I think this is an important point when talking about the accuracy of war casualties. There are lots of things that the U.S. military can lie about, but one thing that can't be covered up is the number of casualties among the regular forces. Between the reporters "embedded" in these units, and the people waiting for them back home, it does not serve the military's interests to lie about the number of casualties. It's an all-volunteer army, and every person going there knew there was a chance they wouldn't come back, and signed up anyway.

Now, they could lie about where and how they died, and they probably don't report on any deaths in Special Forces that are operating secretly. There may also be a day or so lag in the official conformation of combat deaths. But you can be sure that the political price of lying about the number of casualties is so great that the military is better off telling the truth about the raw numbers.

I saw a rather telling statistic on CNN, which unlike a lot of their "reporting", does seem like it rings true. In previous wars (WWI, WWII, Korea, Vietnam), roughly one out of every fifteen boys we sent over did not come back. In the First Gulf War, roughly one out of every 1500 troops did not come back.

When you run the numbers of how many casualties people are "expecting", it is closer to the 1-in-1500 proportion. However, this is a different war than Gulf War I. We are actively engaging the Iraqis in a number of different cities, and quite possibly engaging in Urban Warfare. It's unrealistic to expect that we would have the same casualty rate as Gulf War I.

If 1-in-15 of our troops do not come home, like other wars, we should expect 17,000 combat deaths! The final answer won't be that high, if for no other reason than this war will (hopefully) not last as long as those. Still, faces with numbers like that, losing 100 troops at once doesn't seem that far fetched...