I've been trying to come up with new ways for the player to control units in a RTS game. I was thinking of, rather than order individual or groups of units the player mainly assigns units or groups objectives, which they inturn decide how to fulfill. The units will attempt to fulfill their assigned objectives, and request more resources or units if these are required to meet the objectives. The player would still be able to micromanage the units, how they form battle lines, or what units are posted and where, but they wouldn't have to. Would this be practical/possible to do on a large scale do you think? Would it be fun to play, as many of the decisions ussually undertaken by the player would be largely taken out of their hands? I've also been trying to think about new ways the enviroment and even resources could be used as a weapon/tactic. For instance if you were to make it possible to dam rivers, and water was a required resource, players could cut of each others supply, or destroy dams to remove reservoirs or hydro-electric plants. Resources could also be sabotaged or poisoned, dumping toxic chemicals upstream would kill units using the river for water supply down stream. Sabotaging replacement parts/ammunition could damage/destroy equipment using them. Having a deformable enviroment, being able to dig trenches, build up natural earht walls, underground traps and pits could also add an extra dimension to game play. Any ideas? what do you think, do able, fun?

The way you wrote it, it sounds a little too complex. It might very well be fun, but it sounds like it's have a huge learning curve.

Some good ideas however.

You could make food be a vital resource. Then DAmminf the river or poisioning it would greatly reduce/kill the amount of fish that could be harvested out of the river, cutting off their supply of food...

Also, as far as the deformable terrain, you could make it so that artillery and other powerful weapons deform the terrain when they hit it, but only to a certain extent--you wouldn't want a player destroying a mountain range so they could sneak into their enemy's base. Also, if you were to do the deformable terrain option, I would suggest giving the players an ability to re-level the terrain, like paving it or something. Otherwise, players would robably scout out potential base areas and blow the crap out of it, amking it impossible to build on.

You mean actually putting the s in the rts? [grin]

Ok, this is a bit of a long post. Sorry about that.

I've raved of an indirect control system before. Instead of just giving a pile of links to my older posts (as I usually do), I'll summarize the ideas here. I know it's a bit of a repost, so I'll apologize in advance to anyone who gets the "been there, done that" feeling.

This is the basic premise on which I base the supremacy of an indirect control system in an rts:

Quote:

From The Art of War by Sun Tzu When the general is weak and without authority; when his orders are not clear and distinct; when there are no fixes duties assigned to officers and men, and the ranks are formed in a slovenly haphazard manner, the result is utter disorganization.

If you think about it a while, it is not the general's job to give the orders to the individual units. The general gives the orders to his subordinates, who give the orders to their subordinates, and so on, ad nauseaum, until finally the orders are given to the grunts. Of course, in a computer game, the computer takes the part of the subordinates. This would rise the level of abstraction in controlling the units and could allow you to make your plans more detailed with less hassle.

The utter disorganization is evident in games using the more direct control. If I command my workers to chop wood, the command is interpreted as "chop some wood, but start with this tree here." Not clear nor distinct at all, at least with regard to what I really want to have done (say, I want to get one particular tree out of the way). After a while the general area of the original tree runs out of trees and the workers assigned to that particular area will start to wander about in order to find more trees. They will go to enemy area to chop wood, even if that meant they would die. They will chop down the trees you wanted to keep for cover. If they do not find trees in the immediate vicinity, they will start to stand about, doing nothing. This is utter disorganization within this context.

In order to prevent this from happening, the player has to baby-sit his workers through the process. This is painful and not fun. Instead, it would be better to make it possible to give clear and distinct orders.

Indeed, it would be better to declare jobs with priorities. You could set an area on the map and declare it a wood-chopping area with priorities such as "at least 2 villagers much be here". Also, you could have other parameters as well, such as aggressiveness to indicate the reason why you want to chop down the wood (maximum aggressiveness if you just want to get rid of the forest even if this means a waste of resources, minimum if you want to get the most wood out of it etc). After all, it does not matter who does the job as long as someone does it.

In order to control you army you could declare areas that the soldiers must clear of enemies. You could also have a trigger-event-system in which you could make more thorough plans such as "when group 64 has reacher area 23 and group 89 has reached area 83, then group 64 will proceed to area 92 and group 89 will proceed to area 76". Written as text it might sound cumbersome, but you could build a neat Tactical map -mode in the game in which you could make plans such as that on the world map (Empire Earth 2 seems to have a map on which you can plan your attacks and send the plans to your allies, but I don't know if you can actually control your units this way; I've only given a brief look at the demo version).

All of this doesn't necessarily mean you shouldn't be able to control individual units on a micromanagement level any more. Automation is good only as long as it doesn't take away the power to do things manually. Thus you wouldn't really be taking the power out of the player's hands.

A system like this is used to control the imps in Dungeon Keeper. It's great as you all you need to do in order to get something done is to tell what it is, and the imps will find the job and start doing it. After they finish the job they'll find a new one. No more do your workers stand idle unless there really is nothing to be done. It works well enough for me.

Another game with a similar system is Majesty. Here, you can't really control the heroes that keep wandering about. You can only influence them by telling what you want to be done and how much are you willing to pay for it.

So, to summarize all this: allow the player to delegate the boring chore of assigning the workers to specific jobs to the computer, yet allow him to take over whenever things don't work out. Raise the level of abstraction in control.

References (not that I actually refer to them specifically in the text...):
Current RTS models
Resource Shuttling in RTS Games
Those distant, storm-beaten ships...

And here's a link to a whole thread about a control system very much like this one:
RTS Interface: Strategic Markers

A couple of years ago, Dauntless raised similar concerns, but rather than looking at assigning tasks and letting units decide who does them, we concentrated on letting players create/adjust the in-game Table Of Organisation - the flow of authority from the general down through his subordinates - each unit is either a container for subunits or an atomic (indivisible) unit like a single tank - if a container, one subunit is the primary decision-maker for the unit. Orders can be put in at any level of the command structure, and will get interpreted "by the book" into orders for component subunits as necessary (refinement: individual sub-commanders learn from experience). "The book" should be customisable by the player, allowing them to tweak AI responses - or at the very least accessible. The orders that can be given to a unit are generally high-level concepts like "attack", "defend", "build", "explore", which naturally convert into relatively complex micromanaged behaviour through the hierarchical translation by successive sub-commanders. So an order to a mixed unit of missile and melee troops to attack may be translated to an order to the missile unit to attack, and to the melee unit to defend the missile unit - which the missile unit may interpret as move to high ground then start firing, while the melee unit forms a screen for the missile unit, and may passively intercept melee attackers or aggressively attack ranged threats...

Of course, creating "the book" in the first place represents quite a lot of man-hours (though you may well be able to save effort by consulting Force Manuals from appropriate armed forces...

Grim - You've pretty much listed the idea i was trying to get at - although in a far more understandable way - thanks :).

rmsgrey - Those are some really good ideas, its a bit more complex than my original thoughts, but i'm certainly interested, thanks.

Brokenimage - I see what you mean by steep learning curve, but if good tutorials and possibly helpers are used i'm not sure it would be such a big problem.

thanks for the replies they given me some ideas and a lot to think about.

This somewhat reminds me of Silent Hunter 3, more of a Tactical warfare game, but the micromanagement and decision making in it seems relevant.

The player sets courses and using information recieved from the sonar-man, plots the tragectory and speed of the target in order to properly calibrate the torpedo, along with other various mechanics of running the ship. The tasks can be complex, but the player has the option of delegating them to one of several "Bridge officiers" who do it for him (their qualifications/experience determine how well they do it, this applies to the "underlings" as well), or the player can choose to do it himself, if he knows how. Generally they automatically take whoever is available to do the particular task you set, but you can move whoevers more qualified or available manually if you want to.

When your Submarine gets damaged and your crew is exhausted or injured, then things start getting interesting since you may not have enough crew to stop the flooding and man the torpedo room effectively, or one of your bridge officers could be dead so you have to manually set an evasive course while trying to figure out where the enemy destroyer is, and end up manually moving people from the engine room to damage control.