Member Login Become a Member
Advertisement

COIN Metrics: What Not to Measure

  |  
02.08.2010 at 07:36pm

Kilcullen (I): Here’s What Not to Measure in a COIN Campaign – Tom Ricks at FP’s Best Defense.

When David Kilcullen is at his best, he is unexcelled at discussing how to wage a counterinsurgency campaign. And I think the Australian infantry officer turned political anthropologist/COIN guru is at his best when he gathers field observations, boils them down to distilled principles, and then describes those rules in a clear, practical manner.

So I want to take some time to go through a paper he wrote recently in Afghanistan. (I didn’t get it from him, by the way.) While it ostensibly is about metrics in COIN campaigning, it amounts to a thorough discussion of what works in such warfare, what doesn’t, and — especially — how to tell the difference. It is written about the current campaign in Afghanistan, but clearly has broader applications. …

After some initial throat-clearing (one of my rules when I was an editor was to see if I could cut the first three pages of any long article), Kilcullen’s first major section is about metrics to be avoided…

Continue on for “what not to measure in a COIN Campaign” at Best Defense.

About The Author

  • SWJ Staff searches the internet daily for articles and posts that we think are of great interests to our readers.

    View all posts

Article Discussion:

0 0 votes
Article Rating
79 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Carl Prine

This is underwhelming.

It is time for SWJ to get serious about the metrics underlying our assumptions about “success” or “failure” in COIN.

Major Scarlet

as i’ve said before, it doesn’t matter if we win in Afghanistan and Iraq.. the global terrorist network is still intact because we have failed to deal with the real problem. The real problem is the sponsors of terror.. namely in Iran, Syria, Saudi Arabia, Yemen, etc. I’m amazed that with our global intel network and knowledge.. we can’t identify and attack (by whatever means) our real enemies and put this to rest.

Bill Moore

Outlaw7,

Concur with your comments on open source warfare, but I also think it has been around for awhile. The tools of globalization have magnified its impact and will continue to do so, thus it is an important concept to understand. However, why does open source warfare invalidate our COIN doctrine? I’m not overly sold on our current doctrine either because it seems incomplete, but I don’t see how open source warfare in itself invalidates the doctrine?

William F. Owen

Seriously? Where is the insight here? If you know nothing, then I guess something counts – and sorry, but Dave is wrong as often as he is right.

gian p gentile

Agree with Carl and Wilf.

Is this the best that Tom Ricks can do with Coin metrics, just regurgitate some pithy catechisms from Kilcullen that were actually developed in Tizi Ozou a generation ago? This thinking has become quite stale. It is time to move on and take the debate and discussion to a different level.

Come on. The stock mantra that Ricks highlights that the more civilians killed the more insurgents made and therefore the less pacified the area is fanciful theory. Is it such as Ricks so bombastically states “nuff said”? Perhaps not!

In fact it is hard to know what effect is produced by the killing of civilians in war. During World War II American Airmen believed that the bombing of industrial centers and the killing of civilians (although at the time American’s referred to them as industrial “workers” to be “de-housed”) would weaken morale. But studies after the war based on interviews of German civilians showed that bombing in some albeit complicated and qualified cases actually stiffened German morale to resist.

In Vietnam the myth created by some analysts is that Abrams’s so called radically different “One War” approach pacified the Vietnamese countryside from 69-72 through a hearts and minds counterinsurgency campaign modeled on David Galula in Tizi Ozou with population security as one of its core pillars. Nope, not true, and not supported by current scholarship based on Vietnamese sources. To be sure there was a significant level of “pacification” that occurred between in 69 and 72 in the rural areas. But that was because many rural areas once under VC control were de-populated by the destruction of war and the forced resettlement of hundreds of thousands of civilians. In a sense superior American and South Vietnamese firepower “pacified” the rural countryside by “draining the sea” that the VC swam in.

My point in highlighting these two historical cases is not to explore the moral issues involved in killing civilians in war (which of course is an important and worthwhile subject) but to point out that the theory that underwrites current American Counterinsurgency practice and thinking is broke. Yet that very theory has shaped a New American Way of War and has seduced analysts like Ricks into believing it to be proven in practice. It is this very Coin theory that is driving current American operations in Afghanistan.

It is time for FM 3-24 to be deconstructed and put back together in a similar way as the American Armys Active Defense Doctrine was between 1976 and 1982. That previous operational doctrine was thoroughly debated and discussed in open (not bureaucratic) forums and the result of that debate was a better operational doctrine for the time commonly referred to as Airland Battle . In short, FM 3-24 today is the Active Defense Doctrine of 1976. It is incomplete and the dysfunction of its underlying theory becomes clearer and clearer every day. We need a better and complete operational doctrine for counterinsurgency. The American Army needs one that is less ideological, less think-tank and expert driven, less influenced by a few clever books and PhD dissertations on Coin, and one less shaped by an artificial history of counterinsurgency.

When will the American Army undertake a serious revision of this incomplete and misleading doctrine for Counterinsurgency?

Carl Prine

Take it one more step, Gian.

If we’re being honest, then we’re conceding that those traditional metrics developed by Galula and sanctified during the great 1962 COIN conference at RAND in Santa Monica haven’t been proven to work in OEF.

There might be many reasons for that, including the possibility that we face post-Maoist insurgencies that require new ways of articulating force to achieve results, much less measure this in any meaningful way.

Rather than be the dogged critic, however, I simply would ask that SWJ use its uniquely fertile open format to convene an online conference about alternatives to not only waging war but finding new yardsticks to chart progress or failure.

Scanning the list of the authors to the right, I don’t see any names that jump out as geniuses of complex statistical models. Bing West’s work in Vietnam perhaps comes close, and COL Maxwell’s mind might prove the sort of brain necessary to arriving at a working system of metrics.

Kilcullen co-authored the “Triage” study at Nagl’s CNAS that included a chapter on “metrics” that was almost a punchline, not a serious survey of the issue, so I’m not exactly breathless for his added insight into the problem.

Regardless, it’s time to start asking tough questions about why assumptions about COIN didn’t work in OIF or OEF. What didn’t we know? What still escapes us? What novel ways might be develop to better measure the success of our oplans?

Ricks’ striptease that pulls away one Kilcullen fan to reveal another one from Galula isn’t satisfying to anyone. If he and Kilcullen can’t seriously confect some workable system, then let’s ask SWJ’s peeps to do so.

Carl Prine
Justin Palmer

This is awesome! My battalion is in the chute to deploy to OEF later this year and my intel shop has been desperately trying to find some non-kinetic metrics to measure. Thanks David!

Ken White

Heh. Robert McNamara has a lot to answer for.

‘Metrics for Warfare.’ That’s a myth he largely inculcated over seven years. So-called COIN also badly mutated in that time…

Carl Prine is right, It did get worse.

Not least because those western constructs have only tenuous application to most Afghans — whose country it is…

However, Ricks last paragraph really sums it up quite well:

“…Can you imagine being a new battalion commander in the area trying to keep up with this stuff? Tribes, women, feuds, land disputes, religion — it is just too hillbilly for me…”

Though I suggest that the proposed ‘metrics’ are not at all “hillbilly” (and I are one…). No self respecting hillbilly would waste his or her time trying to apply ‘metrics’ to any volatile human endeavor involving multiple parties and goals particularly if combat were involved.

The hillbilly would simply apply subjective judgment to the situation and arrive at a decision on what to do next until such time as the endeavor was proceeding acceptably or completed satisfactorily.

There is a place for metrics in war. They apply in the fields of technology, engineering, logistics — even in the application of certain aspects of combat power, specifically Artillery and allied support processes. None of those facets can operate properly without at least some metrics.

For ground combat operations involving human combatants and possibly innocent civilians the reliance on ‘metrics’ is as — or more — likely to create confusion as to provide any clarity. That not least because those who demand ‘metrics’ are those most likely to misunderstand their meaning.

Warfare is too volatile and people and their actions and reactions too unpredictable to provide scientific clarity and the pursuit of such clarity is a chimera. Reliance upon metrics in warfare leads to flawed decisions and that is shown by the simple fact that this discussion is occurring — there is no question that logistically, tonnage required is valid; in engineering, the strength of material is an important knowledge — but in the realm of combat success or ‘COIN’ success, no one can decide what aspect of that constantly moving train is important.

Success in war is like pornography. Lacking total annihilation and destruction of the opponent (the exceptional but today unacceptable metric that proves the rule), we know it when we see it.

Carl Prine

I’m struck, Ken, by your missive posted after that of Justin’s.

Let’s assume that he’s a young officer minted for the S-2. He’s searching for a workable means of measuring battlefield progress when confronting the complex human topography of Afghanistan.

The military mind doesn’t often care so much about scrutinizing the highly localized aspects of COIN but, probably rightly, finding a means to check off lists that would eventually conspire to give them a statistical picture of their battlespace.

In the latest effort by Ricks, there’s really nothing new here. Since 2003, the various branches of our military have struggled to find means of measuring progress in OEF and, later, OIF.

What often intrudes, of course, isn’t merely the inconvenient fact that much of the traditional COIN metrics ginned up during Maoist wars of revolution simply haven’t worked when we’ve sought to understand contemporary conflicts. Instead, it’s the more prosaic problem of properly resourcing one’s information gathering; ensuring that the info being collected actually seems to chart progress or retreat; and articulating a means of using force or suasion in concert with the numbers briefed in the TOC to achieve the policy goals articulated by your chain of command.

Based on Ricks’ output so far on this topic, I’m not getting a warm and fuzzy that Kilcullen has split the atom. Indeed, to stretch the metaphor to the breaking point, I’m not convinced that he can even describe what an atom looks like, only what they thinks the structure should approximate if everything is going swimmingly and we had all the time in the world to build the galaxy’s biggest microscope.

Let me make a prediction. When Kilcullen finally publishes this essay, we’ll likely get a rehash of Bernard Fall’s perspective on measuring “administrative control.” Just like Fall, Kilcullen seems to be keying on levels of “securing” the population (assassination attempts on government personnel, attacks on government or NGO projects, rates of enemy taxation), mixing in notions culled from Malaya such as increasing the people’s prosperity and risking their skins providing intel to occupiers’ patrols.

He seems to omit great lessons learned elsewhere, such as Napoleon Valeriano and Charles Bohannan’s insight into scrutinizing government competence and capacity, the legitimacy of leaders following an election and other, perhaps inconvenient in Afghanistan, yardsticks to consider. Or, at least, Ricks hasn’t highlighted this yet.

I feel like the key metric really is how many times I beat a dead horse: The point of metrics remains to see how well we’ve come toward mitigating or solving the causative forces that produce the insurgency in the first place.

What are our political goals? How might strategy link tactics to our desired outcomes? What is causing the revolution and what might fix it? How do we measure the waning or waxing of these forces?

I fear Kilcullen is falling back on Galula’s statements to the RAND gathering in 1962, and all his descriptions seem to echo the paradigm.

But what if we’re in a post-Maoist phase of revolutionary warfare? What if the model needs to be tweaked? Why aren’t those Galula-esque metrics working in OEF now?

Based on these traditional metrics, one would think that the war was won in 2004 and the existence of the Taliban’s shadow governments in so many provinces is a statistical outlier.

gian p gentile

But Carl, it gets even worse than that which is why Ken’s post I think is quite perceptive. Even during the age of Counter-Maoist Coin we couldnt get metrics right. If anybody out there thinks that the Hamlet Evaluation System created by American experts for the Vietnam War accurately assessed and portrayed what was actually occurring in the rural areas then they are not keeping up with current scholarship.

So now we have the Current Coin crowd led by analysts like Ricks who regurgitate the metrics created by Galula, Thompson, and American experts in Vietnam and these current folks assume that the metrics worked then; they didnt. So we have the worst of all possible worlds. A current bevy of metrics constructed in a bygone age that were half-baked even at the time they were used.

Like Ken was getting at an assumption underlying metrics is the idea that the system one is measuring functions according to rules that can be measured. But what if the system itself being measured is broken?

Ken White

Carl Prine:

That sequence was inadvertent but sort of serendipitous. Justin is doing what our current processes tells him he should do. Good for him. Those ‘metrics’ will give the US Embassy some decent information even if the rapidity of change makes that data appear questionable from time to time. As for giving his commander (or his Commander) any really useful information, I’m dubious.

If one measures the effort to obtain said metrics against the benefit to that Bn or the Bde, I’m beyond dubious…

A response to that is “…perhaps, but higher Hq need such information.” Uh, not really, I’ve seen it produced by the binder full. It gets plopped into briefings because the briefers have to have something. Good rarely flow from it…

Having participated in the statistical folly Gian aptly cites (and quite a few others over the years), I’ve seen little good and some harm come from such figures. That harm, frequently, revolves around massive effort to capture ‘data,’ often of little real value and much of which can literally change by the minute and thus give a false picture leading to ill informed decisions. You say:

The military mind doesn’t often care so much about scrutinizing the highly localized aspects of COIN but, probably rightly, finding a means to check off lists that would eventually conspire to give them a statistical picture of their battlespace.

I think that’s a bad misnomer. That predisposition for lists is not from military minds, it is in fact from managerial minds and anyone who truly believes soldiering or warfighting is “management of violence” is in the wrong trade. Warfighting is the practical and targeted application of violence, a thing which is not ‘managed’ at all.

You further accurately state:

” Since 2003, the various branches of our military have struggled to find means of measuring progress in OEF and, later, OIF.

What often intrudes, of course, isn’t merely the inconvenient fact that much of the traditional COIN metrics ginned up during Maoist wars of revolution simply haven’t worked when we’ve sought to understand contemporary conflicts.”

Yes to the first and that’s merely one symptom of my assertion that there will be no effective metrics for the conduct of combat operations (combat, not military. That, too is a difference, the two are not synonymous).

The second quoted item implies that the COIN Metrics during Maoist wars were effective. They were not. Their failure in fact is the genesis of my comment above; “the reliance on ‘metrics’ is as — or more — likely to create confusion as to provide any clarity. That not least because those who demand ‘metrics’ are those most likely to misunderstand their meaning.”

I’d add to that the fact that the pursuit of ‘metrics’ often subsumes or overtakes the real mission in magnitude of effort.

As you also say:

“The point of metrics remains to see how well we’ve come toward mitigating or solving the causative forces that produce the insurgency in the first place.”

Perhaps you’re correct but two thoughts occur. I think that entails being totally sure you really know what those causative forces were (and I submit that little item will skew your ‘metrics’ as it will vary from group to group and day to day…). I’ll also mention that the ‘causative forces’ may be a pure ethnically or religiously based power struggle which an intervenor is highly unlikely to be able to mitigate in any meaningful way.

The second is that such mitigation has to apply to the affected population, not to one’s own ideas of what one would want if one were a member of that population when one is not. Convoluted and ungrammatical but very pertinent today, as it was in Iraq and as it was in Viet Nam — and, believe me, it really was off base there…

As Gian says:

“So we have the worst of all possible worlds. A current bevy of metrics constructed in a bygone age that were half-baked even at the time they were used.”

Carl Prine

Ken, I’m not willing to concede that some metrics failed to be important to COIN commanders during the years of Maoist or other sorts of revolts against colonialism, Ken.

If we agree that all COIN is local, then I’m willing to be persuaded that in Galula’s relatively quiet sector in Algeria a key measurement of pacification was people volunteering information to the colonial occupier.

But I’m not willing to agree that this was so in all conflicts, nor that it might be an important measurement today in ALL of our post-Maoist AOs. I guess I would rather say “it depends.”

I also worry that perhaps sometimes the metrics we think are successfully recording “success” in an area actually are measuring a different phenomenon. Perhaps what they’re measuring isn’t the mitigation or destruction of the causative forces that created and sustained the rebellion but rather occupation itself.

The other problem is my concern that certain metrics might be at cross-purposes to themselves. Let’s say that Kilcullen is right and that he notices attitudinal improvements in a local population after we bring in economic development dollars. Indeed, they start giving the COIN commander more intel about bad guys laying in IEDs.

But what if the local economic development also does other things, such as hardening the anti-occupation ideations of key segments of the population? What if the increased amounts of aid end up providing the population with more income, which is then skimmed off the top by Taliban tax collectors to become cash inputs that feed the rebellion? If not in that AO, then perhaps in the one next door because even the Taliban don’t like to kill a cash cow when she’s putting out so much milk?

This is a problem I’ve had with some of Kilcullen’s wisdom about metrics in OEF. OK, we have reports that the NGO to whom we’ve lavished dollars has hired X number of MAMs, spent X number of dollars and the people are X times more likely to love us and give us info. A key success, the Army would say, is that these NGO programs also have been X percent less likely to be vandalized or raided by the Taliban.

But let’s look at it another way. Could not the lack of attacks on the project signal that the Taliban have been paid off to NOT attack it? Might vandalizing a new factory, school or collective of farmers be a sign of success for the counter-insurgent because it implies the taxman wasn’t given his tip?

What if we were to structure our aid programs so that we only provided money or help to a village that had been pacified, thereby bribing them to stay on our side rather than bolt to the enemy?

What it all really boils down to is understanding the nature of the war. What if the Galula-esque measurements are good at charting progress toward “administrative control?”

That’s great if we assume that bringing the Karzai administration to certain sections of Afghanistan will mitigate or end some of the causative forces that created the conflict.

But what if the nature of the war is different? What if what is really the nature of the conflict today involves a retributive civil war between non-Islamist Pasthun peoples and those formerly in the Northern Alliance and other ethnic groups bought or otherwise wooed to the side of these ethnic militias?

What if a possible cause of the violence is the very extension of the Karzai government (or any government) to these enclaves?

If so, then pacification doesn’t come from establishing governmental control but, perhaps, negotiating a deal wherein not much centralized state control enters that valley, so long as al Qaeda doesn’t arrive, too.

How do we measure that?

I was hoping that SWJ might take up those topics and not blurb the silliness about metrics blandished above.

Gian, I don’t think HES accurately measured the reach of government in Vietnam. Bing West, among others, lampooned the statistics compiled then for this very reason.

But I’m NOT persuaded that in a counter-Maoist war it’s unimportant for the COIN commander to chart the attitudes and effective reach of the host government. In other words, perhaps HES was the right sort of knowledge to have learned, but unfortunately wasn’t a very good program for arriving at that knowledge due to local commander’s bias, the pressure to put out the rosiest sitreps or yada yada yada.

Vietnam was such a stat-intensive war not simply because of the fad of systems analysis or that we had so many programs beating the bushes to find stats. We had so many programs beating the bushes to find stats because we really didn’t understand the nature of the war, the enemies we were fighting or our own role within the populations in which we lived.

Those fat binders Ken faced were there because commanders failed to understand the nature of the war and so collected all the info that they could and asked everyone to just figure it out.

Recently, the MG in charge of intel in OEF came up with the same sort of intellectual shrug that his peers in Vietnam made: We have billions and billions and billions of bytes of data that we’ve collected. So far, we haven’t cracked the code on the nature of the war or a good means to defeat the insurgency. The grail is hidden in all this data!

Maybe not.

Schmedlap

The problem with metrics is not whether we are choosing the right ones. The problem is that people do not know how to use them.

Measures of effectiveness are supposed to measure whether and to what extent certain actions are having the desired effect. They are not supposed to measure whether a particular symptom of something undesirable is subsiding.

Ricks typed, “I think the Australian infantry officer turned political anthropologist/COIN guru is at his best when he gathers field observations, boils them down to distilled principles, and then describes those rules in a clear, practical manner.” No, he’s not at his best when he does that. He is at his absolute worst. Gathering up anecdotes and deriving principles from them is more often misleading and oversimplified than not.

Outlaw 7

In order to correctly discuss current COIN FM metrics maybe we need a thorough discussion on the topic of “open source warfare” that in fact challenges the FM and current COIN thinking.

Especially since a recently released study of “The Ecology of Human Warfare” in Nature magazine draws attention to just maybe how wrong the initial COIN has been. It actually proved the theory of “open source warfare” and it was peer reviewed and challenged which I believe has not been the case for COIN/and the COIN FMs. There have been constant discussions about COIN and the two COIN FMs BUT has it been taken literally apart and peer reviewed?

Currently there must be thousands of individual comments of for and against but has it been challenged by quantitative analysis?

If it has then one would have a set of metrics to look at and further challenge. But since it has not been quantitatively analyzed then the metrics are missing and all we have is chatter about metrics.

“Open source warfare” has now been quantitately analyzed and found to be accurate MAYBE it is time to seriously consider looking at the concept.

MikeF

Justin Palmer,

I’m curious to know what you’re so excited about. What is so awesome about what Dr. Kilcullen said?

And what is non-kinetic and kinetic?

How does any of this help you with accomplishing your mission in A’stan?

Think about these questions as you go through your tour.

Mike Few

Schmedlap

Mike,
I think it was tongue-in-cheek. (I sure hope it was, anyway).

MikeF

Schmedlap,

I hope you’re right. Unfortunately, from what I’ve seen, listened to, and read from the new LT’s, I think they’ve started using Pop-Centric PSYOPS indoctrination during Mounatin Phase of Ranger School. While the Ranger candidates are sleep-deprived, they’re forced to listen to John Lennon’s “Imagine” over and over again.

We might as well move West Point to Santa Cruz.

Mike

Ken White

Carl:

I do not disagree with anything you say and I strongly agree that for SWJ to take a collective look for viable metrics in today’s conflict is worthwhile. Not because I believe in the ‘metric as warfighting aid’ issue but because the current US Army / Government milieu cannot seem to live without them. Harvard MBAs have much to answer for. GWB is the only one of those guys to my knowledge who wisely never got wrapped around that axle. Anyway, if we just have to waste effort on them, might as well make them as worthwhile as possible. Plus, I could always be wrong…

However, one thing you said is key to any success in that look:

“But what if the nature of the war is different? …”

Therein lies a part of the problem with metrics and combat. Every war is different, very much so, yet humans will try against all logic to create a standard, transportable set of rules, guidelines — and metrics. As you noted:

“Those fat binders…were there because commanders failed to understand the nature of the war and so collected all the info that they could and asked everyone to just figure it out.”

True dat. I suspect that could and will happen again. In fact, without being in either recent large theater, I’ll bet it’s already happened a number of times…

Schmedlap has it right in large measure:

Measures of effectiveness are supposed to measure whether and to what extent certain actions are having the desired effect. They are not supposed to measure whether a particular symptom of something undesirable is subsiding.”

Heh. That sort of amounts to proving negatives…

If he’s correct and I suspect he is at least in part, the issues thus become ‘what actions’ and ‘what effect one is trying to achieve’ (both must logically relate, a frequently overlooked area for disconnects).

I submit that in combat, even in a mode as sedentary and low pressure as a COIN / FID / SFA operation, that or those effect(s) become the proverbial moving target and given our tour rotation process, you’re shooting from many different positions with varied weapons (some perhaps inappropriate for task…) at that moving target. i.e you might get a random hit but it’s fate, not skill.

MikeF

I just took a look at Part II- Metrics that work. They are good, but they’re not new. The SF community and State Dept have been using them for years. Hopefully, Dr. K will publish this paper so we can get a first-hand account of what he is trying to say.

v/r

Mike

Abu Nasr

I still do not understand why we insist on calling resistance against the governments of two democratic revolutions “insurgencies.” The Awakening was AGAINST al Qaeda in Iraq, not FOR the Iraqi government. There are areas in Afghanistan where locals reject the Taliban AND the Karzai government. Foreigners surrounding a village and helping locals rebuild wheelbarrows fosters a temporary working relationship, not acceptance of a new national government.

The revolutionary aspects of these two wars are completely overlooked. We are admittedly terrible at information operations and still have yet to stand up any significant political indoctrination programs. The Taliban will settle land disputes and administer “justice” through sharia courts. ISAF will repair irrigation and tell opium farmers to plant wheat. Is modernization necessary for pacification?

“Body count” is the only one of Kilcullen’s recommendations that really bothered me. If this is being used anywhere it is an indication of our inability to institutionalize lessons learned. This is mentioned as a criticism in almost every post-Vietnam paper and is just shy of advising against a strategy of attrition against insurgents.

Carl Prine

Perhaps SWJ might further this pursuit, Ken, by getting on a bandwagon that’s been hurtling down the road from CTC since 2008: Revise FM 3-24.

There has been a movement, now growing, to reconcile the manual (and thus doctrine) to what our combat commanders have learned in the field. There’s also an intellectual jihad to cull some learning from Mackinlay’s “Insurgent Archipelago” and related thoughts on post-Maoism from Steve Metz, et al, into discussion about our “best practices.”

Obviously, a section on “metrics” inside the FM would force us to intellectually come to terms with what we wish to measure and how we would tie our operations to what we chart.

In 1962, RAND convened the world’s top counter-insurgent experts, including Kitson and Galula and Valeriano, to Santa Monica to hash out best practices on the eve of US escalation in Vietnam.

They talked about “metrics” and, I would argue, those sorts of measurements might have helped us guide operations because those counter-insurgent gurus were interested in Maoist and similar revolutions.

But I suspect that much of what our battlefield commanders see today is post-Maoist in the manner Metz and Mackinlay and other people who don’t have last names that begin with “M” have argued.

Revising FM 3-24 would require us to confront that, but if Leavenworth doesn’t want to tweak it then SWJ might be the proper forum. Let’s have that Santa Monica conference again. Let’s do it online if need be and let’s talk honestly abou the wars we see out there and the yardsticks we need to develop to understand whether we’re good at this sort of fighting or not.

Carl Prine

Did I not call it?

http://ricks.foreignpolicy.com/posts/2010/02/09/kilcullen_iii_how_to_take_the_measure_of_an_afghan_official

Rehashed and probably unimportant if the counter-insurgent is facing a post-Maoist problem.

Can we now move on to seriously consider metrics for today’s commanders?

Schmedlap

Metrics are not in a list somewhere on a server or in somebody’s notebook – even if that notebook belongs to Kilkullen. Metrics – or measures of effectiveness – measure how well what you are doing is working and they MUST be developed on the basis of what you are doing. You can’t find a list of metrics and then fit your operations to the metrics. Any commander looking elsewhere for metrics should probably be relieved because he is, by definition, incompetent.

gian p gentile

Forget about it Carl, they are lost causes. Both of Ricks’s recent books are premised on the Counter Mao Coin method. So too is Kilcullen’s.

And it is not just a simple problem of jettisoning a method of Coin that was developed for a previous age. No, the bigger issue is that the counter-Maoist method developed by the likes of Galula, Thompson, and many others was itself deeply flawed. As Peter Paret showed a generation ago in his brilliant but mostly overlooked classic analysis on French Counter-Revolutionary Officers what they and others did was to reduce Mao to a simple sequential check list to counter. Moreover, recent scholarship on Vietnam shows that it wasnt the Maoist revolutionary people’s war that transformed the south but instead the hard hand of the process of war through death, destruction, and forced resettlement that did it.

Mao as a text is still relevant but much of the literature, doctrines, and mentalities etc that were constructed to counter it are for the most part irrelevant for todays operational and strategic problems.

What then does this tell you about Tom Ricks’s discussion of re-hashed counter-maoist methods?

MikeF

Carl Prine asked, “Can we now move on to seriously consider metrics for today’s commanders?”

Here’s something to consider for those wishing to quantify warfare. Where is your data coming from? How accurate is it?

Look at the issues we have in the US with conducting a census or voting. One mistake that I observed in Diyala Province in 2006 was that a majority of the measures of SIGACTs and metrics were collected by the Shia-backed gov’t during a time of a burgeoning civil war. Needless to say, there was not many indicators of Shia on Sunni violence in the daily reports. I literally spent six to eight hours of my day some days trying to confirm or deny the data sets into accurate reporting.

Given that US forces operating in Iraq and A’stan represent such a small percentage of the overall populace, I think it is almost impossible to assume that data can be collected, correlated, and analyzed on the macro-level. Furthermore, bad metrics skew decision making.

Should we strive to have accurate accountability and measurement. Of course. In the sterile environment of a FOB, a DC think tank, or academic classroom, this makes sense.

Is it possible? I’m not so certain that it is.

v/r
Mike Few

Carl Prine

I’m not a statistical Luddite, Gian. I assume that numbers can show us things. No one would suggest that someone who bats .003 is as effective in the clean up role as the slugger who puts out a Herculean .368.

The problem is that I think we’re seeing the stats applied to baseball used to describe football. The game has changed and continues to change. The metrics that might have worked once — had they been properly vetted and found to have some utility on the COIN battlefield — probably aren’t quite measuring up now.

Schmedlap and MikeF are right: We must understand the nature of the war and what we hope to accomplish before we can create cookie-cutter measurements that will chart progress or failure across the conflict spectrum.

I guess I’m not ready to assume that we can’t arrive at some measurements. In Anbar in 2005-06, I realized that a key endogenous input for the various Sunni Arab insurgencies came from illicit fuel sales. It seemed quite possible to me that a commander of the AO could institute patrols and policies that would crack down on the sales and scrutinize those doing the selling in order to choke the snake.

It doesn’t mean that the snake wouldn’t have to move on to other food, only that this one input could be quickly attacked. It also came close to addressing a causative force of the rebellion: Criminality in the form of armed fuel dealers who monopolized the trade helped to not only sustain the rebellion but also deny traditional leaders (not necessarily the Baghdad government) the ability to reseize their local legitimacy and help pacify the place.

During a crackdown on illicit fuel sales, it’s quite easy to chart the rising price of the petrol, cooking gas, et al, to see if you’re taking a bite out of crime. It doesn’t mean that you eliminate it, only that you mitigate the effects of the criminality while transitioning to a program that would put these sales into the hands of the more legitimate, more likely to be responsive and, to our policies, coerced leaders.

Sending out patrols that would check on the price of the gas, recording the numbers and names and whatnot of those detained making the sales and tracking the effects of the embargo on the enemy might show some “success” toward completing a mission.

MikeF

Carl,

I’d include a proper understanding of the nature of war AND human nature.

At a minimum, these questions Dr. K is proposing are much better than the one’s we asked in 2005:

1. How many raids did you conduct today?
2. How many Traffic Control Points did you conduct?
3. How many patrols did you conduct?

I think that we should strive for the proper balance of quantitative and qualitative analysis. Yes, in our western world, we like data. However, Iraq and A’stan are much more about storytelling and the narrative. So, if you can combine proper questions and interviews, then you can begin to attempt to frame some concept of the truth.

Still, some things will be difficult to determine.

“So, President Karzai, how many of your family members are currupt and receiving bribes from the government and their districts?”

v/r

Mike

gian p gentile

The problem with metrics in Coin is that they tend to reflect our own set of tactical methods and procedures. So with the batch of Kilcullen’s as reported by Ricks we see measurements of security, or of economic prosperity, etc, etc. Yet these two things among others are inherent to our current system of Coin. So what do we really end up measuring with these metrics, what we want to believe our own system is accomplishing or instead what is critical on the ground?

I go back to Ken White’s earlier post which was a cold-shower call warning folks about the limits of metrics in these kinds of wars. Should commanders come up with them, of course? But once done we should not turn metrics into a measurement of progress for domestic consumption in these kinds of wars. That after all, was the basic problem with HES in Vietnam.

Carl Prine

I guess I disagree with you on communicating progress or failure to the people in a democracy, Gian.

Obviously, it wasn’t hard for Ike’s SHAEF to communicate progress to commanders or the public because all he had to do was publish a map showing where we were and where the enemies were.

It’s trickier to do so with COIN, but does that mean that it can’t be done? As you well know, we both share concerns about how a journalist’s narrative often replaces reality in describing success in these murky sorts of wars.

But that doesn’t mean that we shouldn’t try! The largest problem with HES — it seems to me — wasn’t that the measurement of government control and people’s attitudes in a counter-Maoist fight isn’t important. It was the fact that the measurments in HES were dubious in their authenticity, and appeared to be so by officers in the field such as Bing West.

Were Marines like Bing West wrong in assuming that VC political infrastructure was the center of gravity in their fight? No. He and they were probably right — not for the entire region, but for their COIN effort. Had the US developed the proper metrics, they likely would’ve been able to finely chart just how deeply entrenched VCI was, how futile it had become to change the ideations of the people and how incapable we were at shutting off the inputs necessary to allow the insurgents to subsist.

Those in the field who have to deal with this firsthand realized the shortcomings of HES. What is unpardonable is that civilian and military leaders then used the numbers as “public measurement of progress” for the democracy’s consumption.

To echo you, MikeF, my larger concern with Kilcullen’s list is that he fails to understand human nature, too. If we think back to HES, weren’t local commanders and district advisors under pressure to assign a good grade to their AO? Beyond the natural desire to see one’s own efforts are better than they probably are, there also was a large dollop of CYA: Lowering a grade wouldn’t find any rewards for accuracy or honesty but more likely some general showing up to find out why your village was so screwed up and seemed to be getting worse, unlike what everyone else was reporting.

Or, as Westmoreland put it in 1967, “It appears that again facts will reveal that there has been distortion of reports on the optimistic side. It seems to be one of the things that has plagued this war from the beginning.”

With Kilcullen’s examples, wouldn’t the local commander on a short rotation be more inclined to give in to temptation when measuring the “outputs” of aid or patrols or whatnot?

Might one fudge the attempted murder of an Afghan government official to make it seem more like a pot shot? Might one bundle all the nightletters into a number instead of deciphering the urgency of what they actually said?

Ideally, we want to measure pacification and not occupation. It seems to me based on Ricks’ limited effort that Kilcullen mostly is measuring the former not the latter because I suspect he’s got the nature of this war wrong and hasn’t quite fathomed the bureaucratic nature of our big green beast.

The other obvious problem, as I mentioned above, is that the assumption that spreading the Karzai government to the hinterlands will pacify the populace seems unduly wrong. One might suggest that it’s the spread of this government OR ANY GOVERNMENT that’s sparking the revolutions in the first place.

Measuring government effectiveness, government “control” or government growth really could be seen not as a “success” in the road to pacification, but as a failure.

Anonymous

Carl:

Gian has some wise words:

“The problem with metrics in Coin is that they tend to reflect our own set of tactical methods and procedures.”

Add the fact that we historically do not do the COIN thing well and you’re on a path to problems…

Re: your 2:04 PM post, while that Commander institutes patrols and policies to crack down on fuel sales, as you say “It doesn’t mean that the snake wouldn’t have to move on to other food only that this one input could be quickly attacked.”

If the opponent is far more flexible and agile than are you you will never catch up as he switches food sources. That has been and is likely to be generally true and was and is certainly true in Iraq and Afghanistan.

Metrics tell you what is happening or has happened, that may be somewhat beneficial for State and NCA but it is of marginal or less utility for tactical units (who get tabbed with collecting metrics of no value to themselves…) — you’re always going to be a step behind.

The object is to get several steps ahead…

Carl Prine

Why should we default to the insurgent the notion that he is inherently more nimble than the counter-insurgent? That seems to be another one of those chestnuts of received wisdom, such as the revolutionary always has time on his side and blah blah blah.

It depends. Sometimes, he does. Sometimes, he doesn’t. Sometimes the insurgent is more nimble and sometimes he’s more clueless than even our most ate-up battalion commander.

The larger reason why I brought that up was to continue my jihad against Hearts and Minds (HAM) as a construct. How do we know that it works? How do we know that cutting off the endogenous and exogenous inputs necessary to sustain the machine of revolution isn’t MORE important than winning HAM?

What if HAM is only important if it can help us find the inputs and strangle them? Are there other means of getting this information that don’t involve long, pop-intensive occupations that sap our national strength and ruin opportunities elsewhere? Why do we assume that the lengthy, intensive process of nation building is the ONLY means to defeat, deter or delay an insurgency, especially in a post-Maoist world?

I also believe that SULs tend to be bright enough to figure things out and adapt what seems to work — if they’re given the time, the resources and the ability to experiment without having things dictated to them by martinets of COIN who think they’ve already cracked the code.

Like Gian, I think the maximalist narrative that’s been constructed that so many battalions didn’t “get it” in OIF in 2005-06 until a learned band of savants arrived to save the day is silly.

Lee Brown

Do we even need standardized metrics to demonstrate success? Might not success look different from locale to locale? Can’t commanders articulate progression/regression/stagnation through a few simple statements in a SITREP or briefing?

Think of QTBs to measure unit readiness, ORAs to measure Iraqi unit capabilities, and body counts as a measure of success in VN. None of these are perfect and some down right poor. I’m not sure if we ever really had anything better than measuring SIGACTs in Iraq, but when things started getting better, somehow we knew. Commanders could see improvement in their AOs and could explain it to their bosses.

Ken White

Carl:

Good post and good comments. There is however, a short answer to your well stated points.

Bureaucracy.

We have it and they do not. That simple.

They do foul up; bunches. Fortunately, they err more often than we do which is one reason we aren’t in worse shape. Some times they pull off a good one, more often they do not — not because we’re more flexible but simply due to discipline and training (and the latter could be better…).

There are always a few US units who are better than average and who rarely miss a beat. They are exceptions because the system encourages mediocrity and the tall poppy who builds and employs a good unit has to fend off his competitors for promotion and retention. That, too hampers our flexibility.

I agree with you on the hearts and minds foolishness — it is a bad idea with no real basis in history or reality. I further agree that small unit leaders are far more capable than they are allowed to actually be. I’m with you and Gian on the savants contribution or lack thereof…

However, I think Lee Brown has the metrication idea sorted correctly:

“…but when things started getting better, somehow we knew. Commanders could see improvement in their AOs and could explain it to their bosses.”

This stuff is not rocket science, nor is it ‘the graduate level of war.’ Attempts by a host of Academics and their clones to make it a 500 level MBA course are a HUGE part of the problem…

Robert C. Jones

Typically in insurgency you have a government that is perceived as illegitmate by the insurgent segment of the populace, and more often than not, the intervening party on the behalf of that illegitimate government is there for its own selfish national interests and is most likely perceived as the source of “legitimacy” of that HN government by their populace.

Not judging, just assessing. This was true in Vietnam, it was true in Irag, It is true in Afghanistan.

The military gets sent into these situations to attempt to stabilize what is by definition an untenable, instable political situation. The metrics that are important are POLITICAL; but military commanders are sworn to be apolitical, at least to the politics of their own country. This is the quandry we put our military commanders into. They cannot assess and measure the political metrics that are truly material to resolving the insurgency, so default to measureing things related to their actions on the ground.

Put up a powerpoint slide in your next Commander’s Update that asks and answers the following quesitons:

1. Is the government of Country X perceived as Legitimate? (followed by appropriately sexy color coded graphs, etc to assess)

2. Who does the populace perceive as the source of legitimacy of the government of Country X?

3. Is our presence here perceived to be more about the national interests of Country X or of our own?

4. What national interests do we believe will be secured by sustaining the current government of Country X in power; and what alternatives to supporting their COIN effort could achieve those same effects?

5. Why does the populace at home believe we are here, what do they perceive the measure of “victory” to be, and is that measure linked more closely to our own national interests, or the interest of the government of Country X?

Military commanders can not ask and assess the very things that are most meaningful to victory and defeat in an intervention in another nation’s insurgency. But when you label that intervention as a “War” then the metrics of warfare overwhelm everything else, until the politicians and diplomats at home are also tracking success and failure by the military measures. At this point you have lost the proverbial bubble.

The critical metrics are the political metrics, and they need to be designed and tracked by and for political leaders. We let the tail wag the dog when we focus on the military metris in such matters.

But like any dog will tell you, its pretty damn hard to ignore your tail when its been slammed in the door.

RamadiNights

You’re all missing the point about our involvement in Afghanistan. “Metrics” is a conversation for another time. What we need to be discussing is that we are still doing nearly everything wrong in Afghanistan from a COIN perspective. Having a 4-star talk the talk isn’t having much effect at the company and platoon level. Metrics would be great if we actually had something legitimate to measure from a COIN perspective, but we’re still in “garbage in, garbage out” mode.

Almost every LT and CPT is still trying to get the locals to tell him where the Taliban are hiding. We still focus almost solely on the enemy, instead of gaining ANY useful insight whatsoever about the population. We are still starting over with every RIP, because there are no worthwhile, realistic assessments of where each AO stands along each line of effort, and the plans on where each AO should go are being completely ignored.

We’re still counting on the military to do reconstruction and stabilization when the two facts are: 1) No MOS is any service is trained to do this work, and 2) we have already proved ourselves to be dramatic failures at this. As I write this a USMC MEB in the west of Afghanistan has approved a $3 million electrification project using DIESEL GENERATORS, this despite the volumes that have been written about what an awful idea this is, to include a recently-released USAID document regarding clean energy in Afghanistan. Let’s be clear, though: the reason generators are an awful failure has nothing to do with pollution. This is an abject failure of the DoD to learn anything at all about reconstruction. And lest we blame everything on the Marines, the PRTs are not only no better, they are worse, because they waste more money, and spend more time patting themselves on the back.

Please let me point out that I am writing this from a company-level COP in Afghanistan, I previously served on a PRT here, and I served attached to the Marines in Anbar, so I have some familiarity with everything I’m talking about. (I’d also like to point out that almost all the places I hear the Marines taking credit for pacifying in Anbar were in Army AoRs, so let’s not get too excited about how great the Corps is at COIN just yet.)

Robert C. Jones

Ramadi,

Sounds like you just need to run a powerline up to Kajaki and turn on the switch, right? Oh yeah, that line would run through a hundred miles of prime poppy growing, Taliban controlled territory…

What to do.

One thing worth considering is that the Taliban are not on the warpath because of a lack of reliable electricity in Kandahar City. Neither do the hundreds of seasonal fighters report to the nearest cache to draw an AK or RPG do so because they don’t have indoor plumbing.

Ineffective government did not cause this insurgency, and effective government will not resolve it. The rank and file resist because we are here and because they get paid an honest days wage to do so. So it has always been in this country, and so it will always be. Those who have overdosed on pop-COIN do not fully grasp that point.

This is politics, and bribing the populace is no more effective than killing the populace. At some point you simply have to enable the development of a government whose legitmacy they recognize, and then hold off the neighbors and other opportunists long enough for them to get their feet on the ground.

It won’t be horribly effective, but then it doesn’t need to be.

MikeF

Ramadi- Thank you for taking the time to check in and please feel free to contribute more as time allows.

One thing that I’ve tried to do the past couple of years is look at people and organizations outside the military that are highly successful and see if their approaches can assists us in our ventures. Sometimes, I found things that are useful. Sometimes, not so much.

Here’s one guy that we could take a look at- Warren Buffett. He’s done pretty well in a profession that can sometimes seem like gambling. Honestly, how many fund managers do you trust these days?

So, how does Mr. Buffett analyze companies for a good deal? Here’s a brief overview.

1. He understands that the market is driven by fear and greed. He gets the human nature of any man-made endeavor.

2. He looks at a company using specific metrics over time. Does this company have a good product to sell? Can he trust what the CEO says? He tries to identify quantitative/qualitative measures that bypass fraudulent claims of success and over-valued/hyped companies.

3. He becomes intimately involved in the running of the company to coach, mentor, advise, and assists in order to protect his investment- Economic FID if you will.

I’ll let y’all debate if his thoughts have merit in our world. I think they do. If not, I guess we can turn to Jimmy Buffet.

v/r

Mike

Ken White

Ramadinights:

Good post and great comments. Thanks for being where you are and doing what you’re doing. Let me echo Mike F and ask you to keep posting here as you are able.

One minor quibble. The delusion of metrics is indeed way down the list of concerns compared to the GIGO problems you cite and which most of us are aware of and agree are bad idea. Problem is that if basic principles are ignored — and that is the gist of your valid complaint — or skewed for odd reasons, that contributes to the GI aspect and thus leads to the GO problem.

Metrics, all too often, are themselves the garbage and precipitators for flawed actions. That elementary point was made in another war long ago. It was ignored then and an often counterproductive process became embedded and an almost Doctrinal mistake. If we put it aside when it is a current issue or problem, it just gets shunted to the next time…

We should drain the swamp in spite of the alligators.

Anonymous

‘”Metrics” is a conversation for another time. ‘

Yes, it would be nutty to suggest that EIGHT YEARS INTO A WAR we might have developed some means of assisting commanders in the field with quantifiable and qualitative guides to help them shape their operations.

Silly me. I guess managers in this year’s World Series really can wait a few more for accurate numbers about how certain batters can hit or the efficiency of fielders beyond what the naked eye will show them. Perhaps we’ll just go with “gut” or “hunch” or “art” or “credentials” or whatever also passes for received knowledge (see the celebrity competence of Kilcullen for the proper example) in lieu of actually determining causation or finding out what works.

The larger complaint, it seems to me, would be to understand the nature of this war. Is this sole means of achieving our goals the establishment of a viable national government in Kabul, complete with adequate security forces and sustainable development projects?

If this is the only (or best) means of arriving at pacification, then one should listen to Kilcullen blather about his “metrics” culled from five or more decades ago and repackaged as something new because they might do some good (maybe not) against a Maoist insurgency.

If, however, one begins to surmise that the nature of the war in Afghanistan is quite different from Vietnam in 1962, then perhaps we should consider other means of addressing our long-term goals there and finding ways that we can measure success or failure in achieving them.

When I hear it seriously suggested that by becoming more efficient at sluicing aid to the “people” currently in rebellion, I almost want to gag because I suspect that the extra funds from this are skimmed by the Taliban’s taxmen to fund ongoing cadre development and attacks against us (and which the UN has discussed iteratively while we refuse to listen).

Perhaps what Kilcullen’s metrics really amount to are the continuous blandishment of David Kilcullen’s career. To Ricks, these oats of Kilcullian wisdom are like catnip because they seem so counter-intuitive to anyone who hasn’t really read the COIN literature compiled during the later Colonial era.

I don’t think this, but I fear Gian and others do. They seem sincerely regurgitated for today’s commanders, relearned wisdom that he believes will help them fathom the people, the enemy and the war.

I simply ask that the editors of SWJ quit making the assumption that this so and instead cast a wider net to glean other ideas about warmaking against irregular foes. Should SWJ get behind the CTC push to rewrite FM 3-24 and make it doctrinally more relevant to the commanders in the field, we might arrive at a chapter on “metrics” that works better than what we see misfiring today.

Carl Prine

Sorry, forgot to stamp that one with a callsign.

Schmedlap

Metrics must be established by the commanders because they are derived based upon what operations the commanders are undertaking. They cannot be imposed or suggested from those of us in the gallery as goals for the commanders to shoot for. If you’re searching for some quantifiable and qualitative guides to help commanders shape their operations then that suggests to me that you don’t think our commanders have developed a plan and are just kind of winging it day by day. Maybe that’s true – I don’t know. But then, if that’s the case, a lack of “metrics” isn’t the problem.

Carl Prine

That makes no sense, Schmedlap. Obviously, those at higher ranks (and the democracy which they theoretically serve) deserve to understand how the fight is going. While all COIN is local, redistribution of resources, sharing lessons learned and assisting those elements that need the most help ARE NOT based on what a platoon leader thinks about the sitrep.

Regardless, I’m flummoxed by Ricks’ latest effort, which parrots Kilcullen’s advice for the ANA’s metrics: We’ll grade you on how well you do things that Kilcullen thinks aren’t worth recording for US units!

http://ricks.foreignpolicy.com/posts/2010/02/11/kilcullen_iv_how_to_measure_afghan_army_and_police_units

This is so open to parody that it’s hard not to get a good chuckle from it. Body counts are bad for US forces but “the ratio of enemy killed vs. security forces killed may tell you how aggressive and confident an Afghan unit is.”

If a tiny dot of pee doesn’t come out while you guffaw over this, you have no sense of humor!

Again, SWJ, perhaps it’s time for the adults to start seriously debating the topic. Use your forum for good!

Schmedlap

Obviously, those at higher ranks (and the democracy which they theoretically serve) deserve to understand how the fight is going. While all COIN is local, redistribution of resources, sharing lessons learned and assisting those elements that need the most help ARE NOT based on what a platoon leader thinks about the sitrep.

I agree with all of that, so I’m not sure where our disagreement lies in regard to my last comment.

Carl Prine

I suspect that commanders at the highest levels ARE winging it. Why do I suspect this? Because in August, we received a draft report on how the staff running OEF, in conjunction with NSC, would begin (BEGIN!) to develop metrics and measures of progress that would chart the “transformative effect” of our warmaking and whatnot in Afgahnsitan.

Part of this was leaked ( http://www.politico.com/static/PPM130_civ-mil_plan_afghanistan_090907.html ), but it had been presaged by the CNAS “Triage” report co-authored by Kilcullen that had perhaps the most depressingly bare and unconvincing section on “metrics” imaginable.

Others commented on this at the time ( http://www.registan.net/index.php/2009/07/29/the-strange-contradictions-of-andrew-exums-afghanistan-trip/ ), but the reality likely remains: We’re hearing from Kilcullen now through the Muppet-hand of Ricks moving his mouth because the Australian had been asked to add something to the discussion about metrics and the deadline is now EIGHT YEARS AFTER THE WAR STARTED.

I’m probably wrong, but reading the tea leaves puts us here.

When I hear that commanders in the field are interested in confirming EKIA, EWIA and other sorts of BDA, I’m not willing to just say that they’re doing that because those numbers are easier to get than, say, attitudinal surveys limning the “people” and their thoughts on the Americans and the Karzai government.

It might be because that particular commander is in the clear stage or that the enemy might get some vote on how the AO is shaped. Nutty as that sounds.

I’ve noted another creeping measurement lately that suggests how “bad” commanders do their jobs — their own KIA or WIA.

The assumption seems to be that the pop-centric/HAM doctrine works and that if there are an undue number of casualties in your unit than you’re obviously “too kinetic” or that you’re doing something wrong that’s getting your men hurt.

My only concern as a layman is that journalism should look beyond this shallow sort of causation and cheap narrative and explore what are really the causative forces of the uprising and seek to explain to the general reader how he or she might notice “progress” or “failure.”

But while my interest in this is journalistic, I suppose that military minds of a higher rank than I ever held might add something on the subject more substantive than the regurgitation of Fall and other colonial savants found in the latest Kilcullen/Ricks tag-team.

Schmedlap

I suspect that commanders at the highest levels ARE winging it.

Well then, I would reiterate: if that’s the case, a lack of “metrics” isn’t the problem. Perhaps proper planning has a role in the outcome. Planning has to precede metrics, not follow from them.

SWJED

Carl said: Again, SWJ, perhaps it’s time for the adults to start seriously debating the topic. Use your forum for good!

We’ve had this conversation before Carl. As a so-called crossroad and “big umbrella” we post items that start dialogue such as the one generated here. That’s what we do.

S/F

Dave D.

Phil Ridderhof

Schmedlap: “Well then, I would reiterate: if that’s the case, a lack of “metrics” isn’t the problem. Perhaps proper planning has a role in the outcome. Planning has to precede metrics, not follow from them.”

Exactly. When “metrics” or assessment don’t seem to make sense its because there’s a real disconnect in understanding the mission. The arguments Ive read in these comments are not so much about whether we measure the correct things, but whether we have the right strategy or have set realistic and achievable goals given the means available.

I’m not so negative on all of the statistics being discussed. Its when we describe them as “metrics,” or try and let them speak for themselves that we lose their utility. Unfortunately, however, we tend to use quantifiable measures as the proof–that the tables can stand for themselves to demonstrate some truth about the situation.

Given that a COIN fight won’t feature clear seizure of terrain or attrition of combat units, you have to gain information about what is going on in some manner. The best is observation, reliable HUMINT, etc. Quantifiable measures are useful if you use them as a starting point to ask why, or focus your observation/analysis on apparent anomalies. As indicated in most of the Kilcullen passages on Tom Ricks blog, these statistics aren’t telling you that you are winning or losing, but they are telling you something about what is going on in the area. Each of Kilcullens observations about some measure is useful or not could only have been established because someone had to investigate further to find out the story behind the numbers. This is good and leads to a better appreciation of the environment and the cause/effect relationships at play. I noticed that the statistics he found most useful were not single counts, but comparisons between two or more. The correlation does not lead to an answer, but to a hypothesis of why there might be a relationship and further effort can be put forth to validate of invalidate the hypothesis–which leads to better understanding. Better understanding is ultimately subjective and that is something we have to learn to live with.

MikeF

PhilR:

“to a hypothesis of why there might be a relationship and further effort can be put forth to validate of invalidate the hypothesis–which leads to better understanding. Better understanding is ultimately subjective and that is something we have to learn to live with.”

Sir, I just typed almost that exact statement today while working on a paper about company level operations. I’ll try to expand a bit. Understanding one’s environment is never fully complete. We gather information through observation, conversation, reconnaissance, surveillance, etc, but there is no certainty in small wars- trust and truth are often elusive.

So, throughout operations (shape, clear, hold, build) we continuely develop hypothesis to test. Then, we conduct further missions to confirm or deny them trying to minimize the gap between what we think we know and what is actually going on. In some sense, we are continually in the first part of MDMP as our facts and assumptions may change or shift based off new information.

Occasionally, we reach decision points where the commander must state, “this is how I see the world.” Sometimes we get it right; sometimes we get it wrong.

v/r

Mike Few

Schmedlap

Mike/Phil,
It’s bad form to pimp one’s own blog, but I figure this conversation is slowly shifting to the council and will soon slip off the “recent comments” list, so fwiw, here is a lengthier response.