Physicist: Predicting Insurgencies Is Easy. As Long as You Dumb Down Your Wars

Insurgencies are easy to predict – no matter where they occur, or why they begin. You just have to assume that all militants care about is appearing on CNN. And that everything there is to know about an insurgency can be found in your local newspaper or in military press releases.

That's the assertion, at least, of New Zealand-based physicist Sean Gourley. The last time we met him, Gourley (pictured above, center) had come up with a tidy-looking equation to explain the chaos of war. Unfortunately, that formula didn't actually work. Last summer, Gourley admitted that he failed to accurately predict the outcome of the 2007 military surge in Iraq – and that his predictions sprang from some rather dubious data. Gourley's formula relied solely on famously sketchy media accounts of insurgent attacks.

Turns out, Gourley was just getting warmed up. Last month, he and his research team published a new paper, "Common ecology quantifies human insurgency," that made it onto the cover of Nature, one of the world's leading science journals. It features an expanded version of Gourley's original formula, and is founded on the idea that insurgencies are "an ecology of dynamically evolving, self-organized groups following common decision-making processes." But strangely, Gourley is repeating – and, in some cases, amplifying – the same missteps he made before. He's still basing his models on press reports and other transparently incomplete data. Now, to make matters worse, he's claiming that militants are fighting just to get that media spotlight.

Gourley and company collected data on 54,679 "violent events" reported in nine different conflicts, including those in Iraq, Afghanistan, Peru and Colombia. The selected events were mostly fatal, because, apparently, "injuries are harder to cross-check." After finding similarities between insurgent attacks in different conflicts, the team came up with a mathematical "feedback loop" model, based on two variables: "global signal" and "internal competition." In Gourley's words:

'Global signal' is an input that each group has access to... This can be understood as traditional news being broadcast (i.e. CNN, AJ etc) and the competition is then a competition for peoples attention. This attention acts to increase the supply of resources to the successful groups - more people join this insurgent group, more money is directed towards it, etc.

Gourley says this "global signal" could, theoretically, be anything that acts as "the primary input for the group’s decision-making process." But for today's militants, at least, media attention is that input. Gourley asserts that insurgent groups commit attacks on quiet days, wait for press coverage, and then plan their next attack based on that coverage and how other insurgent groups react. Which means the model assumes that insurgent groups are mostly interested in making a splash:

Mechanism (2) is consistent with comments by former US Senior Counterinsurgency Adviser David Kilcullen, who noted that when insurgents ambush an American convoy in Iraq, ‘... they’re not doing that because they want to reduce the number of Humvees we have in Iraq by one. They’re doing it because they want spectacular media footage of a burning Humvee.’

Taking into account the size and strength of a particular insurgency and counterinsurgency, and the timing in question, Gourley's team then claims they can predict when an insurgency will strike, and whether or not American troops stand a chance.

But there are a number of king-sized holes in Gourley's argument. First, it assumes that the insurgent is the only one capable of making a first move in an irregular war. But, of course, the counterinsurgent can also strike first. What does that do to Gourley's CNN-driven model of conflict?

Second, as others have already noted, using war reporting to make predictive models is a problem: Coverage, from the media or other sources, is notoriously inaccurate or incomplete. Comparing reportage from one conflict to another is even worse: In Afghanistan, for example, rural environs and rugged geography have led to relatively less coverage than the Iraq conflict. Gourley's team tried to compensate by pulling reports from diverse sources, like media in Afghanistan and a Catholic NGO in Columbia. But that likely created more inconsistency, rather than less.

Then there are nonfatal attacks, whether on people or infrastructure, which are mostly unaccounted for in Gourley's model. He admits that "there will inevitably be some estimations made," but remains convinced that there's no need for exhaustive, accurate data. "We don't need to have complete coverage, we need reasonable coverage."

But garbage in means garbage out, warns warfare authority T.X. Hammes, now a senior research fellow at National Defense University. "They are doing statistical analysis on data with very little control of the quality or even applicability of the data," he tells Danger Room. "Does an attack on a policeman qualify as an attack even it if it was because of a personal issue – while the attack on a civilian doesn't?"

Gourley's model to crunch the data has problems, too. For him, each insurgency has only one replaceable variable: that "global signal." It could be media or religious holidays or weather forecasts – whatever is the insurgents' primary information source. But Gourley's militants are homogeneous, and motivated by one thing, and nothing more. Gourley's wars are just single-variable cycles, playing on repeat. Forget Iraq's religious and tribal divides; never mind Afghanistan's mess of local warlords, jihadist zealots, and Taliban restorationists. This physicist's model for war is clean and simple. If only real conflicts were so easy to predict.

Photo: WhiteAfrican/Flickr

See Also:

• Physicist's Fool-Proof War Formula (Just Add Media Accounts)
• Don't Hold Your Breath for 'Sim Afghanistan'
• NATO Wants Sim Afghanistan to Test War Plans
• “Sim Iraq” Sent to Battle Zone
• Predicting Terrorist Attacks