Lessons in Offline Metrics: The Body Count
Of all the metrics in the world, none is perhaps more chilling, or more somber, than the body count. The number itself is a testament to perhaps the most despicable aspect of mankind’s nature: man’s willingness to destroy others of his own kind, via combat, in order to advance his cause. Reducing the destruction of a human life to a number, and then tallying that number in order to keep score in war, as if warfare were a game of baseball, and each life lost was a run scored, seems an unpalatable debasement of everything mankind holds dear. Or should hold dear.
And that, as they say, is not all. Not only is “the body count” a decidedly unpleasant metric, its value is ambiguous, at best. Why, then, is the US command in Afghanistan turning to the body count to help report on the tides of war in southwest Asia?
Body count as a warfare metric has a long and storied history. As Michael Phillips, writing for the Wall Street Journal, put it, “enemy death tolls have been a feature of war ever since armies stuck heads on pikes. They appear in Thucydides’ History the Peloponnesian War, and in the Old Testament, which enumerates the casualties of King David’s wars, including 360 Benjamites, 18,000 Edomites, and 22,000 Arameans of Damascus.” The body count figure, of course, is not entirely devoid of value. Every enemy soldier killed is an enemy soldier that can no longer wage war against you, which, in a war of attrition, means the side that can replace soldiers the fastest will ultimately prevail.
The body count, however, accounts for far less when the control of territory is at stake—which, of course, is often the case in warfare. In both World Wars of the 20th century, for instance, the metric which most clearly identified success wasn’t the better body count, but the more territory held. During World War II, for example, the high point for Germany’s Third Reich was marked by its ability to control most of Europe’s territory from the Atlantic seaboard south to the Mediterranean Sea, and to the east as far as the suburbs of Moscow. Only when the Allies were able to press the Germans from the west, south, and east, regain that lost territory, and press on to Berlin itself, did the tide of the war, in fact, turn in the Allies’ favor. If World War II were to be measured in terms of body count alone, surely the Soviet Union’s tally of 10.7 million military deaths would appear to make it a loser to Germany, which suffered 5.5 million military deaths. Clearly, and thankfully, this was not the case.
During the Vietnam War, however, where the front lines were blurred, and the taking of territory mattered little as to who was winning and who was losing the war, the body count emerged as the primary metric by which victory was measured. Theoretically, the notion makes sense: in a war where the enemy was not always clearly discernable, perhaps the only way to measure success or failure was to quantify the destruction of those keen to shoot back. Surely then-Secretary of Defense Robert McNamara sought something on which to hang his hat, and as an accountant, economist, and highly-successful businessman, the raw statistical power of the body count metric must have given him a great deal of comfort as a way to quantify success.
Only, as it turned out, the body count didn’t quantify much of anything. Statistics may be useful for predicting the production run of a Ford assembly plant, but statistics often do not take into account the human conditions that create them—especially on the battlefield. Humans, after all, are human, and prone to game the numbers when they can, if by gaming the numbers successfully they can reap some sort of advantage. During the Vietnam War, American commanders in the field tended to overstate their body counts, often because they didn’t have the means to be able to identify, exactly, what constituted a confirmed enemy kill. Why? Because Vietnamese strategy included concealing, from the Americans, the number of combatants they lost to hostile fire. Over time, over-inflated body kill tallies meant less and less because despite the high numbers, and the seeming disparity between American soldiers killed and Vietnamese soldiers killed, the Vietnamese were still able to field a formidable fighting force. And the American home front began to question exactly what they, as a country, were reaping for such an astronomical price.
Thus, in the post-Vietnam era, the body count, as a metric, was relegated to irrelevancy. “We made a deliberate decision to stay away from body counts and not get caught up in that Vietnam predicament,” said Col. Greg Julian, USA, as quoted by the Wall Street Journal. Besides, large-scale 21st century military operations like the invasion of Iraq in 2003 readily leant themselves to more traditional metrics for success, like territory seized and held. Only when the Iraqi insurgency intensified later on, after the entire country of Iraq, it seemed, was pacified, did the means by which to measure get fuzzy again. From the ashes arose the body count.
Even more so in Afghanistan. After the fall of the Taliban regime in 2001, American and allied forces settled down to conduct what it thought would be a smaller, low-intensity war against the few remaining holdouts. As is often the case, however, it turned out to be something else, and by 2007, the US and its allies found themselves in an all war with a refreshed and renewed enemy. To win the war, the US and its allies realized that it had to win for itself and the Afghani government the battle of legitimacy. To show legitimacy in Afghanistan, the US-led coalition forces had to prevent the enemy from having the means to claim victory credibly, and the way to do that was to provide a body count: first only for major engagements, and then, as the war progressed, even for smaller engagements that resulted in the death of even one insurgent.
Whether or not employing a body count to deny the enemy the ability to claim victory will be successful in Afghanistan is a matter for history to decide, but already, the experience of arriving at an accurate body count number in Afghanistan is again re-enforcing the dubious nature of the body count as a reliable metric. To win their battle for the hearts and minds of the Afghani people, Afghani insurgents are inflating the number of civilians killed and understating the number of insurgents killed. In contrast, the American-led coalition forces are understating the number of civilians killed and inflating the number of insurgents killed. Gaming the system is simply the nature of the beast; the true number, of course, lies somewhere in the middle, hidden by the fog of war.
On the web, life and death and the fate of nations are not at stake, but the notion of a body count in warfare, nonetheless, provides valuable insight on the importance of metrics and measurement. What you measure is of vital importance, and how you measure it is equally important. To win online, make sure what you measure actually maps directly to a project’s goal; the goals established at a project’s onset to measure success or failure. And once those metrics are clearly defined, they must be able to be measured objectively. A good metric leaves little room for ambiguity; it’s either one thing or another, and not something completely different when looked at by another party.
To define success online reliably, the metric for success shouldn’t be open to interpretation. If it is, just as in war, the result may very well be a quagmire in which there are no clear winners and no clear losers.
By Robert Pothier