Client-Server and Peer-to-Peer Architectures in Multiplayer Games

When designing a game with an online multiplayer aspect, game developers must choose a network architecture that works best for the game’s intended purpose. The network architecture can either be client-server based or peer-to-peer based. But how do they decide which network architecture to use? Developers must understand the benefits and issues each network architecture brings before deciding which one to implement. Client-server architectures are the most common type implemented in multiplayer games, especially in massive multiplayer online games (MMOGs). There are a multitude of resources to help developers implement the architecture and it provides a more profitable business model compared to peer-to-peer architectures. However, they are a susceptible to network failure and bottlenecks on the server-side, which can cause poor playability for users. Client-server architectures are also more expensive to implement and maintain. On the other hand, peer-to-peer architectures, although less common than client-server architectures, are often implemented in multiplayer games that do not require many players to be connected at once. Thus, they are often implemented in fighting games. Peer-to-peer architecture also provides a lower level of latency. However, their networks are not as secured or as consistent as client-server architectures. And because they are less common than client-server architectures, there are less resources available to help developers implement the architecture within their games. In this presentation, I will cover both architectures and discuss benefits and issues associated with them.