See a Magnum Photos gallery on the capture of Saddam Hussein.
Staff Sgt. Eric Maddox did not want Izzat Ibrahim al-Douri on his chart. As a special-operations interrogator in Tikrit, Maddox had been building link diagrams similar to those created by Col. Jim Hickey's intelligence shop, adding names as he gathered information from detainees. As much as al-Douri, the red-headed confidant who many considered Saddam's second in command, was a highly wanted man—the King of Clubs on the Army's deck of cards—Maddox was convinced that he had long ago parted ways with his boss. Like Hickey's men, Maddox believed the path to Saddam depended on tracking down much lower-profile players in and around Tikrit. There was only one problem: The top command—Maddox's bosses' bosses—didn't agree, and wouldn't be pleased to see a link diagram without big names like al-Douri represented. To avoid conflict, Maddox made two versions of his charts: The one he and the analysts actually worked off of, and the one they displayed prominently any time someone important was around.
Maddox wasn't lacking in names to fill out his diagrams. When he joined the special-ops group in July, an interpreter showed him a list of all the former bodyguards who lived in the area, as well as their kin. Among many others, there were 40 Musslits on this list. This told Maddox two things: First, the Musslits were an influential family in the region. Second, Saddam Hussein had alot of bodyguards.
As he interrogated more and more people, Maddox discovered something curious. While some bodyguards were clearly mixed up in the insurgency, others seemed to have essentially retired. (One very high-ranking ex-bodyguard, Maddox learned, had been hanging around at home in plain sight since the war began.) Now he had to figure out which of Saddam's old protectors were still in touch with him—a task for which studying pre-war roles was not very useful. As Maddox writes in his memoir of that time, Mission: Black List #1, "Iraq was a completely different place than it was before the war."
Throughout the hunt for Saddam, Maddox and his counterparts had to make lots of subjective decisions about who—out of all these bodyguards and all these Musslits—was most significant. Lt. Col. Steve Russell had a hunch that the brothers Rudman and Mohammad al-Musslit were important based on their wealthy Tikrit estates, photos of them with Saddam, and intelligence reports that kept turning up their names. Later analysis by Maj. Brian Reed, the brigade operations officer who would write his dissertation on the link diagram, proved that Russell's intuition was backed up by network theory.
One way to measure the importance of a particular person in a social network is to look at his "betweenness centrality." A person scores highly on this measure if he's a connector—that is, if he's likely to lie on the shortest path between any two other nodes, helping form the connection between them. An individual doesn't need to connect directly to many nodes to have a high betweenness value; he may simply bridge two social circles, such that a path from one to the other will inevitably pass through him. In a network of my work and college friends, for example, my betweenness would be off the charts—I'm the guy who bridges the gap between two otherwise unconnected groups. Serendipitous connections can lead to high betweenness values as well. Just think about all the unexpected connections you've seen on Facebook. For instance, I have a co-worker who went to high school with a college friend of mine. This gives them both high betweenness scores in my network—even if I was out of the picture, co-workers could trace a connection to my college classmates through this link.
Below is an interactive re-creation of the core of Saddam's network. I've pieced this together from Reed's dissertation, public accounts, and interviews with the officers who were on the ground in Iraq. As you can see, Mohammad al-Musslit is clearly the vital link between Saddam and the entire Musslit clan. Rudman also scores highly as the chief of operations for the insurgency. To learn more about a person's role in the insurgency, mouse over his photo. To understand how people are connected, mouse over the line between them. (Family names are listed below given names.)
The calculations behind their scores involve some reasonably complex math, but the reason for the brothers' importance is clear: Rudman and Mohammad stand squarely in the path between Saddam and the extended Musslit family. Adding to their importance, the brothers also connect Saddam—via a marriage in the Musslit family—to a second clan that had a role in the insurgency. (The link diagram constructed by Reed and Maj. Stan Murphy, the brigade intelligence officer, is included in Reed's dissertation in a highly redacted form. I was able to identify Rudman and Mohammad in the diagram based on reporting and inference from surrounding nodes. I was not, however, able to identify the second family that the Musslit brothers connected up with.) Curiously, one of the other men in Reed's diagram with a high betweenness score had been dead for two decades when the 4th Infantry Division arrived in Tikrit. Even so, he was important to the network because he formed one of the few links between the Musslits and the aforementioned second clan that was also active in the insurgency.
Rudman and Mohammad, by contrast, were still very much alive. The network and his instincts were telling Maddox the same thing: These were the guys who would lead him to Saddam Hussein. But instinct alone, no matter how well-supported by a link diagram, is not a very solid basis for organizing a raid and putting American soldiers in harm's way. Maddox had to earn the trust of the special-ops commanders with proof that his diagrams reflected behavior on the battlefield. Maddox did make his share of mistakes—recommending raids that turned up no one and opposing raids that did. But as interrogations continued to confirm what the network had predicted—who was important, who wasn't—the team's commander and analyst began to take greater risks based on the network. Eventually, Maddox became a part of the team that planned and executed raids. It was an unusual role for an interrogator.