One of the things that constantly bothers me about RTS games is the way in which air units are implemented. Typically, air units hover around the map for ridiculous amounts of time, stop and turn on a penny to engage their opponents, and generally behave exactly like ground units, except they ignore terrain. In order to balance out their boosted mobility, they are usually less powerful than their ground based counterparts. Real life is nothing like this. Aircraft do not hover around the battlefield indefinitely, taking potshots at anything that comes near. Nor are they weak compared to ground units. Compare say, a B52 to the most powerful ground based artillery you can think of - there really is no comparison in terms of raw destructive power. So, the question is: how do you implement air units in a realistic manner, without completely breaking the game balance? Think of this thread as a brainstorming session. A few ideas of my own to kick things off. 1. Air units are not directly controllable, and they do not float around doing nothing when idle. Instead, you control them indirectly (perhaps through an airbase building) by assigning missions. Missions would include: recon, air superiority, ground attack, heavy bombing etc. Or perhaps something like RA, where air units can be directly controlled, but need to go back to their base to rearm, or when idle. 2. Air units only do what their mission requires of them: eg, if you set off a bombing mission at a certain point, that''s exactly what they bomb, nothing more, nothing less. After completing the mission, the units return to the airbase. 3. Air units generally go pretty fast - there''s no dithering about on the battlefield, and no turning on a penny/hovering in front of your target. If a target isn''t killed on the first pass, you have to circle around to try again. In the meantime, the plane can still be attracting AA fire from anything that can shoot at it. 4. Air units are expensive, and so are the weapons bolted onto them. In addition to any cost of building the air units, there should be a cost per mission, representing the fuel and weapon usage. A single blanket bombing or precision bombing mission could be very expensive, maybe comparable to the cost of building a light tank.

Air units were one of the not-so-numerous things that Empire Earth did right. They moved realistically, were always moving, and could only stay in the air for a short time. IIRC, they circled the spot you last ordered them to when idling and would automatically head to the airstrip when low on fuel.

I have yet to implement air units in an RTS, although I have ideas for one in 3D. They will have unlimited fuel and regenerating ammo, and will have the basic guns, bombs and air-to-air missiles.

To counter them, other air units or missile towers, I imagine, would work nicely. The air units could be intelligent enough to evade them and even work together to dispatch threats (one draws fire while the other takes it out).

This kind of setup demands quantity of units as the deciding factor. More air units or missile towers would be needed to take out large groups of air units.

Then there is, of course, the use of anti-aircraft emplacements.

I figure the air units could hover, cruise slowly, or go to dogfight speed.

Again, I have yet to see them in action or how it would break the game.

One real challenge is to get them to navigate terrain at a low altitude without revealing themselves, or crashing while they are trying to take out ground targets.

[edited by - Waverider on February 3, 2003 7:04:34 PM]

[edited by - Waverider on February 3, 2003 7:05:03 PM]

"So, the question is: how do you implement air units in a realistic manner, without completely breaking the game balance?"

My question is: why bother? Designers often think of what is "realistic" first and then figure out how to shoehorn mechanics into that, and the result is often neither realistic nor playable. Turn-based strategy games already have so many non-player-deterministic elements, what purpose does realism even serve unless the goal of the designer is to create an accurate simulation of an actual conflict?

I agree that it''s a stimulating mental exercise to figure out better implementations for air units, but don''t let realism hijack the game design process.

This is one of those things few games do right. I found one interesting way to solve the problem. In Total Annihilation (the robot one not the fantasy Kingdoms version) by Cavedog they set it up so that planes (if not over water) would land on the gound and not fire but could be shot at by ground units. Sea-planes were also allowed and could land (under)water. It was interesting because they had weak armor but unlimited fuel, in that respect they were balanced.

Christian

quote:

---

Original post by oneirotekt
I agree that it's a stimulating mental exercise to figure out better implementations for air units, but don't let realism hijack the game design process.

---

I agree 100% - playability should always come first. However, I find that most RTS games are somewhat abstract, and that by removing some of this abstraction (making it more 'realistic' if you like) you can create a different kind of strategy game.

In any case, please consider this thread an 'interesting mental excercise'



[edited by - Sandman on February 3, 2003 7:48:55 PM]

I''d take a look at it from a standpoint of unit design.

I think we tend to see things as a "whole" rather than through its parts. When we say, "bomber", we immediately conjure up huge flying aircraft that are slow, ponderous, but with huge and deadly payloads. But then, what about Tactical Bombers like the FB-111? These are almost as fast as some jet superiorty fighters, can attack air targets, and may not have as huge a payload as strategic bombers, but are often equipped with more accurate delivery systems.

So how do you design this? The more I look at design considerations, the less I see inheritace as the main method of creating unit types. How do you create the Tactical bomber? Do you give it multiple inheritance of both the bomber class and the fighter class? But really, it''s not quite the same as either one, since it''s not quite as powerful as a strategic bomber, nor quite as agile or fast as a fighter.

Instead I think we have to see unit design as a composition of modules. This defines their characterstics better I think, and also allows for more precise function capabilities. Also, by using composition rather than inheritance, it is possible to create new unit "types" easier (I would think, but more experienced programmers please feel free to shoot me down on that assumption) than by introducing new class definitions in some sort of shared library file (dll or so).

How does this work out in game terms? Strategic bombers by their defining characteristics are slow, and virtually defenseless against air-to-air or ground based SAM''s and AA guns. The B-52''s were designed to be mostly invulnerable against ground based targets due to their extreme altitude, but recent missle technology suggests that they are no longer immune to ground based defenses. But even if thius is untrue, they are still extremely vulnerable to air defenses. Also, remember that bombers are slow. And unlike the limited visibility systems for ground based warfare, radar and satellite technology will detect the launching of aircraft from anywhere in the world. In other words, your opponent will KNOW that you are launching an attack when you set your bombers to flight.

For tactical based aircraft, they too are vulnerable but not to the same degree as bombers. Their main advantage is flying NOE (nap of the earth) to limit line of sight and target acquistion times. However, they can still be vulnerable here. That''s why when we hit Iraq in 1991, our first targets were SAM sites, AA guns, Radars and Communications arrays with our B2 Stealth bombers, F-117 stealth fighters and F4 Wild Weasel Phantoms.

There is another method of attack though for ground based attack fighters...slow and methodical. Aircraft like the A-10 are very slow (their stall speed is a little less than 200mph) and can take their time taking aim and blasting away. Since they are built like flying tanks, they can afford to take some AA gun fire (though are still vulnerable to SAM''s). But being able to fly to slow and low means they have a very good chance of evading missles. The same tactic applies for VTOL craft like helicopter gunships and the Harrier attack craft. They fly low hugging the ground evading radar and using the terrain to shield them from AA guns. They then pop up over the terrain to take out targets and duck back down.

I think you''d have to code in flight rules (movement rules) based on the capabilities of the mobility type of their engine (this is how I''m coding it in my game). In my game, all vehicles have a MobilityModule. This is a class object which defines what kind of locomotion the vehicle uses (also, there is nothing saying that a vehicle can''t have two MobilityModules...say to create an Amphibious vehicle...something that would require tricky Multiple Inheritance in pure Inheritance based systems or require more duplicated coding). Based on the class, movement rules and terrain modifiers are extracted. So in the case of Tactical Bombers, they can fly in extremely fast at low (but not NOE) altitudes. However, the faster they go, the tighter their turning radius (and this is something sorely lacking in most games...the lack of turning radii). But conversely, the faster they go, the less time they are vulnerable to enemy fire, and the harder they are to hit. Strategic bombers might be immune from ground based defenses, but will need to have either air escorts, or make sure that they have air supremacy in the area they are assigned to attack.

oneirotekt-
The reason why making games more "realistic" is a design goal of some is that it allows for more interesting game choices. I also tend to feel that "fun" is a very subjective term. For many, going to a Britney Spears concert is "fun" while for others, watching Carmen is their cup of tea.

I read the little link that you had in your post, and I didn''t agree with it on a couple of points. I really think the game industry has to start rigidly defining what "games" are. Afterall, the ultimate goal of a "game" is to make us feel better. This can be through having fun, through telling a story, through giving us a chance to win, to explore new worlds, or to just put our minds at rest with some lighthearted muscle twitching. I feel that the authors viewpoints refelct a very narrow interpretation of what a "game" can and should be.

To get back on topic though, I think that all units should also have maintenance AND operating costs. YI think you''re right Sandman in that bomb runs are incredibly expensive in terms of fuel, munitions and upkeep. Also, by creating movement rules, you can factor in the length of time it takes to recall air units to make another bomb or strafing run. This also begs the point that units should not be able to fight ad infinitum. Fuel and ammo limits will reflect that. But on the other hand, air attacks should be devestating. Basically, if you don''t destroy aircraft, they are going to make mincemeat of your ground forces. The trade off of aircraft destructive power is their own vulnerabilities, and that while their attacks are devestating, it takes a long time to refuel, re-arm or return to make additional attacks.

i pretty much agree with everybody. air units should be devastating, expensive, and something that appear only in certain circumstances. they are not there floating all the time waiting for orders and deal an equal amount of damage of a marine.

i think this should apply to mechanical ground forces such as tanks. they should be expensive, powerful, devastating, but slow as hell and 0 point maneuverability (hard to move around). unlike in C&C:RA, tanks cost slightly more expensive than human ground forces, but they are a heck a lot more reliable and durable.

the thing about realistic, i think it''s better for us to make a game that is realistic. so far, we have been using our fantasy and imagination in a game, creating our own system that we think balanced. and that affects us on how we perceive what games should be, what games should look like. if we put realism in a game good enough, the gameplay will be different. it''s not imbalance, but...different.

---

return 0;

You know, games CAN have unnecessary ties with reality that add nothing to gameplay. They CAN give the player little or no choice in the extrapolation of the story. They CAN rely on unchecked, unjustified conventions.

That''s fine. Viva la difference.

I just don''t see the point in driving the medium into the ground by not *trying* to reassess these painful conventions, which only still sell due to the general public''s lack of better taste and knowledge. The above are not good things at all. To argue in their defense is to sound the deathcry of a frustrated movie director.