Hi Mark,

I tried to apply this with community leaders considering options in Bosnia-Herzegovina (Sanski Most and Prijedor) in 1996. The situation was a displaced persons "march home" which threatened escalation.  I approached it as an iterative consultation process (go-between mediation) to reduce potential for violent confrontation and increase awareness of the aims of others. Before you talk to actors, talk it through with knowledgeable outsiders (I used other IFOR officers and translators, which had the advantage of preparing the translators for the event). The starting point is to ask each party to describe their aims, their options, and what they think the other parties might do. I limited it initially to two actors (Bosniaks, Serbs) plus NATO IFOR, and picked the three most preferred options. Ranking more than three options tends to generate a lot of what-if hypothetical responses and preferences which are too hard to label or code in the first instance.  I checked responses with other actors, ran the software and then discussed results internally first, then with the Bosniak community leaders and the Serb police (Simo Drlaca). The latter consultation was not fruitful. You have to be aware that each side is gaming the opposition and the third party (NATO IFOR) throughout the process. The second round added options for IFOR, and took one away from the Bosniaks.  That's all we had time for before the event.  The reality was more chaotic and unstructured than this clinical description implies.  I think using GMCR effectively implies some education of the parties before they actually get involved in structuring any responses.  Gaming it with staff who are in contact with the parties may be a better initial approach.

Is this helpful?  

Thanks David! That's fantastic. 

I'm modeling an environmental conflict (Northern Gateway), and the biggest challenge is getting to the right level of elegant simplicity. There are so many decision makers involved, but with the whole advent of "social licence to operate" filling in the gaps that regulators leave by not addressing long term challenges, it makes the model murky. I like your idea of keeping it very simple, and letting the details emerge as the process of implementing and analyzing the model unfolds. 

They've also added a new layer to the analysis, where you can chose the ideal state you're looking for as a mediator, and find out what preferences would be required to get there. That's handy when you don't really know what the real preferences are, or when people are gaming each other as you mentioned. 

Okay if I keep you posted? 

Cheers! 

- Mark 

Yes, that sounds interesting, Mark.  I imagine the interface has changed a lot since I used it on a laptop with a 3.5" floppy, too. How many actors are you modelling?  Are you trying to generalize (e.g. environmental groups in general) or do you have one specific group that you are taking as typical? 

Good question. There are so many players involved with Northern Gateway, so when I tried to model them all it just got unwieldy and didn't really end up with anything useful. As I write up my decision maker rationale now I'm finding that I'm spending most of effort explaining why I'm leaving them out of the model, since it's supposed to be focused on DMs that actually have a move to make, and that's constantly changing. 

The new interface is actually pretty darn good. Let me know if you want to connect with the folks that developed it. Here's two articles that might be of interest. 

Thanks again!! You've been very helpful! 

- Mark 

Kinsara, R. a., Kilgour, D. M., & Hipel, K. W. (2015). Inverse Approach to the Graph Model for Conflict Resolution. Systems, Man, and Cybernetics: Systems, IEEE Transactions on, 45(5 May 2015), 734–742.

Systems methodologies to model third-party intervention in international conflicts are developed within the frame- work of the graph model for conflict resolution (GMCR). An inverse GMCR is introduced to utilize the GMCR as a negoti- ation tool by altering the procedure of the original framework. The methodologies presented give a better understanding of how decision makers (DMs) can be motivated to undertake certain actions within the conflict. Moreover, the inverse GMCR tackles the problem of specifying which preferences for DMs lead to a particular resolution, thereby making it easier for a mediator or other third party to influence the course of the conflict. The methodologies are applied to a real-world dispute, a complex water conflict in the Middle East.

Kinsara, R. a., Hipel, K. W., & Kilgour, D. M. (2013). Inverse Approach in Third Party Intervention. 2013 IEEE International Conference on Systems, Man, and Cybernetics, 2433–2436.

A historical water conflict along the Euphrates River in the Middle East is investigated and analyzed using the inverse approach to the Graph Model for Conflict Resolution (GMCR). The essence of this methodology is to inform the mediators of the possible approaches to reach a more desired resolution. The goal is to allow mediators to focus their efforts on influencing the involved parties efficiently in order to achieve an improved resolution. Index Terms—Third Party, Conflict Resolution, Graph Model, Mediation, Inverse Approach.