I think it''s all about the controls: instantly executed, very basic orders, requiring good, fast, action-like dexterity for success. Real-time simulations with turn-based order giving (like the great Laser Squad Nemesis - try the demo!) play just like classical turn based games. Two directions to evolve RTS simulations: 1) Sophisticated but time restricted controls - the realistic wargame direction. The orders get executed with a significant delay, or only at certain time intervals but contain more information, allowing the unit to carry out complex actions without too much baby-sitting. Play LSN to see what can be achieved here. 2) Unrestricted sophisticated controls - allow the player to give many more orders with ease. More powerful controls by the player will change by themselves the way RTS games are played, allowing for more sophisticated plans and strategies. The game design of the simulation behind the game will have to change too to take advantage of a player in complete control of the entire battlefield rather than a micromanaging player having to task-switch between various fronts.

''real time strategy''

It is a strategy game in which orders given by the player are executed in ''realtime'', along with all other game actions (ai movement, automated player units, etc). The AI players (or other human players in the case of multiplay) will not wait for the player to issue orders before issuing thier own. Typically this will force the player to focus on several things at once, ie. base-building/maintenence, unit production, defence, offence.

To aid in this, players are given such helpful tools as radar, allowing them a quick overview of the situation of the whole map (or that which is ''unshrouded'' if a shroud (aka, fog-of-war) is in use), but at very low detail, squad-numbers, allowing the rapid selection and command of pre-selected groups of units, waypoints, allowing predefined patrol/attack-routes to be set up.

Tactical elements are usually "sweetened" with broader-scale concerns - unit capabilities, for example. Fixed defenses and "bases" again up the ante, since the player may have relatively poorer coordination than their opponent but a counterbalance with a better sense of strategic play.

This consideration leads to other avenues of exploration:

1) unit evolution - the "player-made" direction; basically the holy grail here is to allow the user a realistic full-featured physics with a very simple or even automatic research facility at which new variants can be developed.

2) Castle play - the player is encouraged to place a great deal of time and effort into their defensive ramparts. The threat is from turtling, so weapons must be available specifically for destroying and sieging purely defensive player.

ld

I'm not sure that I'd agree that the controls define the RTS genre as such. Many games have instantly executed, basic orders requiring dexterity to succeed and are not considered RTS games; conversely, some games do not have such control systems and are considered RTS.

I think I'd agree with Kazgoroth that the no-brainer definition seems most appropriate - it's a strategy game that takes place in realtime. Of course, 'strategy' and 'real-time' are themselves subject to some interpretation. I'm not going to bother defining strategy, but it might be interesting to think about the definition of 'real time' and its consequences for a strategy game.

Real-Time implies event driven behaviour for the game tokens, as opposed to the explicit behaviour of tokens in a turn based game. In the event driven model, a token does nothing (or performs some default behaviour) unless an event (such as being given an order) causes it to do something else; in the turn based model, the token typically does not have a 'default' behaviour - it's behaviour is explicitly defined by the controlling player each turn, and the turn cannot progress until its behaviour has been defined.

A direct consequence of the real-time model is great need for global awareness of the player's tokens. In a turn based model, a token will remind a player of it's predicament each turn, enabling that player to respond to carefully to any special circumstances. In the real-time model, it is up to the player to keep an eye on each his tokens - failure to do so means that if any special circumstances do arise, he is unable to respond to them at all, and must rely on the default behaviour being good enough. The turn based player can concentrate on one token at a time - the real time player needs to be aware of all his tokens at once.

Of course, this kind of global awareness is actually very difficult for most people - we tend to concentrate on one thing at a time. In fact, this task is so tricky that the finer points of the 'strategy' side of the game often get lost as we try and cope with the difficulties of keeping track of everything in the real-time model. I think one way in which the gameplay of RTS games could be improved would be to add tools which enable the player to improve his awareness of the battlefield.


[edited by - Sandman on May 17, 2004 10:51:10 AM]

Something that I would add to the definition would be that "the player controls multiple units" (for some value of ''controls'' and ''units''). If there is only one ''unit'', the game seems to be called something else (such as FPS, RPG, Action-Adventure, whatever).

I''ve not heard of an RTS with only one ''unit'', though many have one main ''unit'' that is most important.

quote:

---

Real-Time implies event driven behaviour for the game tokens, as opposed to the explicit behaviour of tokens in a turn based game. In the event driven model, a token does nothing (or performs some default behaviour) unless an event (such as being given an order) causes it to do something else; in the turn based model, the token typically does not have a ''default'' behaviour - it''s behaviour is explicitly defined by the controlling player each turn, and the turn cannot progress until its behaviour has been defined.

A direct consequence of the real-time model is great need for global awareness of the player''s tokens. In a turn based model, a token will remind a player of it''s predicament each turn, enabling that player to respond to carefully to any special circumstances. In the real-time model, it is up to the player to keep an eye on each his tokens - failure to do so means that if any special circumstances do arise, he is unable to respond to them at all, and must rely on the default behaviour being good enough. The turn based player can concentrate on one token at a time - the real time player needs to be aware of all his tokens at once.

Of course, this kind of global awareness is actually very difficult for most people - we tend to concentrate on one thing at a time. In fact, this task is so tricky that the finer points of the ''strategy'' side of the game often get lost as we try and cope with the difficulties of keeping track of everything in the real-time model. I think one way in which the gameplay of RTS games could be improved would be to add tools which enable the player to improve his awareness of the battlefield.

---

Hrm, yes, I''d definately agree that tools to improve situational awareness would be an aide to this type of game.

Also, I noticed you mentioned that unless the player jumps in to take control of a particular situation, the ''default behavior'' must be relied on until such time as a player does give his/her attention to the situation. Originally this default behavior (in the case of an enemy attack, there are of course other situations to which different responses would occur, but I wont look at them here) was to return fire to any enemy units that attacked the unit in question. Since then, this has been improved, and often multiple settings are given; often units will by default attack and then pursue any enemy units that come into range, but can also be set to a ''guard mode'', in which they will still fire on enemy units, but wont persue fleeing units. Some other common default behavior selections include ''patrol,'' where the unit moves back and forth between 2 (or more in the case of waypoints) points and engages any enemies that come into range, or ''scout'', in which units will automatically move around the map exploring, but will attempt to evade enemy units without engaging. There are of course more, but you get the idea of what I''m talking about.

Perhaps another way to enhance strategic gameplay would involve allowing players more control of these default behavior settings. Such a system was used in Dark Reign, where various behaviors could be obtained from units by setting some simple variables, which from memory I believe were independance, damage tolerance, and... something else.