Wargaming Small Wars (merged thread)

[sullygoarmy]

Trying to "game COIN" might be like trying to "simulate" marriage just using an inflatable sheep. The lack of realistic responses would show up pretty quickly!

I won't touch the inflatable sheep comment but in terms of simulating marriage, isn't that what 2nd Life is supposed to do? I've never tried it but read plenty about people who literally have a simulated marriage online. Personally one real one is enough for me...I'd rather hunt "tangos" if I'm playing online.

Along that thought process, however, is using something like a 2nd life simulation to mirror COIN. You can replicate every type of personality in a known AO, have events from a car bombing to an election drive and best of all, you can run this over a looooong period of time. For example: you know your unit is deploying somewhere next year. You have an idea of the AO (always subject to change of course). About a year out, you inject your key leaders into this 2nd Life environment. Even an hour a day whether from work or the house over the course of a year would help them build up a huge knowledge base based on actual people, locations and events. If kids sitting at home on a Saturday night can design these types of environments, why can't STRICOM and the military?

May be worth a look. Thoughts?

[JT Clark]

The recent COIN Leaders seminar at the CAC Ft Leavenworth was a great intro to how actual ground-up intel can (and should) be used to model complex COIN scenarios. Combined with a knowledge of basic insurgent strategy it is then relatively straightforward to forecast likely insurgent actions and come up with pro-active rather than reactive strategy on all fronts- military, social, economic and political. Gaming such scenarios was an eye-opener to all that attended.

On a side note, counter-strike and ghost recon fans out there might be interested to see that Hezbollah has their own version: http://en.wikipedia.org/wiki/Special_Force

The interplay between globalization, media, and insurgency is always fascinating...

[Entropy]

I've been a "gamer" all my life, beginning with military board games and going right up through MMORPG's (World of Warcraft being the most recent). I agree with Wilf that "simulation" is probably a more accurate term, though I would suggest "model," at least for certain types of "games."

Like any simulation or model, there are limits to what it can do and there are some things that simply can't be modeled/simulated very accurately. In my experience, those un-modelables are often the human and leadership factors. I think we're all in agreement that those human factors are preeminent in insurgency/COIN, so for that reason alone I think games/models/simulations will have less utility in COIN than, say, naval warfare. In other words, it's easier to simulate hardware than it is to simulate people.

[JT Clark]

Couldn't agree more. It is definitely possible using some basic modelling techniques however to ID the various stakeholders in a COIN scenario and predict their most likely courses of action.

[BayonetBrant]

I think one of these days I'm going to have to write an article about "Game" v "sim" and their best uses in training/learning - already did a bunch of this for my dissertation... I just need to cut it down to a consumable size.

[Rex Brynen]

It is not that difficult to simulate a COIN/peace operations setting in a human-moderated role-playing game, given the flexibility and large number of variables that an RPG allows one to include. (I'm more dubious about electronic ones at present, since player options are starkly limited and there is to much temptation to metagame the AI system.) You can also fairly easily scale for the scope of the exercise, number of participants, presumed level of experience. As I've posted in other threads, I've run them for both large numbers (120) of students and smaller numbers (30) of senior officials/negotiators/technical experts. The real challenge is preparing an adequate amount of appropriate background information and briefing notes, and making sure that the proper dynamic starts among the players early on.

The World Bank has started used a heavily modified version of UN DPKO's mythical "Carana" setting for training mid-level officials on aid operations in conflict-affected countries. It focuses on determining national policy priorities with a diverse group of stakeholders.

(Very advanced heads up—I'll be in DC at some point this fall giving an in-house talk at the World Bank on the use of simulations in education and capacity-building. PM for details.)

[nichols]

Something that needs to be watched is that when a moderator controls certain aspects of the 'COIN' delivery, you get the moderator's unintentional view of COIN. Not sure if that makes sense, I try this: If I was moderating, no matter how much I studied the assignment there were life experiences that I have had that would influence my decision making process.

I' thinking that a base line AI of an area that is editable by the end user would probably be a good thing.

You have orders conduct a NEO in Bangui, the intel bubbas update the AI with current culture issues and you begin the mission rehearsal......

[Rex Brynen]

Something that needs to be watched is that when a moderator controls certain aspects of the 'COIN' delivery, you get the moderator's unintentional view of COIN.

Yes, absolutely--every sim has a great deal of assumptions built into both it and the way it is controlled/run/adjudicated.

A second and related issue is the importance of participants not feeling that their are being maneuvered by the Sim Gods towards some predetermined route and outcome—and how to balance this with the desire by the designer and control team to impart particular lessons (when it is being used in a training or educational capacity).

[defense linguistics]

See Brian Train's Algeria. I think that the simulations and games communities have been so dominated by mathematics-based thinking that language based approaches to the problem have been neglected. Chris Engle's Matrix Games seem promising in this regard. The "End State Problem" is indicative of this. Properly writing desired end states is always a linguistic task. "Victories" and outcomes are ultimately defined in langauge. Perhaps games used to the end of better understanding COIN should be accordingly language-based.

[Rex Brynen]

I think that the simulations and games communities have been so dominated by mathematics-based thinking that language based approaches to the problem have been neglected.

Given the number of people with RPG backgrounds (myself included), I'm not convinced that this is entirely the case. Most good RPG campaigns are almost entirely language based, despite the use of mathematical systems for combat resolution.

It may be that clients, however, like to see heavily rules-based simulation systems. They also tend to be more "portable", since they don't rely so heavily on the skills of a sim moderator.

(Yes, I almost said DM.)

[AmericanPride]

I imagine that a sandbox model would be the most approrpriate for a COIN simulation. My friends and I are having similar issues in developing a simulation for the Dune universe as there are many political and economic models, some of which contradict. The solution I've found is to give the players a wide variety of options with specified effects and the let player find their own goal(s) and path(s) to that goal. Alternatively, modelling a specific theory of insurgency should alleviate some of those problems by providing the framework in which to measure and define action.

[Beelzebubalicious]

posted in wrong string. ignore

[Beelzebubalicious]

Anyone know about this simulation run out of Purdue University:

The Sentient World Simulation project (SWS) is to be based on SEAS. The ultimate goal envisioned by Alok R. Chaturvedi on March 10, 2006 was for SWS to be a "continuously running, continually updated mirror model of the real world that can be used to predict and evaluate future events and courses of action. SWS will react to actual events that occur anywhere in the world and incorporate newly sensed data from the real world. [...] As the models influence each other and the shared synthetic environment, behaviors and trends emerge in the synthetic world as they do in the real world. Analysis can be performed on the trends in the synthetic world to validate alternate worldviews. [...] Information can be easily displayed and readily transitioned from one focus to another using detailed modeling, such as engineering level modeling, to aggregated strategic, theater, or campaign-level modeling

http://en.wikipedia.org/wiki/Synthet...nd_Simulations

[nichols]

Anyone know about this simulation run out of Purdue University:

Talked a little about it here:

http://council.smallwarsjournal.com/...3&postcount=73

We still want to make this happen. We have a ROM for modeling the 17 areas of the CMRS.....it appears funding is available....just haven't figured out how to get it right now.

I'm thinking probably around end of first quarter we might be able to move forward.

[Eden]

I am very skeptical about the utility of gaming COIN at an operational level for several reasons.

1. The complexity of the COIN environment makes model-building extremely difficult. It is also very hard to filter out pre-conceived notions and false assumptions brought to the table by modelers or participants.

2. The 'lessons' and 'principles' from previous COIN operations are rarely transferable - I believe that every COIN op is unique to a far greater extent than more conventional operations.

3. For high-fidelity you would need a very large number of participants, which can be unmanageable, or some very sophisiticated software capable of simulating the COIN environment for a small number of players. The latter does not exist.

4. COIN is slooooooooow. Things develop over months or years. The number one difficulty is figuring out whether you are being successful or not, and why. It requires patience, perseverence, and downright mulishness. You can't game that with people, and we don't have the software to realistically game it using computers.

Now, I think you can usefully game COIN at the tactical level, i.e., to practice specific tasks. But anything above that...well, I would be very suspicious of any 'insights gained' or 'lessons learned'. I think the participants would be better off spending the time readin books and studying history.

[sullygoarmy]

I go back to my 2nd Life Idea. Building a COIN environment in a 2nd Life world is very doable in my opinion. Imagine a brand new 2LT who spend a few hours a week for a year walking around, talking to and training for the environment he will be operating in once deployed? Your role players all have characters in 2d Life, learn the nuiances about the actual leadership from those currently in theater and simulate those characteristics.

IMHO, success is getting that leader a better understanding of the environment he/she is about to enter. There does not need to be a "win" in the game, or even an endstate...just like real life. The goal is to educate and get leaders used to having to deal with an environment that hasn't changed in a century and will barely change while that leader is there.

Personally, I'm more interested in the immersion experience a good simulation can provide for leadership than I am building a better game.

[Steve Blair]

I am very skeptical about the utility of gaming COIN at an operational level for several reasons.

1. The complexity of the COIN environment makes model-building extremely difficult. It is also very hard to filter out pre-conceived notions and false assumptions brought to the table by modelers or participants.

2. The 'lessons' and 'principles' from previous COIN operations are rarely transferable - I believe that every COIN op is unique to a far greater extent than more conventional operations.

3. For high-fidelity you would need a very large number of participants, which can be unmanageable, or some very sophisiticated software capable of simulating the COIN environment for a small number of players. The latter does not exist.

4. COIN is slooooooooow. Things develop over months or years. The number one difficulty is figuring out whether you are being successful or not, and why. It requires patience, perseverence, and downright mulishness. You can't game that with people, and we don't have the software to realistically game it using computers.

Now, I think you can usefully game COIN at the tactical level, i.e., to practice specific tasks. But anything above that...well, I would be very suspicious of any 'insights gained' or 'lessons learned'. I think the participants would be better off spending the time readin books and studying history.

Pretty much everything you tagged out here can also be a problem if you just read books and study history without the proper background and training in the discipline. All history is subject to bias, and without a proper program it is very easy to develop a self-selected reading program that just reinforces what a person already "knows" is true.

Within the gaming community (both online and offline) there are numerous examples (not counting Warcraft and other pay for play MMORGs) of games carrying on for years...often with the same participants. You can actually game some pretty grinding, extended episodes, often with politics involved and some social planning/engineering. The RPG community (though much-maligned by some in other circles) would actually have a great deal to offer in terms of frameworks, methods, and procedures for this sort of extended gaming. It also allows you to simulate things that HAVEN'T happened yet.

Will there be some bias? Sure. But I also contend that there always is bias where people are involved...be it in the structure of simulation mechanics, programming of AI, and so on. There's also bias in the real world. A good control team can minimize bias, or develop mechanics to cover some of the aspects of the simulation that might be more vulnerable to bias.

To me the COIN environment is something that just begs for an RPG-type approach. I can't see systems having the same impact.

[William F. Owen]

errr... Seriously, can we do this? Is it possible for SWC to run a "simulation." We all might learn a lot. Dunno how on earth it would work. My guess is that Mister R Brynen might usefully add something.

- and I have my own inflatable sheep!

[Ken White]

--or depths. I am seriously not touching that sheep...

[AmericanPride]

Eden,

While you bring up sound objections, I disagree that they make gaming COIN impossibly difficult. Games have reached a point of creativity and depth that most things, given the proper focus by developers and interest by gamers, can be simulated.

1. The complexity of the COIN environment makes model-building extremely difficult. It is also very hard to filter out pre-conceived notions and false assumptions brought to the table by modelers or participants.

I think this can be addressed in one of two ways. The first is by using a sand-box mode for the game, which is an extremely popular model. Europa Universalis 2, Hearts of Iron, Victoria, and Geo-Political Simulator are a few examples of complex games played sand-box style. In EU2, the player guides the development of a one of several hundred countries over the course of 400 years, managing its political, economic, social, and military progress as well as balancing the ethnic, religious, resource conflicts within and without the state. GPS, in constrast, places the player in the executive seat of a modern state where he must manage every conceivable aspect of running the nation, including managing the different personalities leading political parties, the military, interest groups, religions, minorities, the media and so on. Both games are real time, with Eu2 running day-by-day and GPS hour-by-hour. Because there are so many options, the player can really take any approach he likes; so different models of politics, economics, etc can at least be simplisticly modelled by the course the player charts. The second option is to build a game that models a specific theory of insurgency. AGEOD's American Civil War, for example, is very detailed in its application and so many concepts and strategies which may be useful in generic strategy games can be severely punishing to the player. For example, in most basic strategy games, the player who attacks first usually wins. in AGEOD's ACW, the design emphasizes operational tempo as opposed to speed. I'm generally in favor of the sand-box design -- the more freedom of action given to players, the more ingeniunity you will see out of them.

2. The 'lessons' and 'principles' from previous COIN operations are rarely transferable - I believe that every COIN op is unique to a far greater extent than more conventional operations.

This would have to be inherent in the game's design. In a sand-box model, the player can apply his own principles and learn his own lessons. That could realistically change from session to session depending on what course the player is pursuing. In a specific model, the designers would have to working with a specific set of lessons already in mind.

3. For high-fidelity you would need a very large number of participants, which can be unmanageable, or some very sophisiticated software capable of simulating the COIN environment for a small number of players. The latter does not exist.

Not necessarily. EU2 and GPS have hundreds of autonomous simulataneously operating participants. EU2 can also run good-sized (16+ players) games. There are several online-browser based RPGs, such as Star Wars Combine, which are even larger. SWC not only drops the player as a single character in the middle of the entire Star Wars universe, it also runs in minute-by-minute real time. The game has been operating for 10+ years now with some of the original players and organizations still active.

4. COIN is slooooooooow. Things develop over months or years. The number one difficulty is figuring out whether you are being successful or not, and why. It requires patience, perseverence, and downright mulishness. You can't game that with people, and we don't have the software to realistically game it using computers.

See above. Most games, such as EU2 and GPS, also provide the option of controlling the speed, though not necessarily the pace, of the game.

Just my 2 cents as one from the gaming generation.