How do we talk about communication in multiplayer games?

I’m posting this now because it’s been percolating for weeks so I might as well.

In other news, I’m getting tired of these titles, but I like the consistency…sigh. Problems in academiaaaaa.

Let’s talk talking. How do multiplayer games implement easy, fast communication? Communication is one of those things that we take for granted in off-line (analogue?) play because it comprises so much of the play experience that it is invisible. For example, a tabletop role-playing game only exists when players communicate to the GM what actions they are taking. Similarly, a game of chess-by-mail only exists because two people write to each other and state the move they wish to make. Communication is, quite simply, the name of the game.

In both of these examples, players communicate intent within rule systems. It’s not as easy as I made it sound up there; communication isn’t the only thing reifying these game instances. The established structures of rules and notation and yes, communication styles, empower players to act in meaningful ways.

It’s never as easy as “give them voice chat and they will come”. Opening up such a system to the public invariable invites abuse and trolling, and does not always work well with the existing game. Players in a group of friends often use outside services such as Skype, Mumble, Teamspeak, or Ventrilo. On the other hand, these things are not friendly to strangers trying to create a cohesive group from within the game itself.

Multiplayer games, especially ones with large and diverse player bases, have taken different approaches to easing communication inside game spaces. There are a few that I’m familiar with through play, and a few that I’ve been pointed towards through exploring this. I do not present these with any extended commentary, but just to show the range of options and the questions raised by each one.

Team Fortress 2 supports voice chat in game through a simple press of the “v” key, though this feature may be abused; players have the ability to mute other players for their own sake. More interestingly, Voice Commands are built in to the default key bindings of the game. These voice commands are often context- and character-sensitive. There are three menus of voice commands, accessible from the “z”, “x”, and “c” keys. Each of the menus has a set of 8 commands further triggered from the 1-8 number keys, with 0 reserved as a cancel. The commands are pieces of information that the designers felt required easy and fast communication between players; they are pieces of information that make the experience of play qualitatively better.

Looking at the voice commands in Team Fortress 2, we see certain categories repeat. Most important are the calls to the medic for help:

Due to its importance and universal nature, the “Medic” call by default has its own separate keybind (default E).

– “Voice Commands”, Team Fortress Wikia

But beyond that, there are a set of commands for informing team movements (z3-z6), commands for revealing information about enemy team movements (x1-x2), requests for aid from the support classes (x3-x7, and the aforementioned “Medic” call), as well as general communication such as “Thanks”, “Yes/No”, and all of the calls under the c menu. With this set of 24 voice commands, players should theoretically be able to communicate their intent in a restricted domain (the game space). Anything left over is then relegated to voice chat or text chat.

With this kind of system where critical pieces of information are bound to particular keys, conveying that information is easy and leaves no room for trolling (except creatively, such as by spamming “Put dispenser here” as a heavy, which leads to a chorus of “pootis pootis pootis”).

In Left 4 Dead/L4D2,as in Team Fortress 2 Valve have taken cooperation as the goal of gameplay. The GDC slides that I’ve linked before (“Replayable Cooperative Game Design: Left 4 Dead”, Michael Booth, GDC 2009) take one slide to focus on “vocalizations” (“Requiring Cooperation: Vocalizations”, slide 25 of 71). As described in this slide, these vocalizations are automatically generated by the player characters and NPC human allies based on contextual clues. They have three main purposes that enhance cooperative gameplay: “Improv[e] situational awareness”, “Communicat[e] short term goals”, and “Encourag[e] cooperation via a baseline of camaraderie” (slide 25).

With Team Fortress 2, we’ve seen what happens when you take communication of situational awareness and short term goals, and map it to keys rather than leaving it to players to type up long statements in chat, or to press a button and then voice chat to the subset of the server that is currently on a third-party service. It makes for easier teamwork, and often a friendlier set of interactions. (I wonder. This is based on personal experience.) It also models the appropriate forms of communication for the system, as these are the twenty-four commands that the designers have chosen as vital to the experience of play.

With Left 4 Dead (and L4D2), we can see what happens when, instead of mapping those to key-presses, you remove those from the conscious decision-making process of the player. The human allies will always produce vocalizations, regardless of whether they’re played by a human player or controlled by AI systems.

We also needed a way for the AI characters to shout out important facts in the world. If I’m playing with my friends, and one of them sees an ammo cache in a dark corner, then my friend can just tell me. That’s an important thing to know; the ammo is in different places every time. … Once we did it for the bots, we realized it was pretty cool for the human-controlled characters to do it also; it’s a bit of additoinal roleplaying, and more convenient than having to call out yourself.
– “Rule Databases for Contextual Dialog and Game Logic

…To be continued…

One thought on “How do we talk about communication in multiplayer games?

  1. This reminds me of the days when I was just getting started exploring the possibilities of the internet. It was around 15 years ago, I think (early teens) and the arena was Quake, specifically Capture the Flag. The strategic aspect of it (as compared to the chaos of Deathmatch) appealed to me a lot, though I must say I never really became any good at multiplayer Quake.

    Anyway, there were no preset communication shortcuts at the time, but people quickly discovered that you could bind chat messages to particular keys using the console. We sort of picked these things up from each other and used them as necessary.

    In my case the play was usually quite casual and though there were clans on my server, there were rarely any organised competitions. That meant that for each game, there wasn’t really any structured a priori strategy or positioning for most players in a team, and things just organically worked themselves out. Maybe something like defense/offense, but rarely anything more sophisticated than that.

    As a result, players adapted to the needs of the situation, and so did the chat bindings. I think my arsenal consisted of basic position calls (“I’m defending”, “I’m attacking”) and responses to particular emergences in the game (“I’ve got the flag, defend me!”, “I’m hunting the enemy flag carrier, assist me!”, “Our flag has been taken, everyone start hungintg!”), etc. Probably shorter version of some of these, but you get the idea. Oh yeah, and obviously “LOL” and “:)” were essential bindings, too.

    So what I guess I want to say is: for me, chat bindings were flexible, and a way of structuring play and strategy on-the-go. Team tactics were only very loosely discussed beforehand, but emerged out of the play itself, and quick chat cues signaling each player’s activity and position helped structure it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s