open to alternatives.
So why do we need to hide information about immediately adjacent stars? At minimum exploring a system should reveal all connecting starlanes (as it does now, ignoring the possibility of hidden lanes) and reveal the star types of all stars those lanes lead to. All problems solved.
I'm not sure why you're saying that exploring a system "should" reveal all connecting starlanes. Why should they? And why should we reveal the stars and types of stars they lead to? I'm giving a reason why I think what I think, whereas I can't figure out why you've come to the conclusions you have.

Taking a guess, if it's the issue of the direction that starlanes point, we can always randomize them. Before you've either traveled a starlane or seen the stars at either end, you don't know where it goes.
tzlaine wrote:
eleazar wrote:In many 4X space games the type of star indicates different probabilities of finding planets of different quality. So instead of exploring (nearly) blind, the player in the early game is making strategic choices: do i send my colony ship East towards one very promising star, or West towards three closely spaced stars, which are each less likely of having great planets.
This is exactly the kind of info that a human player should not have, because she is able to take advantage of it easily, whereas an AI is not. I could live with giving the AI the entire map at the outset of the game too. I'm not above cheating. ;) In fact, I agree that the hiding of info from the player makes the game a bit more tedious -- I just don't want to build in a human advantage.
At anything that involves "foresight" humans and AI are going to operate in radically different ways. If we are ruthlessly going to eliminate any gameplay element which humans or AI would tend to have any advantage in, there is not going to be any game left.
Not true. Checkers is a game in which a human can sum up a player's position at a glance, whereas the computer cannot. However, it happens that it's really easy for the computer to do so after a reasonable period of time. The best checkers-playing computer programs are better than all human players, so there is no insurmountable advantage, and there's still a "game left". However, the graph-layout of a FO galaxy is very different, and a lot more computationally difficult to digest, than an 8x8 grid. Our design should reflect that.

Also, the star type currently doesn't reflect the quality of a system to a noticable degree, afaict. I also don't know of any design guideline that would make that the case in the future.
I admit i can't code an AI, but i have a hard time believing that this would be a significant advantage to human players. It seems mathematically pretty straight-forward calculate which direction of exploration has the best chance of finding good planets. (Assuming that universe generation is fixed to generate planets at the probabilities it promises). I have never noticed that AI players were at a disadvantage in this area in any of the 4X games, so creating an adequate AI can't be too hard.
This is hard for me to swallow. I've seen, not just ocassionally, but over and over again, 4X AIs make stupid colonization choices. Such choices are a function of knowledge of the galaxy map, and the locations of enemies within it. I'd like to make an AI that makes great colonization choices, not just mediocre-to-bad ones.
tzlaine wrote:
eleazar wrote:Does the player expand aggressively towards the galactic center, or does he focus on the uncontested rim, and risk becoming hemmed in.
Note that this strategic choice is available to both humans and AIs under all regimes suggested so far.
Not if the player doesn't know where the center is. The logical outworking of your ideas is that the center should be initially unknown, and after your post, MikkoM suggests outright that it should be unknown.
I'm not suggesting anything like this, MikkoM is. I don't like his idea for the same reasons you do. So again I say that the location of the galaxy center and the general shape of the galaxy are things that the players and AIs all have access to.
tzlaine wrote:I agree that this is a bad idea, and for largely the same reasons. If I have to choose between moving through a system automatically and stopping and lingering for a turn or more, I'm going to have to micromanage the movement of my scouts. Yuck.
You see that there are similar issues with having totally invisible planets/stars which can only be revealed by a scanning ship of an arbitrary and unknown level? I.E. the player has to sweep sweep his part of the galaxy each time he researches a new scanning tech.
You've lost me. What is a scanning ship? What does a "scanning ship of an arbitrary and unknown level" mean? I'm talking about exploring the galaxy by movement alone, just like in all the MOOs.
MikkoM wrote:However even if we do hide the stars from the player and make the gaseous substance appear so that it doesn`t immediately give away the stars positions, the player still chooses what kind of a shape he/she wants for the galaxy and how many stars there will be in that galaxy. (And in my opinion this is a good thing.) So if the player chooses a galaxy with a regular shape can we really hide this kind of a galaxy so well that it isn`t easy for the player to determine at which direction he/she is more likely to find most of the stars? Can we make the starting position so vague that a human player hasn`t got a clue at which direction the galaxy expands.
I doubt we can hide it from the player for long. The human eye/brain is brilliant at pattern recognition, even compared to the best AI. It seems just as likely that the most level human/AI playing field would be in a totally revealed galaxy.
Agreed.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Re: Design: Detection and Visibility

#41 Post by eleazar » Wed Aug 06, 2008 4:15 pm

tzlaine wrote:
eleazar wrote: I doubt we can hide it from the player for long. The human eye/brain is brilliant at pattern recognition, even compared to the best AI. It seems just as likely that the most level human/AI playing field would be in a totally revealed galaxy.
Agreed.
Um, so are you that the other end of starlanes should be hidden even in an initially revealed galaxy,
or that they should only be hidden if we start out by hiding stars?

User avatar
tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

Re: Design: Detection and Visibility

#42 Post by tzlaine » Thu Aug 07, 2008 1:10 pm

Sorry if this wasn't clear, but I only want to hide starlanes if we also hide all the stars. And I favor hiding both until they are explored.

User avatar
MikkoM
Space Dragon
Posts: 318
Joined: Fri Mar 10, 2006 12:32 pm
Location: Finland

Re: Design: Detection and Visibility

#43 Post by MikkoM » Sun Aug 17, 2008 2:24 pm

tzlaine wrote: I agree that this is a bad idea, and for largely the same reasons. If I have to choose between moving through a system automatically and stopping and lingering for a turn or more, I'm going to have to micromanage the movement of my scouts. Yuck.
tzlaine wrote: I'm talking about exploring the galaxy by movement alone, just like in all the MOOs.
You lost me here. So you would like to hide the stars and explore the galaxy by movement alone. But doesn`t this mean that you have to micromanage the movement of your scouts anyway, as you move them to new star systems, when you discover new starlanes? And is this so radically different from moving your scout to a new star system and then deciding whether you want to leave it there for a turn or two to discover possible specials, space monsters, natives etc. or moving your scout straight away to a new star system? Also if you are planning to have some sort of an automated scouting system, where the scouts automatically explore the galaxy, couldn’t such a system also include an option, which would order your scouts to stop at every system to allow more detailed discoveries?

And can you honestly say that you usually have too much to do at the start of a space empire game; to not to be interested in moving your scouts around and finding about the galaxy that surrounds your star systems?

If we would have a delayed discovery system, it should of course be clear to the player that he/she has to spend X amount of turns in a star system to discover the hidden information. Since this way we could avoid situations, where the player has to ask him/herself if he/she has been scanning long enough. Also the X amount of turns should probably be something like 1 or 2, since otherwise exploring a large galaxy could take forever. I am also not suggesting that there should be any sort of a scanning button to start the detailed exploration of a star system. Instead of that I am suggesting that this more thorough exploration process would begin automatically, when a scout ship has stopped to an unknown star system.

As I already mentioned in my previous post I am hoping that a delayed discovery system could offer some options to the player to choose from as he/she explores the galaxy. He/she could either move scouts quickly from system to system, and discover as many planets that would appear useful as possible. And also possibly plan things like empire shape. Or he/she could make detailed scans of those star systems that he/she finds and at the same time be convinced that those systems, that he/she colonizes don`t have any nasty surprises awaiting him/her.
tzlaine wrote:
eleazar wrote:Does the player expand aggressively towards the galactic center, or does he focus on the uncontested rim, and risk becoming hemmed in.
Note that this strategic choice is available to both humans and AIs under all regimes suggested so far.
Not if the player doesn't know where the center is. The logical outworking of your ideas is that the center should be initially unknown, and after your post, MikkoM suggests outright that it should be unknown.
I wasn`t so much suggesting anything here, I was more like asking a question. Our reason for hiding the stars would seem to be that AI`s have a hard time deciding at, which direction to expand, and so we could even the odds by hiding the stars. However if even after we have hid the stars a human player doesn`t have too much trouble at figuring out where the galactic center and probably most of the stars are, can we also offer the AI this same information, so that both human and AI players would have an equal starting position and level of knowledge?

And what comes to the hiding the stars themselves, I have nothing against it if we truly can achieve some in game benefit from it, and we can also come up with some good technobable explanation for it, since as eleazar already said, based on their experiences players will consider stars as something very visible and will wonder why the stars in our game are hidden at first. However I think that Krikkitone already offered us a good start over here:
Krikkitone wrote: Explanation: there are bazillions of stars visible through a telescope, but only starlane ones are accessible/important.. and therefore only those ones show up on the galaxy map)..Basic characters determined (color)
So here is my current suggestion. Star lanes speed up interstar travel considerably, thus allowing galactic empires to reach these stars in reasonable time. Soas Krikkitone said, only these stars are accessible. However when an empire discovers one end of a starlane it doesn`t know exactly to which star it leads to. It might however know the length of this star lane based on some fancy calculations of some particle movement on the star lane (star lane particles :wink: ), if we need to offer players an ETA to an unknown star system. This way we would have an explanation of why at the start of the game the player only sees the back round stars and those star systems that are already known to him/her, but not the real stars.

User avatar
tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

Re: Design: Detection and Visibility

#44 Post by tzlaine » Sun Aug 17, 2008 8:12 pm

Regarding automation of exploration of my idea of low-information starlanes, the automation would just pick a starlane and move down it, same as if it knew the destination of all the starlanes and picked one. However, under my proposal, it's actually easier for the automation to do what the user would do (thus subverting the impulse of the player to micromanage automatic scouting because the automation is stupid), because there is no information, and so the decision to move down any one starlane is always the best given the current information.

Regarding delayed scouting, what does this extra bit of design complexity add to the game? What are the multiple strategic/tactical options available to the player if we do delayed scouting, as opposed to the "normal" Moo-type scouting?

User avatar
Krikkitone
Creative Contributor
Posts: 1504
Joined: Sat Sep 13, 2003 6:52 pm

Re: Design: Detection and Visibility

#45 Post by Krikkitone » Mon Aug 18, 2008 4:14 pm

As for delayed discovery of specials,.... I wouldn't limit it to 1 or 2 turns, some specials might take 10-20 turns (if you had early sensor tech)... so they are a surprise/are only going to be detected in nearby systems.

This would give you a reason to have Multiple scout ships.... one for every system (that you hadn't colonized)
It would also have a reason to improve your sensor technology besides detecting enemy ships.

It would mean that even though you scoutd this far off territory in your neighbor's back yard, they know more about it than you do. because they actually parked some scout ships there.


Now some specials should be instantly detected, some should take a turn and some should take a intensive survey effort.
(those would be the three 'automation' options for scout ships... go on to new starlanes ASAP, stay for a turn, or move to a system that doesn't have a survey ship and just 'Park'.

Locked