As Congress prepares to hear testimony from Gen. Petraeus on the situation in Iraq, the White House and Pentagon have been pointing to several statistics that they say show progress as a result of the surge. Some military experts, however, say those numbers only tell part of the story.
Sometime around February 2004, a top military official in Iraq estimated that there were about 15,000 total insurgents. About a year later, U.S. military leaders in Iraq announced that 15,000 insurgents had been killed or captured in the previous year.
In private, a skeptical military adviser pointed out to commanders that the numbers didn't make sense. "If all the insurgents were killed," he asked, "why are they fighting harder than ever?"
The adviser, who couldn't speak on the record, recounted the story as an example of how statistics can easily become misleading.
Here's a few statistics that military officials have cited in the past few days.
From Gen. Richard Sherlock: "Overall violence in Iraq has continued to decline and is at the lowest level since June 2006."
From Gen. Kevin Bergner: "On a national level, sectarian deaths are about half of what they where in December of 2006."
And from Gen. Ray Odierno: "Total attacks are on a monthlong decline and are at their lowest levels since August of 2006."
And here's how those statistic translate into political rhetoric.
From South Carolina Sen. Lindsey Graham: "Well, the surge has worked; it's provided a level of security I haven't seen."
And from President Bush: "Anbar is a huge province. It was written off as lost. It is now one of the safest places in Iraq."
But other numbers tell a different story.
This year, Anbar is actually the second-deadliest place for U.S. troops in Iraq. Baghdad is the deadliest.
And while there's no doubt the numbers of troops killed in Anbar this year is lower than last year, troop casualties have spiked dramatically in other provinces.
Twenty American service members were killed in Diyala Province last year. So far this year, 100 U.S. service members have died in Diyala. Every month this year, more American troops have been killed as compared with the same month last year.
Pentagon officials argue that numbers like these are meaningless, that they don't give a sense of success. The problem, though, says former Army Col. Doug MacGregor, is that the Pentagon uses statistics selectively to bolster the case for success.
"People are making claims and assertions that don't stack up when they are viewed in the context of the last four years," MacGregor says.
Here's an example: The Pentagon says sectarian deaths in Iraq were sharply down in August. But the military's definition of what constitutes a sectarian murder is narrow.
Last month's massive bombing in northern Iraq that killed more than 500 ethnic Yezidis made August 2007 the second-deadliest for Iraqi civilians. Yet the Pentagon doesn't consider large bombings like that one an example of sectarian violence. The result is that it can show that sectarian murders are down.
"What we have right now is an illusion created by the White House, created unfortunately with the help of many people in the media," MacGregor says. "And the result is, people pick up on what is said [and] it becomes conventional wisdom."
The military measures stability in Iraq by looking at total attacks daily — attacks on U.S. troops, Iraqi forces and Iraqi civilians. The Pentagon says total daily attacks are now at a one-year low. But last year was the deadliest for Iraqis since the invasion, so the comparison, says retired Army Col. Paul Hughes, is somewhat misleading.
"Even with the security that's improved in the Baghdad region," Hughes says, "they are still not getting the electricity and the water that city's citizens need."
Before the war, Baghdad had round-the-clock electricity. Today, more than four years since the invasion, the city averages about six hours of electricity a day.
And then there's the issue of Anbar province. Both the White House and the Pentagon have attributed the changes in Anbar to the surge strategy. But several military advisers who worked in Iraq until late last year have said that is simply not true. MacGregor says that the increasing cooperation between U.S. forces and Sunni tribes in Anbar started more than 18 months ago, long before the "surge."
"And they were done on the initiative of the Marines and the Navy who looked at Anbar and said, "There's gotta be a better way to do business here," he says.
So is the surge working? The short answer is that no one can know for certain because statistics only tell a small part of the story.