I’ve made some tweaks to my rules for Fardh al-Qanoon. The updated rules are available on Scribd, while the Map and Counter Sheet are available for download as usual. A changelog can be found at the end of this post, with my design notes available here.

Now that I’ve submitted a version of my simulation for my course, my plan moving forward is to create a separate version of Fardh al-Qanoon not bound to the original assignment’s parameters. Future versions will include more playable factions and, hopefully, a more accurate representation of the insurgency.

Map & Counter sheet

Download Map on Google Docs
Download Counter sheet on Google Docs
Download Rules on Google Docs

Changelog

• Reinserted critical missing rule about advancing Sunni Relations Counter in G-12.
• Clarified difference between securing Sunni neighborhoods and districts in B-6.
• Added clarification in E-3 about Militias moving from Sunni neighborhoods into the district the neighborhood is in.
• Added clarification in I-2.
• Fixed typo in example of play.

As someone with no prior manual wargaming experience, I found the process of designing a counterinsurgency simulation profoundly frustrating and, simultaneously, immensely gratifying. I believe the limitations imposed by the assignment and the medium of a tabletop game force a designer to develop a deeper and more thorough understanding of a conflict than they would otherwise have to with an analytical essay. A written argument is flexible and there is a level of inherent ambiguity in its presentation, whereas a conflict simulation is a testable hypothesis (though there is, of course, plenty of room for debate in its design as well). In retrospect, I probably would have chosen a far more traditional conflict for my introductory experience to simulation design, but I found the challenge of recreating a counterinsurgency operation in game form to be thoroughly rewarding and learned much about not only counterinsurgency and the Baghdad Security Plan itself, but about creating and testing models.

My first observation about simulation design is how much more rigorous the research process was than for other academic assignments I’ve had. The amount of information required to design a useful simulation demands a very thorough research process given the complex dynamics at play in a war or battle (perhaps even more so in an insurgency). While any piece of academic work requires thorough research, I believe the variety and depth of information required for a simulation distinguishes it from an essay. In my own simulation, I found the lack of quality information for certain aspects of the conflict (in particular, anything to do with non-Western forces) to be a challenge, and it was fascinating to see how using the simulation helped me partially “fill in the blanks” and gave me a basis for modifying my research. For example, finding information about opposition manpower in Baghdad was nearly impossible, with sources varying from being simply unavailable to outright detrimental because they presented unreliable guesses. As I reviewed other counterinsurgency simulations and created a storyboard version of my own, I realized that I could approach the problem of “quantifying” the opposition from another angle: looking at their effects rather than the means they used to achieve them. The Coalition collected ample data about violence throughout Iraq, and while the usefulness and methodology of collection is a topic unto itself, the overall trends in the data provide a picture of how the Mahdi Army (the primary opposition of my simulation) changes over the period of the Baghdad Security Plan. Verifying my analysis meant plugging that information into my simulation and seeing if it fit with my overall conception of the conflict, an option I would not have with a traditional essay.

The most valuable part of my experience designing this simulation has been the exercise of a different set of analytical skills. I have always had an interest in game-based models, having played computer strategy games from a young age and “modding” them for realism. My interest was always a hobby, however, and the results never had anything resembling academic rigor.  Attempting to design realistic and useful simulations certainly involves skills applicable to a wide range of disciplines beyond conflict simulation, since you are essentially creating a testable model. The same principles can be applied to modeling politics, economics, or any number of systems (albeit with very different characteristics), and the analysis skills employed are universally useful. I believe the reason simulation design is a constructive intellectual exercise at all is the avoidance of counterfactual history. Modeling a conflict is about finding what did happen and creating a system that shows you how (which goes a long way to answering the why). It is tempting at various stages to ask “what if?” My biggest concern with simulation was how a designer could plausibly “prove” a set of rules, since history only provides a single trial’s worth of data. In creating game mechanics that model particular dynamics of a battle or war, the designer seeks to explain just one historical outcome. While doing so may provide some insight into how things could have gone differently, a simulation does not aim to provide predictive answers to the “what if?” questions that inevitably arise. When I came to terms with this fact, I found designing my simulation to be far easier. I was too worried about being theory-agnostic with my design, since different writers claimed different elements of the Baghdad Security Plan were the drivers of the reduction in violence. Was it the increased troop presence, the change in tactics, the relatively demure Shia response, or perhaps the “Sunni Awakening”? Instead of trying to say “here is what happened, here is why the other one did not happen” the simulation says (or attempts to say) “here are the elements, and here is how they are related.” The magnitude of each element’s effect on the outcome is debatable, so the simulation must be flexible enough that someone can alter a rule or tweak a variable and thus create a simulation that models their hypothesis instead.

As a pupil of the computer age, I found the limitations of a manual simulation to be chafing at times. At every stage of the rules design, I could not help but think of how I could model a mechanic more realistically with a line of program code. The simplicity requirements of the assignment bordered on the austere, especially for my chosen conflict, but I am tentatively satisfied with the end result after several playtests with someone who was not familiar with wargames or Operation Fardh al-Qanoon. After playing the game as both the Coalition and Mahdi Army several times, this person was able to identify the key elements of the conflict and the challenges faced by each player faction (and, to some extent, the non-player Sunni faction). While I don’t think a manual game can fully cover the hideous complexity of Fardh al-Qanoon (not to imply that a computer-aided simulation would have a much better go at it), it is far more accessible to the audience and players, since they are aware of everything that determines the outcome. They know every rule, every die roll, and every decision that led to a certain result; my playtester, for example, correctly saw that there were no traditional “fronts” in the people-centric conflict I was trying to simulate, but that each player’s relationship to a third actor (the population) created a similar dynamic as the Mahdi Army was rolled back district by district until they were only able to easily operate in one or two areas, usually including the Shia stronghold of Sadr City. Knowing what effects Joint Security Stations have, what Mahdi Stacking Limits were and what they were based on, and what caused Mahdi influence to increase and decrease made the conflict accessible in a way that a computer simulation would not, in part because computers programs are opaque (although this can be mitigated with detailed logs), but even more so because of the complexity it enables. The nature of a manual simulation necessitates simplicity; anyone who has had to debug program code knows how incredibly complicated it can be to identify the cause of a single outcome in a complex program, much less the sequential series of outcomes that the result of a wargame entails.

To say I learned a lot from designing and playing simulations in this course would be an understatement. Whenever I read about a battle or war now, I immediately start thinking about how I could model it. As someone with no military background, I find it much easier to comprehend and absorb new operational concepts because I can conceive of them as being part of a model with certain rules (in the systemic rather than game sense), thus making them more accessible to me. I have seen this same thing occur as a TA running wargames for undergraduate students, who connected their academic knowledge of history with game mechanics to inform their in-game decisions. By looking at games for the first time as an academic tool rather than strictly as a source of recreation, I have found a rich new analytic method with which to think about war (and indeed any social science where laboratory experiments are simply not feasible). The challenge of designing a counterinsurgency simulation within the assignment’s restraints has been an educational experience, but I also look forward to continuing my development of the game and seeing how I can further improve it with different parameters such as having more than two players, longer play time, and perhaps limited computer-aided calculations.

Other games to consider

• Anderson, John. (2009). Battle for Fallujah. London: King’s College London.
• Anderson, Michael, Jon Compton, and Joseph Miranda. (2010). Battle for Baghdad. Mission Viejo, CA: MCS Group.
• de Gunzbourg, Antoine. (2008). The Battle of Algiers. London: King’s College London.
• Herrschaft, Tom and Mark H. Walker. (2008). Lock ‘n Load: A Day of Heroes. Henry, VA: Lock  ‘n Load Publishing LLC.
• Miranda, Joseph. (1996). Greenline: Chechnya. Game Publications Group.
• Miranda, Joseph. (1991). Holy War: Afghanistan. Bakersfield, CA: Decision Games.
• Miranda, Joseph. (1993). L.A. Lawless. Bakersfield, CA: Decision Games.
• Ruhnke, Volko. (Forthcoming 2012). Andean Abyss. Hanford, CA: GMT Games.
• Ruhnke, Volko. (2010). Labyrinth: The War on Terror, 2001 – ?. Hanford, CA: GMT Games.
• Train, Brian. (2006). Algeria: The War of Independence 1954-1962. Toronto: Fiery Dragon Productions.
• Train, Brian. (1999). Battle for China. Toronto: Fiery Dragon Productions.
• Train, Brian. (1996). Civil Power. Strategy Gaming Society.
• Walker, Mark H. (2004). Operation Iraqi Freedom. Armchair General.

I took a bit of a break from blogging, but in the interim I’ve completed the first version of my conflict simulation, Fardh al-Qanoon. Take a look at the rules, map, and counter sheet, which are followed by my design notes.

As a reminder, this simulation was designed as part of the Conflict Simulation course for King’s College London’s MA War Studies program. My professor, Dr. Philip Sabin, recently released Simulating War, a book that deals with the very topic of the course.

Please feel free to leave comments or email me at mail@smartwar.org with your thoughts.

The Rules

Fardh Al-Qanoon Rules 1.0 on Scribd

Map & Counter sheet

Download Map on Google Docs
Download Counter sheet on Google Docs

Design notes

Fardh al-Qanoon attempts to simulate the 2007 insurgency and counterinsurgency efforts in Baghdad, a complex conflict with a variety of important elements that must be addressed if a simulation is to be useful in any way. The game has gone through several major design changes, with the key theme in each revision being the paring away of complexity while representing the basic dynamic of the insurgency/counterinsurgency. The initial conception of Fardh al-Qanoon involved limited intelligence, a map of Baghdad down to the neighborhood level, and several player values to keep track of (other than units), including factional support, intelligence, Sunni support for the Coalition, Sunni empowerment, Coalition operational capability, Mahdi Army operational capability, and Coalition casualties. There were clearly too many moving parts for a manual simulation, and certainly not all of them were vital to representing the conflict’s essential elements.

Defining the conflict’s essential elements

The first and foremost element of the simulation is the actors involved. In reality, a variety of parties were involved, and the 2010 multiplayer game Battle for Baghdad takes this into account. The Western-led Coalition, Iraqi government, assorted Shia militias (including the Mahdi Army), Sunni militants, foreign militants, and al-Qaeda in Iraq were all players in the 2007 operation. Given the assignment’s constraints and the narrow scope of my simulation (relative to the nationwide insurgency in Iraq as a whole), I felt it was reasonable to have a two-player simulation with the Coalition and Mahdi Army as live actors and the Sunnis of Baghdad represented through game mechanics. In order to prevent the Sunni element of the simulation from being “gamed” (since the rules for its behavior are laid out and known to each player), I included a flexible element of chance that allows player actions to affect the general trend of Sunni actions without specifically determining the outcome. As the Coalition improves its relationship with the Sunni community (advancing the Relations counter on the Coalition-Sunni Relations track), events like terrorist attacks and Sunni-initiated sectarian violence become less likely through a die roll modifier, but they are still possible. After all, the “Sunni Awakening” did not lead to the complete cessation of Sunni sectarian violence, nor is the Sunni faction a unitary actor. While this basic, three-faction formulation involves a good deal of abstraction (essentially placing all non-Shia combatants under the Sunni game faction), I believe it is sufficient for the limited purposes of this simulation.

The second element is the setting. I started with a map of Baghdad and its 89 neighborhoods (combining some neighborhoods together for the sake of simplicity), but it became quickly clear that this was unfeasible for the assignment’s parameters. I reduced this to 25, then 23 larger neighborhood zones that included Sunni areas within them. Even this turned out to be too complex; the first in-class playtest made clear that there were simply too many options for a player to know what they should do in a given situation. In the end, I decided to pare it down to the nine districts of Baghdad, including Sunni neighborhoods as individual spaces within the districts. I believe this solution, in conjunction with the unit mechanics, works well because it presents the geographical setting in a hierarchical (city → district → Sunni neighborhood) rather than lateral manner. A district-level simulation also removes the issue of having to differentiate between individual neighborhoods/zones, which would be difficult to do because of the lack of quality, open-source information about specific areas within Baghdad. Representing the city on a neighborhood-level without this information (neighborhood demographics, level of urban buildup, lines of communication, landmarks, etc.) would mean not being able to answer certain questions, such as “why is neighborhood ‘x’ a hotspot for violence?” or “why is neighborhood ‘y’ so difficult for the Coalition to secure?” Thus, making the map and game simpler also makes it more useful, since it does not seek to tackle questions it cannot plausibly answer.

The third element is what I would call the “field of play.” In most simulations of more traditional battles or operations, this would be the same as the setting. Two or more military forces have a competition of arms over a geographic area, with a result dictated by maneuver and firepower. The field of play in Fardh al-Qanoon is the population rather than strictly territory (a good example of this is the GMT game Andean Abyss). One player can have a superior combat force in an area, but that alone does not determine whether they are in effective control of it. Each player’s relationship with the population at that moment that decides their capabilities. In the initial version of my simulation, this was represented with factional support tracks for each district measuring each player’s level of support from the population, in addition to a separate track altogether for the Sunni population’s support of the Coalition. Altogether, that comes down to 19 individual variables, which somehow sounded like a good idea once.  While perhaps not more elegant, my simpler solution was to use Mahdi Stacking Limits, secure/unsecure states, and Manpower Requirements to achieve the same end results that the complicated factional support system would have.

Mahdi Stacking Limits are determined by district demographics and urban density. Heavily Shia areas such as those in eastern Baghdad obviously support more militias, but they have fewer targets that militant Shias would be interested in attacking. On the same note, built-up areas support more militias because of population density and the greater ease of disappearing into the urban terrain. Stacking limits also reflect the Mahdi Army’s status as a non-governmental militia that does not exert perfect command and control of the militants that claim allegiance to it. In an ideal simulation, I would have liked to include a mechanic for militias to sometimes act independent of the Mahdi Army player (e.g. attacking the Coalition in a district rather than staying hidden as the player might have decided to do). In the end, I decided the added complexity such a mechanic would involve (and additional play time required) did not justify the value it would provide. Related to stacking limits are the Mahdi Army’s portion of the Combat Results Table and recruitment mechanism. A Mahdi player gains no benefit from using additional militias in combat after the fifth, since it lacks the command structure necessary for a large militant force to work as a unit. A five militia battle in the game represents a widespread battle throughout a district rather than an intense and overwhelming local battle (this is also why larger engagements are always worse for the Coalition than smaller engagements, where its doctrinal and technological advantages make it only marginally less effective with a smaller force). Mahdi recruitment only occurs in Sadr City, the stronghold of the Mahdi Army and Shia population of Baghdad, and districts that experience an incident of violence (representing Shias who decide to take up arms in the aftermath of a sectarian incident). As the Coalition secures districts and pushes the Mahdi Army back, these mechanisms make it naturally more difficult to raise and use militia units in the “rear” areas of the city. On the other hand, negative developments in a district  can be exploited to recruit militias.

Manpower requirements provide the mechanism through which the Coalition interacts with the population. A fully-developed “3x Joint Security Station” counter can only be achieved after at least two months of fully meeting the manpower requirements to secure a district, meaning enough troops to patrol neighborhoods and interact with the population. Ideally, this would mean including a mechanism for ensuring the same units are patrolling a district for a lengthy period of time, since one part of the COIN tactics used in 2007 was that military units would “walk a beat” and become the government’s connection to a local population, similar to a neighborhood’s assigned police officer in a Western city. In addition, the rules currently do not require the manpower requirements to be met in consecutive months, nor is there a provision for JSSs to be “downgraded.” Again, I felt these would add too much complexity to a manual game without providing sufficient simulation value, so they were abstracted: the consequences of an undermanned district are that the area becomes vulnerable to terrorist attack and Sunni violence, and each unit removed from a district represents a greater loss (i.e. removing a battalion before a JSS means losing 1 point of manpower in the district, whereas removing a battalion after having a 3x JSS means losing 3 points of manpower). Manpower requirements also force the Coalition to make difficult choices in its deployments; the districts themselves require a total of 69 points of Manpower to fully secure, while the Sunni neighborhoods require an additional 21 points. By June, the Coalition has 30 points worth of units, which are technically worth 90 if every district has a “3x JSS.” During the early part of the game, the Coalition is faced with several opportunities to choose between focusing on a district as a whole or a Sunni neighborhood, and between focusing on one district versus another. By the end of the game, the Coalition should have more than enough Manpower points to secure the vast majority of the city, but its decisions until that point are crucial, since they determine if the Mahdi Army will have the “Will” necessary to deny the Coalition full victory or even attempt to achieve its own outright victory.

The Units

Ideally, the simulation would use the actual designations of Coalition units deployed in Baghdad between February and November 2007. After compiling a list of the battalions and BCTs involved from order of battle maps by the Long War Journal, I found that there were 39 distinct units (not including Iraqi military or national police) deployed and withdrawn from the city at various points. Given the constraints of the assignment, realistically portraying arrival and departure of individual units would be a cumbersome addition for so little simulation value. Instead, generic battalion and BCT unit counters are used; the arrival and departure of units is abstracted, and instead the ratio and change in troop levels is simulated via the reinforcement mechanism. It should be noted that the unit counters do not represent all government forces in Baghdad, but simply the ones that are available for command. In addition, the Iraqi units are abstracted and are assumed to be operating alongside the commanding Coalition forces.

Reliable information about the Mahdi Army’s manpower and unit formations is sparse; most open source information about them deals with the aftermath of street battles and the results of arrests or raids. Since any calculation of the Mahdi Army’s available manpower is inherently suspect, the simulation is vague in that it represents each militia unit as being a single formation of armed men who work together, while a dispersed militia is the latent ability to raise a formation of armed men quickly. The exact number and armament of these militias is one of the variables of the game that people can debate: how many irregular Shia militants would it take to successfully face off against an American Brigade Combat Team? How many men does it take to clear out a Sunni neighborhood with no Coalition presence? The game defines this in an open-ended way as “1 militia” or “2 militias,” and players can debate what the real-world equivalent of that would have to be.

The Combat Results Table

There are two important outcomes for combat in this simulation: are Mahdi militias successfully driven out of an area, and does the Coalition suffer serious casualties? With that in mind, the Combat Results Table offers a required roll for each event. In order to come up with the values for the table, I read through news stories reporting on clashes between the Coalition and insurgents and reviewed the Department of Defense’s quarterly “Measuring Stability and Security in Iraq” reports for 2007. I came to the conclusion that in small engagements, the Coalition almost always prevailed easily with light casualties, and that widespread resistance was where results were mixed. Since the simulation operates on a district level, the use of more militias in a single combat represents widespread violence. The Mahdi Army can accept lower risk by not using all of its available militias in a combat (and also give itself the opportunity to launch more numerous, smaller attacks), which represents smaller local engagements that the Coalition has an easier time with. In small engagements, the Coalition’s combat chances degrade slightly as its Manpower Points drop; in larger engagements, the degradation is rapid. This gives the Coalition a disincentive against immediately engaging in large, urban street battles with the Mahdi Army to defeat it in one fell sweep. Instead, the best use of its forces is to focus on securing districts, which eventually frees up forces for other missions through the manpower multiplier effect. The Mahdi Army is not in a position to start a large urban battle at the outset of the game either, and must build up its reserve of “Will” by engaging in smaller, local acts of violence against Sunni neighborhoods. By the end of the game, an opportunity for a massive battle in Sadr City or another Shia district arises; the Combat Results Table makes it an iffy proposition for either side, so the decision of whether or not to engage in the battle ultimately rests on the position of each player after their decisions in the early months. For example, if the Coalition neglected the Sunni community earlier in the game, it may find itself needing to secure Sadr City to have enough Victory Points in the end, thus necessitating a large, even more disadvantageous battle there. If the Mahdi Army has saved up enough “Will” by the end to risk trying for an outright victory, it might attempt to hold 2 districts to the end to ensure the Coalition falls short in Victory Points.

The simulation is by no means comprehensive, but I believe the essential elements of the insurgency/counterinsurgency conflict are represented through the game mechanics, and that the game is simple enough without detracting from its purpose. A better simulation (in the sense of providing a more complete model) would go beyond the parameters of the assignment and include more than two players, a larger number of units (to accommodate actual unit designations), and, among other things, a longer allowance for play time. Considering how complicated the initial drafts of the design were, I believe the relative simplicity of the final version is a drastic improvement.

Counter sheet

As you can see, there are a total of 68 counters involved in the simulation (the limit for the project is 100, any my first version came dangerously close to that threshold with 96). Unlike before, there are no civilian counters, nor are the Mahdi Army units hidden (thus removing the “limited intelligence” aspect of the game). Hat tip to Rex Brynen at PAXSims for the suggestion to create black militia counters with white print, which looks very thematically-appropriate.

Updated map

The map has been revamped since the version I posted previously, and the changes look even more stark in comparison to the version I included with my first rules submission, which listed all the neighborhood numbers and track values. First of all, all the tracks have been simplified and each district’s factional support is no longer measured. Since the goal of the operation was to secure Baghdad, I decided to simplify how the game works by measuring Coalition Control rather than its level of support in each of the 9 districts. On the other side, the Mahdi Army’s level of dominance (its presence, reputation, and ability to recruit new fighting members) is measured in lieu of support.

The Mahdi Army has a stacking limit in each district, which is identified by the white number on the green silhouette of armed men. The limit was previously determined by the Mahdi Army’s level of support in each district, with an essentially endless supply of units available up to the district limit. Now, the availability of units is limited by the Mahdi Army’s level of Dominance overall. As the Coalition boxes the Madhi Army in one district at a time (as, ideally, they should be doing in the game if they hope to win), they’ll find themselves facing ever-larger numbers of militia as they reach the core areas of Mahdi control in Sadr City and other Shia-populated districts in east Baghdad.

Under the previous rule system, it was possible to try to lock the Mahdi Army out of a district by securing support in it and lowering Mahdi support, thus limiting the number of units they could deploy to build their support back up. With the factional support system gone, districts are now locked by securing the neighborhoods themselves. The Coalition has 18 units (which do not represent all personnel in Baghdad, but rather the personnel they could use for patrol, raids, and reconstruction at any given time in a month) and there are 25 neighborhood zones. They must, therefore, work on securing neighborhoods by flushing the Mahdi Army out and establishing a permanent physical presence. The establishment of Joint Security Stations allows the Coalition to free up units for other missions while keeping a neighborhood secure, though the Mahdi Army can attempt to break the secure atmosphere by sneaking forces in.

What has not changed since the last version is the Sunni faction, which is still present and affected in two ways: Coalition-Sunni relations, and Sunni empowerment. As the former increases, Sunni acts of violence and terrorist acts become less likely, while the latter determines how many Sunni events occur each month. The Mahdi Army player will want to suppress Sunni empowerment to decrease violence against Shias (since defending against Sunni attacks takes resources away from hampering the Coalition) and make terrorist attacks less likely, while the Coalition player will want to improve its relations with the Sunni community for the same ends (since sectarian violence and terrorist attacks hamper its ability to secure control and increase Mahdi power).

I will post version 0.3.2 of the rules after I receive some feedback from Dr. Sabin on the current version tomorrow, so look for that late on Wednesday or on Thursday.

I’ve previously talked a bit about designing a Fardh al-Qanoon simulation and what the elements of the conflict’s simulation would be. Today I discuss the map, an essential component of any manual simulation.

In Professor Philip Sabin’s Conflict Simulation course, we had to submit an initial draft of our game map well over a month ago. Dr. Sabin, who has designed many simulations, gave us each detailed comments about our initial submissions. I’ll admit, I thought my first draft was rather clever and would make a decent foundation to work from. The passage of time and hours of revision have given me the clarity to realize how silly I was.

The initial map

(Note that many labels and game-related values have not yet been inserted. This is a very rough first draft.)

I believe one of Dr. Sabin’s initial remarks was to describe the map as “science fictiony,” which in retrospect was being too kind. The miasma of colors and shapes do serve specific purposes, but those purposes could be achieved in much subtler, less seizure-inducing ways.

I went with hex-shaped boxes overlaid on a satellite map of Baghdad to represent neighborhoods, since there is such a wide variety of neighborhood shapes and sizes. Some districts have been broken up into more neighborhoods than are listed in this document (pdf), while other districts have had neighborhoods of similar ethnic makeup combined.

As you can see, this version of the map distinguishes between Shia-dominated, Sunni-dominated, and mixed neighborhoods (purple = Shia, yellow = Sunni, white = mixed). The green hex represents none other than the Green Zone, while the red hex represents the FOB at Baghdad International Airport.

There are many tracks littering the map, but the one of primal importance is the collection of tracks on the bottom right. Since this simulation is about a counterinsurgency operation, the most important metric is how much support the two opposing players (representing the Coalition and the Shia faction of Baghdad) can garner in each district.

As I made progress on the rules and reflected on Dr. Sabin’s comments, I ended up making quite a few changes to the map. The most recent draft of Version 2 drastically changes almost every element.

Map Version 2

(Note: once again, many labels and game-related values have not been added yet.)

As you can see, the timescale has been changed from weeks to months, and the game now covers 10 months rather than 12 weeks. This reflects a change in the “turn” mechanism; initially, I planned for the Coalition player’s use of “Command Points” (since done away with) to determine the passage of time. Instead, each round (lasting one month) will have set phases during which each player has a chance to perform their specific tasks.

The number of neighborhoods has been drastically reduced and now stands at 25, with 12 in the western districts and 13 in the eastern. Many districts have 2 or 3 neighborhoods, while Mansour and Rashid each have 4. I felt this was a fairly important disparity to include because while the districts with 4 neighborhoods gives the Shia player more “space” to hide and maneuver, they have to balance that opportunity against maintaining their influence in districts with fewer neighborhoods, where the population density gives the Coalition the chance to make a big impact.

The overall concept with the district influence tracks remains the same. There are 7 “levels” of support in each district, and players reap benefits from building up support depending on which faction they are playing. The “Intelligence” tracks have been combined into one, and two separate counters will track western and eastern district intelligence (I don’t know why it didn’t occur to me to do this in the first place; two intelligence tracks was such a waste of space).

Two new tracks are the Coalition-Sunni Relation track, which is exactly what it sounds like and has the effect of changing what kind of Sunni actions take place during the game (better relations means fewer terrorist attacks and sectarian operations, and more community-building efforts and cooperation with Coalition forces). The Sunni Empowerment track determines how many Sunni actions take place each month, and this track can be negatively affected by the Shia player executing sectarian operations to drive Sunnis out of their neighborhoods.

The last new element is a pair of mission hex sections, one for western and one for eastern Baghdad. Now that I’m using “phases” during rounds, the Coalition will have to decide how to use their forces in a given month by assigning them different missions at the beginning. Having a high level of support in a district, however, gives the Coalition the flexibility to change missions on the fly. If, for example, the Coalition player sees that the Mahdi Army is making a play for a district and massing militia forces, he can reassign units in the “Build” hex to “Raid” and attack the Mahdi Army.

Rules, rules, rules (and counters)

Of course, none of this matters without a solid system of rules, and there are still map counters to consider. I’ll be done with a draft of the rules and the counter designs by Thursday, since that’s when all of this needs to be submitted for class. Look forward to those later this week.

I finished the first draft of my simulation rules for Operation Fardh al-Qanoon (aka the Baghdad Security Plan) more than a month ago, which I no doubt thought was exceedingly clever and put me ahead of the curve. After putting together a map and a draft of a historical analysis (as required by my Conflict Simulation course), I realized I needed to scrap everything and go back to the drawing board.

Unlike many of the other scenarios chosen by my fellow students, Operation Fardh al-Qanoon is not primarily about the combat. Combat is an element, but one that I focused too much on in my first draft rules. In its initial form, the simulation devolved into a game of whack-a-mole, with Sunni and Shia insurgents, al-Qaeda terrorists, and foreign jihadists playing the role of dastardly moles. Coalition and Iraqi troops obviously had a lot more on their hands than chasing and killing militants once the operation began in February 2007. No COIN simulation can be complete without a robust portrayal of the political dimension, and the political developments in Baghdad in 2007 were certainly interesting.

Players

The first element of a simulation is the actors involved. A classmate is designing a simulation of the 2008 terror attack in Mumbai as a solitaire game, which I think works well for that scenario. The Baghdad Security Plan practically demands an intelligent, responsive adversary, but the constraints of the assignment mean there can only be up to 2 players. I’m scrapping my plan to have an “OPFOR” player representing all elements working against the Coalition in Baghdad, as the only way to make it work would be to have very restrictive rules that force the opposition player to use his assets in a historically proper way (i.e. not having Shia militias cooperating with al-Qaeda operatives against Sunni civilians).

There will be a Coalition player as usual, but the opposition player will represent the Shia insurgency, specifically Muqtada al-Sadr’s Mahdi Army. The Sunni faction, al-Qaeda, and foreign extremists will be handled by the “game system” through rules, so they will still be represented, but there won’t be an intelligent, actively-plotting personality behind them. A problem with this approach is that it makes it possible to “game” these factions, since the rules for their behavior are known to both players.

Simulation Dynamics

Like any counterinsurgency scenario, the Baghdad Security Plan was a contest between opposing political forces in an arena composed of people. At the same time, it was a battle of order vs. chaos. Regardless of who was “winning” faction-wise, high levels of violence and instability meant everyone lost. An ideal simulation would measure both factional support and stability (which takes security into account) within the 9 administrative districts of Baghdad, but given the constraints on complexity for this particular simulation, I will be measuring factional support only.

All other elements of the simulation, namely combat and politics, can be implemented through a relationship with the factional support system. Each player’s goal is to ultimately improve their relative standing in the city, and the other side of this coin is forcing the other player to take actions that make him lose support.

Unit Tactics

Uncovering a weapons cache (Source: Wikimedia)

A Coalition player has the basic options of patrolling neighborhoods to identify and remove militant forces, raiding insurgents & militants to remove leadership and material resources (when their locations are known), and building (a generic term for all reconstruction, development, and community engagement activities). You’ll notice combat is not listed as a primary activity, but combat does occur when a Coalition unit tries to clear an area of militants during a patrol action. Targeted raids are ideally meant to take out militant leadership or resources without a fair fight, but combat action requiring resolution can occur during a raid as well. The other option for combat occurs when the Shia militia player decides to engage the Coalition or another faction’s forces.

While the Shia player does not have the resources or highly-trained personnel of the Coalition, they have a larger base of human support to draw from, giving them men, materiel, and space for maneuver. The Coalition operates from an inherently vulnerable position, while the Shia player’s fighters, leaders, and equipment has a city of millions to hide in. The Shia player must decide whether to engage Coalition forces or Sunni & foreign extremists (or all of them at the same time), where to risk militant forces, and how best to maintain their factional support and increase it.

Say, for example, that the Shia player decides to shore up its support in a district and attract new recruits by using violence to drive Sunni residents out of a mixed neighborhood. This increases the chance of a Sunni reprisal attack against Shia civilians; the Shia player can use its fighters to protect Shia civilians and attack Sunni militants, but will then expose themselves to Coalition interdiction. By doing nothing, they give the Coalition an opportunity to fill the gap and gain factional support. Both players have an array of options at their disposal, and they can be as proactive or passive as they want.

The field of play: human terrain

The simulation takes place in the 9 administrative districts of Baghdad, offering a rather unique terrain. With most wargames, you would differentiate between plains, forest, and hills, but Fardh al-Qanoon will have to rely on human terrain. While there likely will be a differentiation between dense and less dense urban terrain on the game map, the most important feature of the map will be the demographic makeup of each represented neighborhood. I’m working off of the administrative and neighborhood districts outlined in this document (PDF), but many of the neighborhoods will have to be combined to form one “area” on the map for the sake of simplicity.

A map showing the Coalition Order of Battle in Baghdad in February 2007, with neighborhood demographics (Source: Long War Journal)

Some neighborhoods are largely populated by Sunni Muslims, others by Shia Muslims, and others still by a mixture of both. The Shia player aims to shore up their support in Shia and mixed neighborhoods, while trying to drive Sunnis out and prevent Sunni extremists from harming Shia civilians or holy sites. The Coalition player has to play a delicate balancing act between Sunnis and Shias, protecting both without favoring one over the other or becoming embroiled in open war with the Sunni or Shia factions. Al-Qaeda and foreign extremists will also require local support in Sunni or mixed neighborhoods, which they can lose if Sunnis in a district come to view the Coalition as a better option.

A measure of reality

The game “engine” has to be flexible enough to allow the operation to develop in any number of ways, the vast majority of which obviously did not happen in reality. The conditions of game victory are what will provide the incentives necessary for players to act in historically analogous ways without binding them to that course. The Shia player may decide not to temporarily stand down in February 2007 as US “surge” units start arriving in Baghdad (as Muqtada al-Sadr did in real life), and instead take their chances bogging Coalition forces down in ugly street fighting before all the new troops can arrive.

The game engine, under the limited influence of chance, will give the risk-taking Shia player a certain set of outcomes, most of which would certainly seem terrible for him in the Coalition player’s eyes. Such a move would certainly involve a heavy loss of forces for the Shia player, but if it manages to inflict massive Coalition casualties (and thus make the operation an immediate political failure) or forces the Coalition to abandon protecting the population in favor of fighting militants, then the Shia player can achieve its victory objectives.

My goal is to design a simulation that will reliably answer some “what if?” questions without being prescriptive. A COIN simulation is inherently difficult because COIN and anti-COIN advocates disagree on the underlying math behind a COIN operation, and thus they disagree on what the results would be. Was the surge successful because of an influx of US troops and the tactics they used, or because of the Sunni Awakening?

The questions behind a Fardh al-Qanoon simulation should not be about the efficacy of counterinsurgency. They should ask about the relationship between the various factions at play, about violence and stability, and about political support. By modeling the underlying dynamics, this simulation will hopefully be “valid” without having to make an initial assumption about whether COIN works or not.

Considering my most advanced experience with manual war games thus far has been Axis & Allies, I have quite a bit to learn about simulating conflict before I can even think about solid game mechanics. There is also the issue of navigating between COIN strategy & tactics (or as Gentile put it, a strategy of tactics) and actual results of the practice of COIN in Baghdad. If the Coalition player were to do everything according to historical guidelines, ideally the end result should reflect history.

The problem with getting the simulation results to match up with reality is creating the baseline for results in the first place. One of the perks of the modern age is the abundance of data about anything and everything you can imagine. War, however, tends to throw a truck-sized wrench into the machinery. Even military data (much most almost all of which is not generally available) is not guaranteed to be accurate. Secondary sources are even less reliable, often filled with conflicting accounts or large “guess” ranges (such as for civilian casualty counts). I’ve decided to measure Coalition KIA casualties as part of the simulation, but not anti-Coalition militant body counts. For one thing, enemy body counts are not as important when you’re dealing with a COIN operation, as the end goal is to secure the population rather than destroy the enemy.

Civilian deaths are very much relevant to a counterinsurgency simulation, but there are certain constraints in the class-assigned project that make too much complexity undesirable. Instead, civilian deaths (whether by Coalition or anti-Coalition hands) will be reflected generally by a negative change in stability. Since the Coalition goal is to “stabilize” all 9 districts of Baghdad, civilian deaths are still punitive, if not strictly measured.

Choices, choices

The asymmetry between Coalition forces and the irregular enemies they face means that engagements will not be of the same attritional nature as conflicts between state militaries. Professor Philip Sabin’s World War 2 and Eastern Front games (both available for free download here) were basically my first foray into academic (as opposed to hobby) manual war games, and they nicely reflected the nature of attritional conflict. The choices faced by Coalition and anti-Coalition players will have to be a bit different in my simulation of Fardh al-Qanoon.

The Coalition forces will have to balance the risk to friendly forces vs. stopping violence within a certain timeframe. There will also be the balance between killing militant forces and preventing civilian deaths; since engagements will not always occur on the Coalition’s terms, this is an especially important factor.

The anti-Coalition side (which will include everything from Sunni/Shi’ite insurgents to al-Qaeda operators) is too diverse a group to have institutional goals in the game. No player will be able to objectively separate assets from different groups in their mind; thus you would end up with strange situations like Shi’ite insurgents supporting foreign Sunni fighters to destabilize a Shi’ite neighborhood. Instead, the anti-Coalition player’s goal is simply to make sure the Coalition fails to secure the city by the end of the game. Game rules will ensure that the proper militant forces will be used in particular situations to destabilize the city (e.g. Shi’ite insurgents can’t be used against Sadr City, while al-Qaeda and Sunni insurgents can). The choice for the anti-Coalition player thus becomes how best to utilize his limited militant resources and escape detection by the US and Iraqi troops.

The anti-Coalition player can also “win” by inflicting massive KIA casualties against Coalition (non-Iraqi) troops, so far above the historical levels that the operation is deemed a political failure. President Bush was already experiencing significant pressure from a Democrat-controlled Congress; if the “surge” failed to stem the tide of violence and resulted in increased US casualties, domestic pressure would force a change in policy.