Multiplayer Conversation Modelling 


Indraneel Kale

 Shashank Gargeshwari

I am a relatively social person. I am approachable, fun to be around, and have a large circle of friends. There is one caveat though. I. Don't. Call. Or. Text. My. Friends. Never did. This means that if I am not physically around someone, our contact teeters. I have not been in touch with people I consider to be my best friends for months now, and it feels unfair that all we get as means of communication is text or call. Although I have recently tried putting in more effort to stay in contact with people, it still feels unnatural to talk to someone over the phone.


There are various discrepancies I've observed in communication using phone calls, like the inability to see the other person's facial expressions, the chance of cutting each other off due to lack of visual cues, and the ever dreaded bad connection. There was one other subtle aspect though that I felt contributed to a lot of my ultra-short catch-up calls: the lack of opportunities to pause and NOT talk. You see, when you are on a call, you better talk, because if both of you can't think of anything to talk about, then that's the end of that call. 

Sooooooooooo... How are, like, your ancestors and stuff?

It's a paradox innit? To get people to talk more, we must allow them to not talk. Now how are we to solve it? My initial hunch was to move talking away from the limelight and make something else the center of attention. What would take its place? A game of course! Now, lots of games have team chat, but most are really fast paced, not allowing for a slower pace of conversation. I envisioned a game that compelled people to have slower, more deliberate conversation.


I decided to make this game for one of my Design Projects at NID. This lead to Two Fold, a game based on what I call Multiplayer Conversation Modelling. In fact, me and my classmate, Indraneel, decided to take it further for the next Design Project and created a working prototype, Two Fold 2.0, in which he took up the art production while I did the programming. The game design was driven primarily by the types of conversations it could evoke from the players. Here's how I went about designing the game. 

The Design Process

Hey, look Abhi, I can fly! (Two Fold v 0.1)

My initial brief was something like this: "Two Fold: A Game to Replace Phonecalls!!!!", but would it replace all phone calls? Surely it wouldn't replace, for example, a phone call to book an appointment at a hairdresser.  What else would it not replace? It wouldn't replace a call to inform someone that you got a new job. It wouldn't replace calls to invite someone to a party. It certainly wouldn't replace calls to emergency services, so what kind of phone calls was I talking about? I guess when I was conceptualizing the project I was thinking about bonding-talks, the kind of talks you just happen to have without a particular agenda. Talk for talk's sake, if you will. You'll notice this is exactly the kind of talk that would fall victim to the can't-think-of-a-topic-so-let's-end-the-call-here phenomenon I know too well. Keeping this in mind, I went on with creating the core game mechanics of Two Fold. 

How come he get to shower while I fight a lion? (Two Fold v 0.1)

Since the core mechanic of the game was talking, I had to create situations in the game that would compel people to talk. I decided to go with a two-player real-time game as a starting point. I could have chosen other formats, but I decided on a 2 player action platformer because I am a hardcore fan of the genre. The story was non-critical, so a typical searching for lost treasure refrain was used. Once all this was decided, there seemed to be two dominant things the players would be doing while playing the game :

a) Talking to each other

b) Performing in-game actions


I needed a simple model of both conversations and actions, so I decided to pick from among a few of them outlined on the internet. The first one was a simple model for "Elements in a Conversation" picked from I'll reproduce the list below to make it easy for you. 



This relatively simple list would prove tremendously helpful in creating game mechanics and level design for Two Fold, but it was only one half of the formula. The other half was the action-modeling. For this, I headed over to to borrow from their take on player interaction. They talk about two axes of player interactions, one denoting the intensity of interaction with the other denoting the friendliness of interaction. For Two Fold, I had decided to go with co-operative play, so the first axis was deemed more important. I'll outline what that axis laid out below:


  • No Interaction: Players can't affect each other's game state or strategies

  • Implicit Interaction: Players can't affect each other's game states, but can influence their strategies

  • Explicit: Players can explicitly affect each other's game states


With all this laid out, it was time to dig in and start designing the game mechanics and levels for Two Fold. 


The Game mechanics

From the list of elements in a conversation outlined previously, I found Asking to be quite potent. This led to the design of selective information hiding in the game. If each player in Two Fold had only partial access to the information in the game world, it would urge them to communicate to ensure that both have a complete and accurate understanding of the world that they inhabit. Cue The Wall. Two Fold's world consists of a corridor for each player with a wall separating them. Players would still be able to see each other regularly, but the wall will be designed to hide key elements in the level that needs communication between the players to solve.

[Placeholder Image] : Peekaboo!

Players can move left and right, jump, and strike with their sword. The same sword strike is also used to interact with other elements in the game, like activating levers or attacking enemies. Player co-operation is needed to solve environmental puzzles and 

The levels

With the game mechanics mentioned above, a series of 3 levels were designed. The Ancient Temple Puzzle level was prototyped in Two Fold 2.0, with further development planned to produce a complete game. The levels that were initially conceptualized are :


1 - Sharaba Boss Fight
  • Hero - Victim: One Player has to rescue the other from a seemingly invincible beast by pushing boulders onto it from the other side. One player is powerless and is dependent on the other to rescue them.

  • Speech Elements: Asking, Informing, Building, Supporting

  • Player-Player Interaction: Implicit


2 - Ancient Temple Puzzle
  • Interactive Puzzle Solvers:- Both have to fiddle around with levers to open a door for each other. Both have to think about it together while solving individual puzzles on their respective side

  • Speech Elements: Asking, Informing, Supporting, Asserting, Proposing, Summarizing, Checking, Building

  • Player-Player Interaction: Implicit


3 - Lion Boss Fight
  • Brothers-in-Arms: Both players fight a Lion that can jump between both sides of the middle dividing wall. Players can heal from a spring on both sides, but it takes time. This creates an on and off state for action for both, and they depend on each other to stay alive and kill the Lion.

  • Speech Elements: Asking, Informing, Supporting, Asserting, Proposing

  • Player-Player Interaction : Implicit

[place holder image] : v 0.1 and v 2.0

Of course, all of this is just words and sentences, and it is time you and a friend of yours tried out Two Fold and left some feedback on my contact page. After testing with players, it was apparent that Two Fold was an enjoyable and unique Two Player experience that offered a lot more than what has been explored here. I am planning to continue development on Two Fold and release it for free when it's done. The comparison pic above shows how far the game has come over the course of  6 weeks. I wonder how it will look with another half an year of development.  

Graduate Student | Digital Game Design 

©2018 by Shashank Nagaraja Gargeshwari.