What I want to discuss is the future of multiplayer adventure games. I've been experimenting with the TCP/IP plugin for AGS with a few others in the #ags chat room, and we've found it works quite well. It has quite a few nice features, yet it's kept very simplistic in its functionality and implementation.
I recognize that some people will disagree with me about this, but I personally don't like the idea of making a multiplayer epic quest. Regardless of the game's design, you would probably be dependent on the other player(s) throughout the entire game. Each player would have to stay caught up with the others. You couldn't advance through the storyline without the other players being on the same wavelength. This degree of dependency is not a desirable feature, in my opinion. Instead, multiplayer adventure games should take a different angle if they're going to be recognized as an enjoyable part in the future of gaming.
Any multiplayer game needs either (1) teamwork and/or (2) competition. In some of the most widely-played multiplayer games, both teamwork and competition are incorporated. The players understand that by working together they can "beat" the game faster or have a better chance of doing so. Competing against a human adversary is the ultimate challenge, since no AI yet has been made that can imitate the unpredictability of human thinking.
These are the proven features of multiplayer games that players tend to find enjoyable. How do we transfer these features into an adventure game? One idea in particular seems quite promising to me. This is all purely conceptual, and is just a brief introduction to the idea. Any comments and suggestions you might have are most welcome.
A Multiplayer Murder Mystery
I'm quite excited about this concept. Have you ever played Mafia, the party game where a group of people sit in a circle, draw cards to randomly determine who are the cops, the mafia, and the townspeople, and then try to eliminate the Mafia each turn? The idea here is similar. The setting of the game is, e.g., a mansion estate. Appropriate characters are chosen at the beginning of the game, such as a butler, a doctor, and so on. One player is randomly selected as The Killer. The others have to figure out who the killer is, and avoid being murdered in the process.
The game is very much an adventure game. Puzzles are solved, normally consisting of gaining access to certain areas of the play area, like the attic's drop-down stairs or the locked garage. Some items, such as a fireplace poker, a machete, or a gun can be used as murder weapons or as items to solve puzzles. Each player has an inventory, but only 2 items are allowed at a time. Because of this, players will constantly be dropping items. Every time someone touches an item, he/she leaves fingerprints on it. Certain items such as gloves or a handkerchief can be used to avoid leaving fingerprints or wipe them away, but these are difficult items to find. Fingerprints can be scrutinized by other players.
General Philosophy
As far as I'm aware, a similar game like this has not been made. It goes beyond mystery games like Clue!, since the killer is still very much active, and the life of each player is threatened with each step.
The idea is to invoke paranoia in each player. You don't KNOW who the killer is. You don't KNOW if the crowbar you see a character holding is going to be used as a murder weapon against you in a few seconds or just to open the locked trunk. You don't KNOW if the guy lurking in the shadows is going to burst out with a chainsaw or just passively observe you. Teamwork is possible, to figure out who the killer is, but the paranoia is always there, since you can never be sure you're teaming up with the right people. Whenever the player is selected as the killer, he would have to be very devious and put on a fake identity for everyone, trying to convince everyone he's just a normal non-killer like the rest of them.
The Server-Client Network
As a software developer, I'm very much aware that creating a multiplayer game is no easy task. The most effective way to create the game would be to designate one player as the server and the others as clients. Every time a player wants to do an action, it would poll the server, which keeps a record of where every item and every character is. It would simplify the game process and avoid duplicate actions (such as two people trying to pick up an object at the same time). Typical stuff, as well as a proven effective method.
The Atmosphere
The game undoubtedly would contain a survival horror element. Normally, survival horror games have to be carefully planned out and the atmosphere thoughtfully constructed. In this game, enough elements can be given to the players so they can construct their own survival horror atmosphere. In some rooms, lights can be turned off, casting shadows all over. Doors can be locked, so you can avoid the killer chasing you. Dialogue with other players should be restricted to the same room, rather than a global in-game chat, adding to the element of seclusion. Plus, this would avoid the problem of a sneaky player saying at the beginning, "BUDDY SYSTEM! Everyone choose a buddy and stick with him/her. If someone dies, we'll know the buddy did it!"
Replayability
The game can't be the same each time, since that might get boring. Items could be placed randomly (yet still in appropriate locations) whenever the game starts. The fact that someone different is the killer each time would keep the game interesting and replayable as well.
I'm going to draft a detailed design document for this game (when I get around to it), then post a link to it for comments and criticism. Hopefully, this will invoke further interest in the idea. Would anyone be interested in working on a project like this down the road? I think it has enormous potential. It could definitely leave a mark and be a good way to bring adventure games into the multiplayer world.
One thing you might notice is the lack of storyline. Stories are typically important to adventure games. There might be a way to accommodate a story, such as giving each character a purpose in the game, or adding a secondary quest for the players such as finding a hidden treasure, and throw in some NPCs. This may or may not be a good idea.
Eventually, I'd like to see different versions, i.e. different levels, of the game made. It doesn't have to take place in a mansion. The setting could also be a desert island, an abandoned amusement park, or (my favorite) a teenage party.
But first, I'd like to make a smaller multiplayer game. One idea is a Spy vs. Spy type game where two players try to sabotage each other. It would let us experiment with incorporating multiplayer features in AGS, and be a nice stepping stone before launching into the larger multiplayer murder mystery project. Some of the network features could be carried over, too.
