A Game Theory Analysis
by Chief Warrant Officer 3 Chad Machiela
Download the full article: Gun Control in Counterinsurgency
Throughout 2006, Multi-National Corps-Iraq (MNC-I) coalition forces and the farmers of the al Jazeera Desert of Iraq struggled to cooperate while pursuing separate goals. Consistently, the desire by MNC-I to impose populace and resource control measures to limit use of the area by insurgents clashed with the needs of the populace to survive and care for their families. One issue in particular resulted in the repeated arrest of farmers who intended no crime but to protect their families and left the farmers with no choice but to support the insurgents—the coalition's policy for gun control. Game theory provides commanders and policy officials a methodology to analyze the options available to disparate actors within a competitive situation or conflict, to predict likely adversary and population reaction to plans or policy, and to help develop courses of action beneficial to all.
The al Jazeera Desert is a sparsely populated region, bordered by Lake Thar Thar to the west and Main Supply Route (MSR) Tampa between Samarra and Tikrit to the east. Because of the coalition's top-down method of controlling Iraq, this rural area hosted no coalition forces. Coalition patrols instead focused on protecting MSR Tampa and the pipeline between the population centers of Samarra and Tikrit. Because the area was without cell coverage, residents could not call on security forces for assistance when threatened by insurgents or criminals, providing insurgent forces an ideal area for hiding, training, and reconstituting before traveling back into the larger population centers to resume direct conflict. Criminals flocked into the desert to remain out of the reach of government forces and prey upon the isolated farms.
In 2006 the coalition's populace and resources control measure for management of privately owned weapons was to allow each Iraqi household to maintain one AK-47 or AK-74, with two magazines with 60 rounds of ammunition. Ostensibly, this would allow the family to protect itself against local criminals and insurgents, while limiting the number of armed individuals who might oppose the forces of the coalition and the Government of Iraq. Instead this policy ensured that local residents were left helpless to resist the insurgents, who cared little about limits on gun ownership and generally travelled in armed groups of four to twenty.
Download the full article: Gun Control in Counterinsurgency
CW3 Chad Machiela is a Special Forces warrant officer assigned to 1st Special Forces Group (Airborne), Joint Base Lewis McChord. He holds a M.S. in Defense Analysis from the Naval Postgraduate School and a B.A. in Public Law from Western Michigan University. The opinions expressed here are the author's own and do not reflect the views of the Department of Defense or the U.S. Government.
About the Author(s)
Comments
Mike,
Your point is completely true point that game theory interaction plays out in a drastically simplified model that cannot possibly capture all the complexities of the environment that COIN occurs in. Of course, that is why we create models, to take environments that are too complex to understand as a whole and identify primary levers that we can use to begin to test our understanding. The value of the model is not based on the number of calculations it spans, but on how well it works. To quote George Box, "All models are wrong, some are useful." Game theory does not require actors be be rational to model, provided you understand enough of your opponent to order his priorities correctly. His choices need not be "rational" to predict, provided you have learned enough about him to be reasonably sure of what his ordered preferences are--however crazy or alien they may be to us. The parables and quotations listed above (and other published works and organizations like HTS) serve to offer clues about what those ordered preferences might be in future games.
Thanks for the input, I look forward to seeing your monograph.
Chad
Well, Jesus certainly tiptoed thru the heresy/treason minefield with his 'render unto Ceasar' construct, but I always thought that there was room for a 'He'll get what he's got coming interpretation'.
Re the Eastern mystic Stalin, we do well to recognize that his model remains much in evidence on several continents today, so there must be functionality for some in it.
Re his imitator Saddam, I'm reminded of the arabic saying, which I'll paraphrase:
'Twenty years of a lousy Caliph is better than one night without one.'
I look forward to hearing more on game theory, Mike. I never got past "The Prison Rat' game. Instructive, but a tough school for teaching the Golden Rule.
WW,
"But I digress ;)."
I suppose that you missed the "western" adjective in western judeo-christian model in your digression of modern day politics spawned from 2000 year history and political socio-economic evolution.
I'd ask you to consider one CURRENT eastern Russian adage that holds true and easily applied to an Iraq state where Saddam Hussein worshiped Stalin,
"I rather my own house burn down rather than see my neighbor have a bigger home."
This mentality defies western mentality and is certainly irrational in a Rational Actor Model (RAM) that defines the necessary constructs of mutual reciprocity, respect, and equal competition for capitalism.
Moreover, the Rabbi Jesus, the Mesiah in my book, declared that one should turn the other cheek to such irreconciliable and non-cooperative fools. A wise old first sergeant versed it in a different voice,
"Sir, do you know why you should not wrestle with a pig in the mud? The pig likes to get dirty."
Best,
Mike
RE 'equations based off western judeo-christian thought.'
Both Judaism and Rabbi Jesus' gospel are Eastern religions, as described in the Bible. Jesus was denounced by propertied local officials, tortured and executed by a Western military occupation, as a rebel. Decades of unrest and insurgency eventually coalesced to defeat occupation and counterattack, before falling prey to an epic Roman 'peace.'
I grant you that most Americans are in denial about the Arabic dialect spoken in 1st century Palestine, and that Jesus' resistance network looked more like Arafat's relatives than Bibi's. Rewriting history is one of our Western talents, elevated to an official function of government.
Not even Seldon's Psycho-History game theory could have anticipated the full revival of the Hebrew language by Europeans in Palestine, more than 2000 years after it had declined to mere ceremonial use among Levantine semites and their Greek and Persian cousins.
But I digress ;)
WW,
"I wasn't aware that game theory assumed both parties played by the same rules. I thought it sought to explain why party B might rationally behave differently than party A."
You're correct, but the rules of the game are defined by the Rational Actor Model through expected value equations based off western, judeo-christian thought. If you look at earlier 1990's models, that's where applied mathematicians get confused with the hijacking or terrorist models. Sorry for no direct references, I'm on vacation right now with no books.
Shortly, I'll be publishing a monograph on "Small Wars and the Theory of Games." Hopefully, I'll provide a model that may be useful for us all. I've been studying this stuff for twelve years and developed my model two years ago. And please, be sure to step in and correct any shortcomings!!! I really liked the rest of your post.
Mike
I wasn't aware that game theory assumed both parties played by the same rules. I thought it sought to explain why party B might rationally behave differently than party A. For instance, why guerillas don't wear uniforms like gov't troops, or why populations might support insurgents (or gov'ts) that are robbing them.
Re Iraqi/Afghan response to attacks in their community.
One of the things that strikes me when viewing video at the scene of a bombing or Israeli missile strike, is that Arab men seem to be running to the emergency, to render assistance to wounded countrymen. This behavior was part of the notorious wiki-leak of the helo gunship footage, where a driver risked his daughter to stop and aid the blasted militiamen, and got blasted in turn.
Limiting guns and ammunition clips to 1/house still leaves them lying around for an insurgent to invade a home, grab the family AK, deliver suppressive fire against soldiers 'attacking into the threat', then retreat unarmed and do it again from a different house. The family AK, used in isolation, might just piss off local bad guys, but still provide insurgents with pre-positioned firepower, for serial use, in a 'collective defense' culture. And get a lot of civilian homes assaulted by US troops who only know they are receiving fire, without understanding 'why here, why now.'
(The VC used a collective defense doctrine, distributing worn weapons to unreliable militiamen, who were instructed to put poorly aimed fire towards US helos, when they came close.)
A rural farm in Anbar or Kunar is a somewhat different story, since neighbors aren't near enough to render common defense. What struck me about one scenario, a dozen men showing up to steal the car, is that motor transport is involved in the raid, and is also the target.
There are far fewer cars than guns in an insurgency. A car or pickup is what leverages insurgents, enables the Taliban to concentrate and overwhelm police, co-opt or recruit an entire village.
Every gov't in the world puts strict limits on driving privileges, which gives terrorists both cavalry and a mobile car-bomb.
Any motorbike and tractor in an active AO out to be named and tracked, matched against the usual occupants. Mag-stripe or bar-code is perfect for logging and collating data points, sifting for an agent of terror. A large external bar-code or smart pass can be read without the occupants even knowing they've been scanned. A road-block crew should have identified a cars occupants, the driver at least, before approaching to search it. Just like a patrolman in the US, who runs the plates before cautiously approaching the driver.
In a war zone, a truck that raises a flag for being out of area, or being driven by an unfamiliar/unlicensed driver is the one to look out for, without any need for firing warning shots at local families, who want to exit an ambush site, just like any HMV driver. Automatically collecting mobility data won't stop attacks, but it can reveal patterns and networks, without going into folks bedrooms.
Just some thoughts.
Chad states,
"My real motivation for writing the article, which I never spelled out, is that game theory is indeed useful for working through strategic interaction of potential lines of operation, but more important to groundpounders and SOF professionals is that it offers a method to explain what we understand intuitively to skeptical audiences."
So true, and I think that Chad drives his point home with his essay. With that said, I'm gonna play Debbie Downer and offer some advice for those that would seek to study game theory as a means to understand conflict.
As with every other academic theory, heuristic model, or mathematic equations attempted to be translated into practice, we must be cognizant of its limitations. In the case of game theory, as well as much of economics, the limitations are in the assumptions.
Game theory assumes:
1. A fair game. Both sides are playing by the same rules.
2. Perfect communication and information. Both sides have access to the same information and both sides actions and threats are perfectly translated.
3. Rational Actors. Most of game theory translates rational actors into the western, judeo-christian mentality.
4. Closed system. In Chad's case, the player's actions of gun control would be mutually exclusive to any other actions taking by either actor during the game.
In practice, none of these assumptions hold true. Nobel Laurete, Maynard Smith, was perhaps the most pragmatic about this reality in the preface to his award winning theory, "Evolution and the Theory of Games." He thought that his model could never be translated and applied to human behavior.
Chad- this is not a dig at your most excellent paper, and I hope that you still consider another article on the other projects that you were working on at NPS.
Best,
Mike
Thank you, everybody, for the comments. My real motivation for writing the article, which I never spelled out, is that game theory is indeed useful for working through strategic interaction of potential lines of operation, but more important to groundpounders and SOF professionals is that it offers a method to explain what we understand intuitively to skeptical audiences. While a resolute convential force battalion commander might remain confident in his stance comparing his own intuitive assessment of various PRCM to mine, game theory offers a way to illustrate a line of reasoning that, if he accepts the evaluated goals for each side as correct, will result in an outcome he must accept as likely (however unpleasant). I was not trained in game theory when I was there, and I saw weapons taken away weapons from farmers that I thought should have been allowed to keep them. The GPF commanders I worked with were not unreasonable or unintelligent, they were simply trained in doctrine that didn't address the likely courses of action of farmers caught between insurgents, criminals, and coalition forces.
As far as the 2nd Amendment discussion, remember that the English Common Law definition of militia in the 1700s was "all males between the ages of 16 and 65," and did not describe a particular standing force. The 2nd Amendment was not meant to cover the actions of the standing militias, it was intended to provide a means of rapidly establishing a militia from the people should the need arise, be that for the common defense or to protect against the tyranny of an overreaching Federal government.
Excellent work. Unfortunately I was one of those people seizing weapons. It went against everything I believed in, but orders were orders.
At some point in late 2005/early 2006 the residents of South Dam Village and Haditha were finally allowed to retain "enough" arms. Very soon after that they reported that they were able to resist, on two separate occasions, insurgent attempts to steal vehicles (presumably for VBIEDs) and money.
One worker reportedly said, "We chased them away with our AK's. Are we allowed to do that?" The unit commander replied, "That's why we let you have them."
That my $.02
Dayuhan you said:
"A less ideal but more likely scenario would involve multiple militias fighting over control of various profitable enterprises, such as drug production, smuggling, etc. For an example of a state where the balance of coercive force rests with militias, see Somalia, not the prettiest picture around."
You might be right, but as long as they are not supporting attacks against the United States or our allies; I think we tolerate it. If there are Afghans who want law and order; I think we should make an effort to assist them.
Dayuhan you said:
"Isn't such a separation implicit in the idea of an armed populace as a check on government authority? If the people are the government, why would they need arms to defend themselves against themselves?"
Because there may be competing governments, foreign governments, and people of your own government can be bad.
<i>Ideally it is a balance of power. So if one militia gets out of line, there will be many other militias that will have to be convinced one way or another. So no one group really holds the power or has the control.</i>
A less ideal but more likely scenario would involve multiple militias fighting over control of various profitable enterprises, such as drug production, smuggling, etc. For an example of a state where the balance of coercive force rests with militias, see Somalia, not the prettiest picture around.
<i>I find it interesting that you phrase your questions in the sense that government is separate from the people.</i>
Isn't such a separation implicit in the idea of an armed populace as a check on government authority? If the people are the government, why would they need arms to defend themselves against themselves?
Dayuhan you said:
"how is that government going to control the militias if they get out of line... which they will?"
"who decides what the dividing line is between "legitimate" resistance (insurgency) and "illegitimate" resistance"
Ideally it is a balance of power. So if one militia gets out of line, there will be many other militias that will have to be convinced one way or another. So no one group really holds the power or has the control.
I find it interesting that you phrase your questions in the sense that government is separate from the people. It seems to be that is what the United States is pushing in Afghanistan. I think this is wrong, and opposite from the principles our country was founded on.
Principles of freedom are not perfect, but on the whole they work much better than the alternatives. We should not let fear of "what if", move us away from these principles.
I don't think either point is silly. A well-regulated milita requires someone to do the regulating. If a government's armed force consists of militias, how is that government going to control the militias if they get out of line... which they will? There's no reason whatsoever to assume that militia units will share the same view of national interest or submit to government control unless government has the capacity to do so.
The second point is also reasonable, if reduced to absurdity. If we posit that armed citizens have the right to resist government if they believe government is intruding on their rights, who decides what the dividing line is between "legitimate" resistance (insurgency) and "illegitimate" resistance (crime). Obviously the government can't decide, because it considers both illegitimate. So what divides the two? Is it purely a question of how many citizens are shooting?
D., if you want to have a serious conversation, you need to not be so silly.
Obviously (to me anyway) the term "well regulated" was placed quite intentionally into the US Constitution to avoid just what you describe in your first comment. The NRA crowd tends to gloss over that point as they merely want every Jackass in America to have whatever weaponry he desires, which is a perversion of the both the letter and intent of that fine bit of law. Afghanistan needs a system of well regulated militias to best address its most likely and most dangerous security concerns IMO.
As to your second point, come on. I know you don't think that is what I meant. While all insurgency is crime, not all crime is insurgency. The fear of insurgency is what keeps governments in line; the fear of crime is what drives them to adopt gun control positions that may well open up a nation to future governmental abuses that lead to insurgency and foreign intervention in the form of UW.
<i>Afghanistan is a country that screams for a militia-based military built upon an armed populace </i>
In many ways they already have that, don't they? Of course those militias are mainly fighting for their own immediate perceived interests, which gets a bit sloppy.
<i>
An armed populace is critical for keeping government in line. </i>
If I shoot an IRS tax collector, or a traffic cop that pulls me over for speeding, am I a murderer or am I keeping the government in line? Where isd the distinguishing line to be drawn?
Liberals tend to see the best in man and government, so are quick to disassemble essential tools for preventing insurgnecy that they do not understand for their primary purpose.
I stongly encourage any who are serious about insurgency, UW or COIN to review the American documents that we used to explain why we were becoming insurgents (Declaration of Independence) and then to counter and prevent our own growing insurgencies and stabilize the new country (Constitution and Bill of Rights)
One of these COIN tools was the right to bear arms. Now, at the time, every man, from his late teens to mid-40s was required both to be a member of "a well regulated militia", and also to own, maintain, and know how to operate his own firearm. This served two purposes: It was the right type of land force for the threats that we faced at that time; and it also was a sure and effective reminder to any government, at any level, that the populace whom they were sworn to serve also possessed virtually all of the guns as well (OK, the state possessed guns stored in armories, but had no forces to employ them, thus why these armories were the first stop for guys like Mr. Shay and Mr. Brown some years later...)
This was by design. An armed populace is critical for keeping government in line.
Now, another lesson is that in the interim we had articles of Confederation, that were fatally flawed due to having a focus not for COIN, but rather for a focus on preventing what we feared the most as insurgents :A strong centralized government. Once we became counterinsurgents (as all successful insurgents do) we realized our error and corrected it.
In Aghanistan, quite similarly, the initial constitution was written to prevent what they feared : warlords and militias; so it was written to vest all power in the central authority and to ban local militias in favor of a large standing army. Just a suggestion, but it is time for a re-write from the perspective of the government's new role as counterinsurgents, and ask not "what did we fear", but rather "what must we divide and protect between the government and the populace to prevent future instability." Afghanistan is a country that screams for a militia-based military built upon an armed populace (afther all, the only threats there are such that completely befuddled the Soviet and US conventional forces, are the ANA better?); and also more decentralized control and power back to the states and the people. To "support and defend" their current constitution is to tie oneself to failure.
Brother Mac,
I'm down on the beach with my daughter right now. Interesting enough, I checked your reply as we were watching the sun go down on the intercoastal. As I was reading, she thought it would be funny to spit water on my face :).
The Shakespeare quote was good. I will challenge some of Chad's assumptions eventually, but right now, I'm simply happy that he's contributing to the greater narrative. He's wicked smart, and his ego will not be detered by our jousting no doubt :).
I'm just trying to keep you in line as well as COL Gentile, Steve Metz, Bernard Finel, Jim Gant, Carl Prine, and COL Bob Jones over the next year as many of your well-worn and heart-spoken truths become truth.
As iron sharpens iron brother. Perhaps, one day long after all y'all old men are gone, I'll figure it out and voice my own truth. For now, I opine once in a while, provide some snapshots of reality, play referee, and overall, just keep being me.
Mike
Damn... I just learned that "all the world's a stage, and all the men and women merely players"... While I was busy learning and practicing my lines ... I totally missed the fact that I am merely a game piece responding to the roll of the dice... Who knew? And why didn't those that did know tell me sooner!!!
Frustratedly yours,
MAC
This is a very well written essay. As a student, I am most intrigued with the idea of using game theory to analyze the problem and present possible solutions. Game theory is both simple and complex. It's a simple concept, but as more and more possibilities and paths are examined, it can become a complex problem.
In my opinion, I think the main point is that commanders should evaluate their actions from different perspectives to determine if other parties support them. Game theory is one method for commanders to do this.
MAJ Hinkle:
I did not get the same read of this paper. I think that the author was showing through game theory the choices that farmers had and were willing to make in regards to retaining guns for personal protection given the coalition's ability to protect the population. Understanding that relationship can help the commander to devise the best policies to control guns without necessarily taking them away or criminalizing their possession (which results in then only the criminal having guns). I think his concluding paragraphs are most instructive:
"In conclusion, whenever any authority criminalizes a legal activity in an attempt to reduce the incidence of an illegal activity, second and third-order effects are generated which may result in an outcome not only less effective than hoped for, but even counter to the desired effect. In the case of the al Jazeera desert in Iraq during 2006, the coalition forces attempt to minimize the number of weapons available to the insurgents operating in the area contributed to the insurgents means of support. The farmers were unable to do anything but support the insurgents regardless of whatever preference they might have for a functioning Government of Iraq. The presence of criminals and the coalition forces inability to protect the population resulted in alack of support for both the legitimacy of the Government of Iraq and coalition forces, and provided the farmers no incentive to follow the rules of an authority that made criminals of a group without criminal intent. By analyzing the options available to the farmers, local commanders could have shifted policy to provide incentive to the farmers to protect themselves, perhaps eventually resulting in less of a need for weapons at all, and a willing reduction of arms."
Pursuing gun control of the law abiding citizen to curtail the illicit activities of insurgents in Iraq is a frustrating dilemma that have parallel arguments in even our peaceful society in the U.S. Do we take away the ability of a citizen to protect himself and his family until help can arrive, or do we allow the insurgent/criminal take advantage of the helpless law abiding citizen without this protective right?
I worked with the initial push to stand up the Iraqi Police Force and saw the reactive versus proactive tendencies of their culture. Literally, the Police were viewed as the report takers and body receivers of those that were killed. Even after returning a few years and continuing the mentoring of the Police Force, the fact was that the Police were responders of incidents and could do little to initially protect its citizens the critical few minutes that were needed to protect life. An armed citizen and his neighbors could save his family during the time of the crisis until help comes.
This is not just a battlefield problem.
I also was a Police Officer in Ohio which was in charge of protecting a large metropolitan city that was ravaged by crime and gangs and saw the difference between a family able to protect themselves until we arrived to help and those that could not due to not being armed. With the ongoing budget problems across America there continue to be few police forces capable to have a response time under 5 minutes. Imagine how long a couple of minutes are when confronted with an aggressively armed criminal. A citizen with a gun can "hold down the fort" until help arrives.
We must continue to balance our Soldiers' protection and the citizens of Iraq need to protect themselves from those who wish to due them illicit harm. Taking away their only means of survival (a gun) is not the best solution and will create a false sense of security that may cause their demise.
MAJ Ronald Hinkle is a Military Police Officer currently assigned to the Command and General Staff College in Ft. Gordon, GA. He holds a M.A. in Business and Organizational Security Management from Webster University and a B.S. in Social Psychology from Park University. The opinions expressed here are the authors own and do not reflect the official policy or position of the Department of Defense or the U.S. Government.
Another factor is the lack of continuity from unit to unit. In Baghdad, in 2003, I not only wanted homes in my sector to be armed, but I had to actively assist arming them. I think the folks in my AO were the only people in Iraq who didn't have 15 AK's in their living room. They were Sunnis outgunned by thugs in nearby Sadr City. Anytime we captured weapons, we distributed them to the neighborhoods most in need and tried to work cooperatively with them to maintain security, as I often explained to them that my platoon could not be everywhere at once. After a month or so of that, we were sent to a different sector. I visited a week later and the unit that replaced us was confiscating weapons, against my advice. 180-degree reversal in policy. Aside from making the security situation worse, what kind of a message did that send to the people?
Chad,
Good to see you here and congratulations on the promotion. If you have time, then I think it would be value added for you to write an essay on the other work you were doing at NPS for the SWJ community.
I'd like to add a couple of comments for further explanation to Chad's essay.
By 2005, Coalition Forces (CF) successfully identified two significant enemy training camps and safe-havens. One was Lake Thar Thar; The other was Turki Village in Eastern Diyala Province. Another camp existed along the Diyala River Valley, but it was but a rumor at the time.
These denied areas sat on unit seams and boundaries or otherwise un-patrolled areas in the desert provide perfect cover and concealment from the human terrain. They sat well below what Dr. Kilcullen labels the "detection threshold."
In 2006, CF began continious operations to destroy these safe-havens and deny the enemy ground in which to train new recruits. Chad speaks to the combined efforts of the Marines, Army, and Special Forces to handle Lake Thar Thar. My unit was responsible for Turki and the DRV.
Restrictive Population Control measures (Gun Control, Curfews, and Coercive Civil Affairs) proved necessary for us to assist the government to take back control of these areas.
Chad explains it with game theory. It can also be explained through psychology in the analogy of a family conducting an intervention on a family member with a substance abuse problem.
Mike