Multiplayer games can be the most enjoyable games to play because the challenge comes from human intelligence as opposed to the often-predictable AI present in competitive single player games. It's this element that keep some of the best video games alive well after their technological novelty has worn off.

The following article discusses how to design a game such that immediate conflicts are richer, more fulfilling, and encourage experimentation and learning. This article is highly theoretical, and will present ideas "in a vacuum". Some sections include the phrase "Where we end up.” These sections discuss the strategy realized by players who spend a considerable amount of time playing a game which has implemented ideas presented in this article. There are a few games which may come close to what I describe, but none are perfect implementations.

My goal is not to claim that certain games are bad, nor that this article outlines the only way to make a good game. Rather, this is a theory which tends to apply to most competitive multiplayer games and can help guide you as a game designer when creating your own competitive multiplayer game.

Assumptions

Before I begin discussing theory, I'd like to point that these ideas do not apply to all games. Here are some elements which must be present when applying the ideas I discuss below:

Competitive Multiplayer

Obviously, the games must be competitive multiplayer. For team-based games, each team must have at least one human-controlled player because artificial intelligence is not able to play the game to the depth indicated in this article.

To reach the full depth of game play, an player must be able to predict opponent behavior, develop behavioral patterns, dynamically recognize behavioral patterns, intentionally trick, and make mistakes. This kind of behavior is very difficult to develop with an AI opponent.

Fast Paced Play

The game must have fast paced play. Turn-based play will eliminate some of the depth of play when implementing these ideas. Interestingly enough, many real time strategy (RTS) games attempt to implement this method. Due to the popularity of these games, and the widespread implementation of the soon-to-be-described method, I will also discuss why this method fails for RTS's.

Variety of Attacks

This is usually not an issue, as most games implement a variety of attacks simply to keep the game interesting. That is, multiplayer competitive games in which the player can only attack one way are very rare.

Keep in mind that a variety of attacks doesn't necessarily refer only to a variety of weapons, but it can also refer to positioning, rate of attacks, and recovery.

Rock Paper Scissors - The Origin

The basis behind this method is the simple Rock Paper Scissors (RPS) game. If you are not familiar with the game, check this article.

The core idea behind the game is that no one strategy can win every time. Regardless of which "attack" you pick, you can either win or lose (or tie in the case of both players picking the same attack). Many game designs follow this pattern. This keeps players from finding one specific strategy which will win every time. It encourages players to play dynamically.

RPS will be used as an analogy for game play; however, my discussion will not be limited to the simple implementation of RPS game play. I will also consider how this can be modified to allow for more interesting play, and even how RPS is at the core of almost all games we play including sports like soccer, basketball, and American football.

Unmodified Rock Paper Scissors - What It Gives Us

Without any modifications, RPS already provides some interesting game play and can actually result in player strategies.

Attack Variety

In an RPS system, a player must vary his attacks. If the player repeatedly picks rock as his attack, then the opponent will catch on and continually pick paper. The attacker will either lose interest in the game or realize that he must pick a different attack so that he doesn't lose every round.

Blizzard Games' Starcraft

Most strategies in competitive games provide “tradeoffs to the player”. StarCraft implemented the RPS system by giving each unit strengths and weaknesses. Players can defend against ground units by building defensive buildings like bunkers in the entrances to their base. However, attackers can overcome this strategy by attacking with flying units. Units such as Goliaths are effective against air units so they can help strengthen a base's defenses. Relying on just one kind of attack is not an effective strategy.

Prediction

As indicated previously, prediction and "reading your opponent" are valuable strategies in RPS games. To predict, one must either be able to recognize a pattern or have some kind of indication of what the opponent will do.

This is the most valuable strategy when playing some games. When playing poker with experienced players everyone will know the rules, the value of their hand, and probabilities of winning given a particular hand. If all players have a "perfect understanding" of the rules and the cards are truly random then in the long run no player has any advantage over the other. At this point the game is not as much about making wise decisions with your cards as it is about reading your opponents and being able to bluff. The game becomes less about the cards and more about the players - and being able to predict what your player has planned will make you a better player.

A common attack in Street Fighter II is to attack by performing a strong kick in the air followed by a sweep (duck and strong kick). This is effective against beginners because it requires the defender to block the first attack while standing and the second while ducking. However, this common attack pattern is easily recognizable and players familiar with the game quickly learn to defend against it.