I''ve been thinking about how best to implement visibility systems in an RTS. It seems to me that there are a wide variety of options... Fog of war Unexplored terrain is blacked out, explored terrain which is not currently visible is greyed out. Limited vision This is the system used by 99% of RTS games. Each unit has a ''vision range'' which determines how far it can see. Enemy units can only be seen if they are within range of one you your units. Limited visibility Rather than each unit having a different ''vision range'', units have a different ''visibility range''. A tank has a larger visibility range than a soldier for example, and therefore it will be visible from a greater distance. Terrain might also have an effect on a units visibility range. Line of sight only Visibility is not limited by distance, but line of sight only. All visible All units on the map can be seen at all times. I''m curious to hear people''s thoughts on how these systems effect the gameplay in a real time strategy game. Also, are there any other systems that need to be included? Other features, e.g directionality? (with all the systems above, a unit is just as likely to spot someone standing behind him as he is likely to spot someone standing in front)

There''s also:

Player vision
For 3d games the likes of Shogun or Myth. The camera''s movement is restricted around a commander unit, limiting the player''s visibility. Interesting twists like seeing rising dust over a hill, but not the army that created that dust, or the active use of 3d shadows or other graphical effects are possible.

The scouts in Dark Reign that could morph in any game sprite around them also fit here, because they depend on how believable their cover is to the player.

Radar/Sonar
Total Annihilation. Enough said.

The effect of a visibility system is very dependent on the other gameplay rules. What it all boils down to is how difficult it is to get information and how useful is the information. Both of these are determined to a large extent by the rest of the rules. For instance, extensive use of units with longer firing range than view range in TA makes constant scouting vital. Counter-units in AoE and very long range cannons or nuclear weapons in TA make even a peek of the enemy base very important. Anti-scouting defense measures such as walls in AoE or arrays of Missile Towers and fighter patrols or the Radar-AntiRadar game in TA become important.

I think there must be a balance between what is known and what isn''t known in a game. Knowledge isn''t solely defined by the game-state (what units are where), but also by the plans of the enemy. In a game like chess for instance, even if the game state is completely known, the enemy plans are hidden, and very hard to discover. In a RTS, there is a lot more knowledge in the game state, and that''s why limiting visibility becomes important. If you see a large army moving in the direction of your base, you more or less know what the enemy wants and what you must do as a response. Games like RISK or Panzer General where luck is very important can also live without a lot of visibility limits, because even with complete information of the game state the future of the game isn''t predictable.

quote:Original post by Diodor
In a RTS, there is a lot more knowledge in the game state, and that''s why limiting visibility becomes important. If you see a large army moving in the direction of your base, you more or less know what the enemy wants and what you must do as a response.

Why do you say there is more knowledge in the game state in RTS games? This sounds like a symptom of missing depth to me.

Imagine a game in which visibility is limited only by line of sight - distance is irrelevant. Units are able to hide in certain types of terrain however.

Player 1 parks his artillery on top of a big hill, so that it has a commanding view of the terrain.

Player 2 parks his artillery behind a hill, so it is harder to spot. Of course, they have a fairly crummy line of sight in this spot, so he uses hidden infantry units to spot for them.

It''s hard to say which is better, since a great deal of it depends on the exact layout of the map, the placing of hills and the concealing terrain etc, although I''d probably put my money on player 2. His artillery and scouts are all hidden, so player 1 will have to send his own scouts out, or else guess their location, in order to win. In the meantime, player 1''s artillery will be taking a pounding.

The purpose of this example is to show that a high visibility system (ie, a system which hides relatively little from the players) can add a level of subtlety and strategy which is lost with a low visibility system.

quote:
Original post by Sandman

Why do you say there is more knowledge in the game state in RTS games? This sounds like a symptom of missing depth to me.

Lack of depth is exactly what I implied. I don''t think I need to prove the original comparisson between chess and a RTS. Limited visibility makes up for lack of depth. Your own example of an interesting choice is based on (lack of) visibility. In a game with complete visibility, both players would just fire their guns at each other.

That being said, I prefer games with high visibility. Risk, chess, go, Z are some examples.

quote:Original post by Diodor
Your own example of an interesting choice is based on (lack of) visibility. In a game with complete visibility, both players would just fire their guns at each other.

Well, yes and no. The example was of a high visibility game. Take the same example in a FULL visibility game and it becomes uninteresting: as you say, both players just shoot each other. Take the example in a very low visibility game (e.g the standard fog of war + limited vision) and it also becomes uninteresting - it''s just a case of wandering your scouts around blindly until they walk into the enemy base.

In an RTS I''m `working on` (still thinking it out, but planning on eventually actually creating it) I''m planning on creating a hybrid of complete and limited vision. The minimap will show the location of every unit, but instead of showing where they are, it will show where they where X seconds ago. It will only show that something is there and not what unit(s) or exactly how many (aka if many units are positioned close together it will just be a larger dot than normal).
Since I think I''ll have something simmilar to the `creeps` of warcraft 3 (which are computer controlled units hostile to all players that can''t build structures or more units), it won''t be easy to know if a dot on the minimap is the enemy player, a `creep`, a bird flying around, an animal running through a forest, etc. I will probably have the dots dissapear one in a while to reappear Y seconds later, to make it impossible to track enemy units perfectly. Some units might even be able to ''hide'' to avoid being shown at all on the minimap.

Each unit will also have a vision cone (more like the field of vision in an FPS than the normal circle of vision, but it will start slightly behind a unit to simulate ''feeling a presence nearby''), and you can only see units/structures on the actual map if it falls inside a vision cone of a unit you control or an allied unit. Unit facing will actually matter, and it will be easy to change facing in-game. Anything visible in a vision cone will be shown in the current position on the minimap in a special color depending on its owner (in addition to being shown in its old location as a `generic something`)

quote:
Original post by Iron Chef Carnage

Z?

Z Steel Soldiers is an old RTS (a DOS game) without any kind of fog of war. The terrain is divided in areas (4x4 areas on the large maps), capturing an area also captures all the factories in the area. As I''ve said, to have game you need a combination of certainty and uncertainty. If there is too much uncertainty, you have nothing to base decisions on. If there is too much certainty, there is no game anymore. Obviously, removing the fog of war removes a lot of uncertainty. Z managed to make up for this uncertainty loss in more ways. All the factories were pre-built, and more factories couldn''t be built, only captured (each area has a factory or two). Each factory could build units in a certain time frame. This time decreases with the number of areas the player controlls, so controlling more areas was extremely important. Capturing a factory one second before the production time is up means the unit is produced for the other player. The combat model which is fairly random adds a lot of uncertainty as well. All these possible twists keep the game in the "everything can happen" area.

Another great thing about Z is the way the end-game is handled. Many games lose uncertainty at some point, yet the game keeps on going on and on until the final victory is achieved. As an example, in Dune II the uncertainty ended soon after the game start (if you can survive the initial onslaught, you most likely win eventually). Actually achieving the win takes a lot of pain. In Z, at the point where the uncertainty is lost (typically after a larger fight), you have a player owning twice as many land areas and twice as many units as the other. The first player already produces four times as many units, and this difference increases a lot more in a very short time. The end-game in Z is almost a triumph march. Short and glorious as it should be.

So what has losing the fog of war done for Z? Instant reaction to enemy actions and a lot more planning ahead.

Returning to categorising visibility systems. I think the micro-management involved is often the most important gameplay element in a visibility system (it directly determines how much is known). I hate when games force me to handle all the scouting by hand. It''s one of those annoying things that turn the RTS genre into action arcades. Here''s where Total Annihilation really shines. The share brilliance of guarding a combat unit with a scout plane to have the scout circle around the combat unit and follow it around the battle field. Or the patrol routes. Or the perfect radar/sonar system (with static radars as well as dynamic radar).

The cost of scouting units, their strength and speed are other significant factors that can make the difference between low and high visibility games.