What do people think of the current move toward RTS games having physics simulations? I've heard that the next game in the Age of series is going to feature real-time physics courtesy of Havox. Personally i think that a certain degree of physics is a good idea, so that damage from a projectile can vary depending on its range weight etc, and some great effects with troops being sent flying by explosions, catapults and the like, but i'm not sure that a full physics simulation to the level of havox, effects like the ones seen in Half-Life 2 are necessary. For instance the AI for a single combine soldier in half-life is far superior and more complex than the AI for a single Archer or Swordsman in an RTS game. The way i see it this is for two reasons, because of an RTS games scope the individual units do not need to show complex behaviour inorder to show believable behaviour, and because of the sheer number of units you can't afford complex 'thought processes'. Can't the same be said for physics calculations? To produce a believable simulation they don't need to be as complex realistic or accurate as they do for a FPS. What do you think?

Since I know that the development time will detract from making better AI, a better skirmish mode (which is all I really ever play), more weapons, more options, and more strategies all for the sake of making what already looks good enough better-- I'm against it.

EDIT: Not to mention whatever egregious licensing prices they probably have to pay for Havok, which drives the cost of the game up and forces a more water-down, mass market appeal approach...

yeah i see your point, but at the same time, some level of physics, although i'm an advocate of simplified physics, not close to real-world, could give the player extra and unconventional play tactics, for instance causing a landslide to wipe out an enemy town or army rather than fight it out in a straight fight.

In my opinion, unit AI is much more important(and generally less resource-consuming) than physics in RTS games. Why, you ask?

Take the 'infantry problem', for instance. Most RTSs include infantry, but some value them more than others. See, in many games trying to be 'realistic' ala Sudden Strike or Blitzkrieg, infantry gets mowed down by pretty much everything and it's only respite is an AI that drops to the ground when shelled and the ability to put it in buildings.

Of course, the first and easiest solution to make 'realistic' is the horde solution, where infantry come out fast and cheap and charge like the stereotypical Soviet Army with little regard for their own safety. These sorts of games go to the all 'tank/vehicle/robot..etc' pretty quickly.

The second solution is the 'terminator' solution, where infantrymen are strong and can take many hits from tanks, rifles, etc. and still be alive and well. You see this in lots of RTSs where they want to make every unit useful and use the easy-to-balance rock-paper-scissors system.

There's the main problem with your average RTS. A man on the ground will not simply stand out there and get shot at or just drop to the ground and sit there in the open firing back but in RTSs that's exactly how combat is: Two units sitting in a field firing at each other or dancing around like in a UT deathmatch.

The only game i've seen that's remotely adequate in the unit AI field is the Close Combat series, and it's got a few kinks with it, too. Too bad they won't be making more of 'em.

Well, i'm off on a tangent.

That would be interesting. AI that reacts without the players command. Aka if the player is invaded by tanks, the villager units flee...

Thats largely my point, many modern RTS games that are just coming out are going down the cram in as many cool features (such as havox driven physics) as possible.

I can't remember playing an RTS where the AI was actually satisfying, whether it be for the single unit or the for the entire civilisation AI.

I think physics will be (in a cut down form) as big boost to RTS games, but other things, such as AI are more important. Unfortunatly it appears that what sells games are graphics and physics and other essentially 'eye candy' effects.

one game that is absolutely outstanding when it comes to realistic physics, but not really is an RTS, is Soldiers: Heroes of world war 2.
the engine of this game is really astoundingly impressive, alas the game as a whole gets rated down for the quite mediocre campaign missions does not really utilize its potential.
the physics effects in this game are really remarkable, especially when it comes to vehicles, and how infantry utilizes the environment for cover. one does notice it in the very steep CPU requirements for it to run smoothly though.

also, this game makes infantry quite dangerous to vehicles, have they the right equipment, but without tactics they will be killed relentlessly. (the way the AI does.)
it really should be seen. it is not really an RTS though, as one does only control 3-6 men at most, which can be eighter at foot or in a vehicle/emplacement of some sort. also, one does usually play with "direct control", and at least i pause the game frequently to plan ones tactics.
one description that i find quite fitting is "Operation Flashpoint in third person RTS view with a WW2 mod".

oh well, hope i dont disturbed anything, i am mostly a lurker here.
this game is not all that excellent, but it certainly has impressive physics. in a kind of RTS style world. still, it would probably not be applicable to the hundreds of units of ordinary RTS:s.

Quote:

Original post by CheeseWeaver the physics effects in this game are really remarkable, especially when it comes to vehicles, and how infantry utilizes the environment for cover.

Do you mean the AI? I think this conversation shifted a bit. It would be nice to see AI take advantage of the physics of the new games coming out. The gravity gun sounds like a ton of fun in half life 2 for being able to do things like picking up razors and throwing them at enemies, as well as pushing grenades back at enemies. However, does the AI take advantage of the physics?

Physics could be an extremely interesting feature in an RTS, but only if the gameplay takes advantage of it.

Specifically, I'm thinking of buildings. Say that instead of being given a bunch of premade buildings that you just 'drag and drop' onto the landscape, you've got equipment/room types (a la Dungeon Keeper) and "construction blocks." So you actually play architect in creating your base buildings - can build a single castle-like structure, a citadel, or just a load of small buildings.

If players are doing that (and the AI, too) then physics becomes interesting because you can start looking for weak points in building architecture. Enemy built a big wall around their base? Instead of just assigning a bunch of units to 'attack' a wall (which isn't exactly realistic), you can look for the places where there's a small cave running under the wall, or a stream or something, and plant explosives (a la Helm's Deep). If you're trying to destroy an entire base, you could take things out a bit at a time or you could attack the foundations of the building, causing the whole thing to topple.

RTS doesn't just mean something with units and stuff, of course. Bontago is a physics-based RTS that plays extremely well...

I think another great feature RTS games could take advantage of are physics damage. If a trebuchet fires and hits a tower, how awesome would it be for the brick and morter to crumble to the ground crushing players caught under it?

Not wanting to side track, anoter great RTS feature would be to make the world more dynamic. I can see this split up into 3 features:
-Nature. Seasons change during an RTS game. This brings in a whole new element of strategy when you must gather the crops before winter. It also raises the issue of when does one want to attack? It would be suicide to attack in the winter, where food would be sparse, and the fighters could suffer from the cold. That is unless you were the Vikings - Introduces a whole new slew of climate benfits.
-Building Improvement/Detoration. A problem with RTS games is the empire you build does not feel real. To static again. Making buildings should require upkeep. This shouldnt require the players attention, but to set funds to keep buildings up to shape. In a nod to a technology tree as the player makes discoveries and if the buildings are in the proper condition, they should be upgraded as a visual enhancement. I would love to see cities neglected and start to fall apart, while other well crafted ones gleam with polished marbel from the resources put into upkeep and advancement.
-People changes. In the 4 page rpg roleplay thread we are discussing how to make characters feel more alive. The first step in an RTS game is to make the people more than objects. ;). A great idea is people's will. This could include such possiblitys as the peoples philosphies - (and oh how cool that would be. Imagine in a game like Rise of nations where a country where people oppinions go from feudalism to democracy (key players like Locke show up - NPC private property :) ) and beyond. Other needs can be more basic but importaint like the people are getting feed, etc.

I think the reason why games like The Sims are so popular is they give the player the option of building a true strategy. While sandbox mode is popular, A game that requires a strong strategy to sucess could be immensly popular, at least to the hard-core RTS auidance. I am not a fan of the progression towards this new system of only one or two resources and the emphasis on building tons of troops. I guess Caeser just has a place in my heart.