(Read parts one and two)

In this post, I’m going to discuss what is perhaps the most essential element of an insurgency/counterinsurgency wargame: political mechanics.

Most wargames understandably focus on the combat aspect of a conflict. After all, recreating an epic naval battle or demolishing your enemy’s tank formations in a pocket is a lot more interesting than simulating negotiations on a peace treaty or writing a constitution for a new government. All wars are about politics, but simulating the political aspect of a war would not add value to the vast majority of wargames. Insurgency games are some of the few exceptions.

Before detailing the political mechanics of the insurgency wargame, I’d first like to talk about a game that mixes politics and war just right. Crusader Kings 2, a medieval grand strategy game by Paradox Interactive, simulates feudal dynasties from 1066 (just before the Norman invasion of England) to the mid-1400s. Unlike Paradox’s other strategy games, and indeed unlike most other grand strategy games in general, Crusader Kings 2 focuses on character-based play. Landed nobles living in a feudal system scheme to usurp each other’s titles, assassinate each other, curry favor with the Pope, carve out a Christian kingdom in the Holy Land, and a wide variety of other era-appropriate activities. Players set their own goals, whether it’s to mend the schism between the Christian churches, get elected as the new Holy Roman Emperor, or simply keep their rivals from amassing too much power.

Part of Crusader Kings 2’s appeal is that players have a variety of methods available to achieve a political goal. (For an example, see this post by a Reddit user describing how he managed to get an Irish Catholic to ascend the Byzantine throne without participating in a single battle.) It is often easier to expand through strategic inheritance rather than war, which requires either levies (which make vassals unhappy) or mercenaries (which are expensive). A powerful rival noble that would thump you in a war can be taken down a notch by convincing the Pope to excommunicate him, or you can simply plot with his unhappy courtiers to assassinate him. The simulation of dynasties and extensive family trees adds another layer of possible intrigue; strict inheritance laws mean that you can predict a line of succession and identify possible troublemaking claimants to a title.

A screenshot from Crusader Kings 2. By simulating noble dynasties, the game can represent politics in historically-appropriate ways, such as the formation of alliances through marriage.

Crusader Kings 2 translates the political and military dimensions of medieval Europe into a system of mechanics that allow for a great deal of experimentation while recreating the constraints and pressures that historical nobles faced. This is quite a feat, keeping in mind that as a commercial product it strives for recreational value rather than historical or simulation value.

The tabletop format restricts how detailed the political-military system can be in an insurgency wargame. This is not necessarily a problem, and can in fact be a benefit. Abstracting elements that are not crucial to the simulation (for example, in a wargame about insurgency, you may not be interested in the intricacies of the supply and logistics chain that sustains a military effort) makes the entire system more accessible and easier to balance. To use Crusader Kings 2 as an example again, the balance between Christian and Muslim realms often changes dramatically after an update patch because of the sheer number of variables at play. In the initial release, Andalusian Muslims were more likely than not to overrun the Spanish kingdoms; in the current version, the Reconquista can easily be accomplished more than 200 years ahead of time. It’s easier to design a game that conforms to realistic processes and constraints if it has fewer moving parts. On the other hand, simpler games are at greater risk of having design predetermine their outcome.

And with that ramblingly long preamble out of the way, let me finally explain the fairly simple political mechanics of the insurgency wargame. The current Afghan war will be used as an example for illustrative purposes.

Political Capital

A variety of resource concepts are abstracted into the basic unit of “political capital.” This represents different things to different factions; thus, for example, the social capital of strong tribal relationships can locally match the effect of the incredible financial capital of a foreign, first world superpower. Political capital (PC) is required for a variety of actions, from training and deploying troops to swaying public opinion to win an election or buy legitimacy. Political capital can be traded between factions as a bargaining chip, though there are some caveats, as not all political capital is equal.

PC given by a foreign occupier, such as the US in Afghanistan, represents money, weapons, and advisory assistance that can be universally redistributed and utilized. Other player factions, representing tribal or ethnic factions, are more limited in their political power. A southern Pashtun faction has less influence with a northern Uzbek faction than it does with eastern Pashtuns, which in turn it is less influential with than its own southern Pashtuns. A player faction that actually governs, such as the Government of Afghanistan, has PC with more universal application than local factions, making the goal of holding political power an attractive one. If the government is not popular or is not considered legitimate, its PC is devalued (i.e. more is required to accomplish a task), making it a less attractive bargaining chip.

For simplicity, PC is represented through token pieces that match the colors of factions. Thus, a Red player (local tribal faction) that possesses a great deal of White (Government of Afghanistan) political capital has a strong interest in making sure the White player’s government prospers. A Yellow player (local tribal faction) that is being supplied with ample PC from the Black (foreign occupier) player is incentivized to help Black fulfill its political goals, but White has no direct power over him. If Green is capable of generating more than enough of its own PC because of its incredible popularity in a given territory, it has little incentive to accept PC from White or Black—unless it has ambitions to national power beyond its local territory, where its own PC holds little sway.

The relationships between factions, their corresponding populations, and their political goals determine player strategies; political capital is the medium through which those elements interact.

The Focus Matrix

The most important distinction between player factions is their political goal, as everything else essentially stems from that desired outcome. Players can move toward their goal in a variety of ways, but the relative simplicity of a game compared to the real world means a significant loss of cues and signaling to be interpreted by other parties, especially if each “turn” represents a large period of time. There must also be some degree of representation of the strengths and drawbacks of factions. After all, insurgency/counter-insurgency is not chess—the players are not identical. Insurgents are not as well-organized as government forces, but they are more effective than if they were. Government forces generally have better equipment and training, but have a harder time building support in a contested area.

The solution is the focus matrix, which allows players to organizationally orient themselves toward subgroups of the population to facilitate generating more political capital. There are two areas of focus: the faction divide (this represents whatever classification is used to divide the population in the scenario, whether economic class, ethnicity, tribal affiliation, religion, etc.) and the urban/rural divide. Within the two areas of focus, each level represents a collection of efforts including propaganda and outreach, recruitment, patronage, and other activities. Among other things, levels of focus affect a player’s ability to generate political capital from the group in question.

Players can spend political capital to attain more levels of focus, going for either a deep focus (building up levels in one or two factions and in either urban or rural focus) or a wide focus (building up levels equally among factions and in both urban and rural options). Deeper focus on one group comes with the drawback of negatively modifying the player’s orientation toward other groups. Territorially-limited groups will find it easy to focus deeply on their own population and build-up type (urban or rural), but a government player without an abundance of political capital will have to choose their focus carefully and make the right agreements with other players.

In my previous post, I discussed in general terms my objective in designing a new manual insurgency wargame and some of the basic concepts it should seek to simulate. I know this is a very belated follow-up, but in the meantime I have been learning Python and considering turning this project into a computer simulation. That would be a daunting task, so for now I am sticking with the manual format with the aim of turning it digital down the line.

The wargame can be divided into roughly two realms: the environment and the players.  In this post, I’ll discuss the environment and mechanics involving it.

The environment is divided into regions at a level that make sense for the design of the scenario, whether that be villages and neighborhoods, cities and towns, or entire provinces. Choosing this lowest level of representation is a balance between complexity and utility; representing politics or combat at the provincial level involves far more abstraction and less choice than representing them at the local village level.

The main feature of the environment is terrain, and for simplicity each region contains a single type of dominant terrain. In this simulation, terrain will affect the results of combat events and the amount of effort required for a faction to gain support in a region. Each region also possesses a certain level of urban build-up, from none (uninhabited) to extremely dense. As expected, this affects a regional population’s size and demographics, as urban populations are more likely to contain significant diversity. This allows the possibility of an urban/rural divide among insurgent or loyalist factions.

Regions possess an objective value based on factors such as economic output, fertile land, presence of resources, etc. The subjective value of a region depends on a particular faction’s objectives and requirements, which are not necessarily known to other factions in the game. For example, a nationalist insurgent faction may have no interest in controlling a prosperous region held by an opposing faction because it contains an incompatible population type and is not part of the historical nation the faction seeks to establish. (These factors are taken into account in the player’s victory conditions, which are on a card given to the player at the beginning of the game and cannot be shown to anyone.) Instead, the faction might use violence to suppress the region’s economic value to the enemy without ever seeking a local base of support. Discerning the objectives of other factions is crucial to the political aspect of the game, especially negotiations with enemies and allies. Insurgencies do not happen in a vacuum, so a scenario should include historical context that suggests each faction’s likely end-state goals and territorial aspirations, but the specific victory conditions are confirmed only to the player. In claiming a set of objectives to other players, a player can tell the truth or bluff, but other players can never see his card to confirm the truth.

Control over a region is determined at two levels. First, there are governing structures that determine what faction exerts official control of the region. Just as a military force is sustained by supply, governing structures require support of the local population to function, though this does not necessarily mean the government in question is representative in nature. Support can also be projected from neighboring regions, though local support or lack thereof naturally has a stronger influence. Structures by competing factions can exist in the same region, making official control impossible. Alternatively, structures by cooperating factions allow for split control based on whatever agreement the factions make (e.g. a federal system).

Second, a region can be effectively controlled with a military force presence that suppresses existing governing structures. The disadvantage, of course, is that militarily controlling a region requires constant presence, along with the usurpation of governing structures and the complete removal of opposition forces. On the other hand, military control is not dependent on the local population’s disposition, making it ideal for a powerful but foreign player attempting to restore or install a local ally into place.

Control of zones will play an important part in combat, as changes in control can affect “momentum,” a mechanic that will allow factions to capitalize their successes or exploit enemy missteps. Tactical retreat from a region may make sense to preserve a limited military force, but it will affect public perception about the conflict (especially for a foreign faction) and boost enemy confidence.

In the next post, I’ll discuss political mechanics.

Ever since completing my first wargame, Fardh al-Qanoon, I’ve been eager to tackle the topic of insurgency once again with a more complex manual simulation. Fardh al-Qanoon simulated a local counter-insurgency campaign in a single city during a relatively short period of time, and while I hope it may play an educational role to someone trying to understand the Baghdad Security Plan, such a game is not terribly useful for analyzing counter-insurgency at the national level, which inevitably stretches over a substantial period of time.

Fardh al-Qanoon also lacked a critical element: politics. Insurgencies, like all wars, are about political objectives. The short time period of the game meant that the insurgent player’s victory conditions could not be based on actual political objectives of the real insurgents, since there was no way for the player to achieve them within the game’s timeframe. The Coalition player could not co-opt the insurgency through a mix of kinetic action and political accommodation, limiting the toolset the counter-insurgent had at his disposal to win.

The map for “Fardh al-Qanoon,” a simulation of the Baghdad Security Plan.

The two-player game also meant the non-player “third” faction (Iraqi Sunnis) could not be negotiated with. Rules governing such negotiations could have attempted to emulate a live human player representing the Sunnis, but then the problem of “gaming” a non-player faction arises. Randomization mitigates this somewhat, but the more randomness you introduce into a game, the less outcomes depend on player decisions, and thus the less useful the game is for education or analysis. There is, of course, a hard limit to how much you can accurately simulate in a wargame before it becomes far too complex and cumbersome, so carefully-designed randomization systems that abstract complexity can definitely be a benefit rather than a drawback.

These issues lingered in the back of my mind as I mulled over the idea of a new insurgency simulation. How should regular and guerrilla forces be represented? How should different quality forces be accounted for? How does a player decide when to use political methods or violence to achieve their objectives? How do players decide when to cooperate or compete with other players?

But most importantly of all: how can the simulation be universally applicable?

Developing a system rather than a scenario is the most important task of this project. Philip Sabin, my Conflict Simulation professor at King’s College London, wrote a book called Lost Battles that creates a model for simulating ancient battles, both Greek and Roman. Ancient Greeks, Romans, and their various opponents were obviously not uniform entities on the battlefield, but careful design allows the Lost Battles system to account for these differences in a relatively simple ruleset. Most importantly, it allows players to gain valuable insights from recreating battles ranging from the Battle of the Granicus to Cannae.

The objective of designing a “universal” insurgency wargame is essentially to create a system that can be tweaked to accurately recreate insurgent conflicts across time and space, though for simplicity it will focus on the post-1950s period. Afghanistan and Iraq, perhaps the two most popular contemporary examples, are obviously very different in culture, history, and terrain; the specific circumstances of each country constitute a “scenario,” and the simulation will serve as the “system” for the scenarios to be designed in. (This is similar to how Volko Ruhnke and Brian Train’s Afghanistan game, A Distant Plain, uses the same system as Ruhnke’s Andean Abyss, or how many computer strategy games come with map editors that allow you to create your own scenarios.)

Basic concepts

The basic concepts of the simulation should be general enough to apply to any insurgent conflict, with the state of their attendant variables distinguishing one specific conflict from another and creating the unique scenario that each insurgent conflict ultimately represents.

First, there is the population. The representation of the population in the simulation should give us an idea of its composition relative to the various factions represented in the simulation and the level of support of the population for factions. The population will be divided in zones of terrain that can represent administrative areas, historical regions, or whatever division of territory makes the most sense for a scenario. Each zone’s primary terrain type is an important factor in determining things like a population’s support for factions (e.g. remote, mountainous terrain is more difficult for a central government to extend its power to), results of military operations, etc.

Military force and political capital represent the two dimensions of a faction’s ability to shape the environment. Military forces can be conventional or irregular, with the former requiring greater resources to operate and the latter possessing some of the unique advantages guerrillas enjoy. The ultimate goal of a counterinsurgent or a competing political insurgency is to establish a governing presence in the disputed territory backed by local support.

In the next post, I’ll begin discussing mechanics of the system and how they might simulate the concepts outlined here

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’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.

After looking at a few historical battles I found interesting, I thought I was set on recreating the Battle of Ayn Jalut, a turning point in the Mongol conquests where Mamluk forces successfully defeated the horde. Of particular interest in the battle was the similar number of different quality of cavalry on each side.

Mongols were famous for their cavalry, but their steppe horses were smaller ponies compared to the larger horses used by Mamluks, mostly of Arabian stock. Mongol warriors also traveled with 6 to 7 horses, a number easily sustained in Central Asia, giving them a tactical advantage as they could switch worn-out horses for fresh ones. The Mamluks obviously did not have as much space to maintain so many horses, so their cavalry fought with 1 to 2 each.

Another popular image of the Mongols is that of the horse archer, sniping down opponents as they blaze past them. While Mongols were indeed accomplished horse archers, the quality of their weaponry (typically fashioned by the users themselves) limited the range at which they were effective. Mamluks, being professional soldiers, extensively practiced archery and were effective at longer range. With fewer horses, however, they could not match the maneuver potential of the Mongols.

After doing some more reading about the Battle of Ayn Jalut (of which there is surprisingly little material available, considering its importance), I got quite excited about the prospect of designing a simulation to recreate it. At the end of the day, it simply wouldn’t work, since the crux of the Mamluk victory lay in pulling off a tactical ambush that the Mongol commander obligingly walked into. Anybody who has read about the battle or even played the simulation at least once will know not to fall for the Mamluk’s feinted attack and retreat, thereby throwing the Mamluk’s odds of victory way out of favor.

Instead, I managed to latch onto a much more modern scenario. First, I should explain my long-standing desire to create an “updated” chess. I’ve always thought that chess was a bit of a silly game because pitting evenly-matched forces against each other without fog of war makes it impossible to use any of the tactics or tricks one would find useful outside of an 8×8 square board. A modern version of chess should reflect asymmetry (because after all, who seeks to meet the enemy when the odds are not in his favor?), uncertainty (the fog of war, or friction as Clausewitz liked to call it), and flexibility.

With that in mind, I’ve decided to model Operation Imposing Law, aka Operation Law & Order, Operation Fardh al-Qanoon, or simply the Baghdad Security Plan. General Petraeus’ revised counterinsurgency strategy was just beginning to make a positive impact in Iraq as I developed an interest in the serious, academic study of war, so I have a bit of a fascination with COIN. Plus, I figure creating a manual simulation of a COIN operation in Baghdad would be a good challenge. The constraints around our simulation assignment make abstraction and simplicity a necessity to some degree, so the idea is to capture the basic elements of conducting counterinsurgency and obstacles to succeeding in that effort. While I haven’t thought through the entire scenario yet, I’m fairly certain the main element of the simulation will involve the Coalition player being blind to the anti-Coalition player’s fighting units, which will all look like civilians to him.

Creating a manual, tabletop COIN simulation in a setting like Baghdad will certainly be an interesting process. Much like Iraq, failure is not an option. There’s a Master’s degree hanging in the balance!