Conflict simulation, peacebuilding, and development

Engle: Iterative matrix games


Yesterday, matrix game inventor Chris Engle provided an overview of the history of the matrix gaming method for PAXsims. Today he offers some more thoughts, this time on bringing narrative games and quantitative analysis together.


Iterative matrix games: Bringing together narrative games and quantitative analysis

reserach-charts.jpgMatrix games are a simple narrative simulation technique that maximizes the creative input of players by allowing them to make up what happens next. They say what they imagine and this allows those watching to get a glimpse of the inner workings of a player’s mind. The game is a conversation between players and is not built on an underlying mathematical algorithm. It is the opposite of a mathematical game theory approach. Consequently it is easy for designers who work in quantitative sciences to dismiss the technique.

There is a potential to bridge the gap between soft narrative approaches and hard quantitative ones. It is in iterative play. This has not yet been done but here is how I think it could work.

The game is played on an electronic platform. Participants register for the site by giving their demographic details, whatever details are appropriate for post-game analysis. Players gather on the site as they would on social media. Each game starts at a preset time. Players are presented with a brief scenario and asked to start.

Players then type in the actions they want to happen. This happens in real time so participants are both audience and actors. They may add to and alter the story as they see fit. Each addition adds to the narrative but also is linked to the writer’s demographics and sits in relationship to the actions that precede and follow it. In addition the platform could allow off side comments to further supplement the data collection. If games are limited to no more than twelve players it should work. More than that and play could spin into chaos as players are unable to keep up with the flood of new actions. Too many participants and information overload results.

If the same scenario is run repeatedly, for the same players or for new sets gamers, the electronic record builds up a data base of outcomes that can be subject to statistical analysis. So while any individual game may be “just made up”, when aggregated they form a picture of a crowd and crowds have wisdom.

As of now I know of no way a computer can breakdown individual stories to categorize them but I can imagine that coming. When it happens matrix games can be a way of melding together the strength of people and machines. The game combines human imagination with machine data crunching. The potential for useful research here is intriguing.

Chris Engle 

5 responses to “Engle: Iterative matrix games

  1. Chris Engle 25/08/2016 at 8:06 am

    An interesting thing that might show up in the data is how participants call for rolls. This would in aggregate show biases. Which shows another part of the matrix.

    It all comes down to coding doesn’t it. No computer can do it. It is either the players or a coder.

  2. Rex Brynen 29/07/2016 at 9:32 pm

    You could already do the gameplay and data collection on the existing ICONS simulation platform. You would still to need to code (and analyze) all the player actions and discussions, though.

    Yuna Wong (RAND) and Sara Cobb (George Mason University) have done some interesting work on the analysis of narratives in seminar wargames, which is a rather similar process.

  3. Chris Engle 29/07/2016 at 9:08 pm

    There would be the investment in server space and some programming skill but I could see reusing programming from something like Facebook. The server space would be for the data base. The labor would come in coding the individual arguments. That has to happen before number crunching. You could use an underpaid grad student or ask the player to code their own action after they made it. The researcher would set the coding elements they wanted to study. Maybe some days computers will be able to do this without human intervention but not now. I think this could still be done comparatively inexpensively.

    I can imagine how to do it even though I can’t do it. Hopefully some with the research skills will step up for this part. They would get a lion’s share of credit for matrix games. I’d love to see this.

  4. David Manheim (@davidmanheim) 28/07/2016 at 7:17 am

    Very interesting points; it seems similar to “Social MUDs,” or MUSHes ( ) especially as they were used in the 1990s – though some are still active today – see . For a more modern incarnation, see also:

    The point I would add is that having a role that is similar to Game facilitators / RPG dungeon masters / MUSH wizards seems useful for both arbitration and guidance. If that role is served by one of the deisgners or runners of the game, this could pose some obvious questions about how they bias results, but having independent parties, or players assigned to the role, might alleviate that concern.

  5. Robert (Bob) CorderyBob Cordery 27/07/2016 at 2:40 pm

    This would be an interesting development, although I cannot see how it could be achieved without a considerable investment.

    The big advantage of Matrix Games is their ability to include ‘the human in the loop’, the most difficult element in any game system to replicate. If it is possible to meld that into a game system that could be quantitively analysed, it would be a great leap forward and would answer some of the criticisms of Matrix Games.

Leave a Reply

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

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

Google photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s

%d bloggers like this: