The Numbers Simply Don't Match Up... Posted by Adam Blickstein

A graph (which is pasted below) Petraeus presented today seems to paint a diametrically opposite picture for March in terms of Iraqi civilian casualty numbers than what has been broadly reported. The graph suggests that civilian casualties in Iraq stayed somewhat level during March, and in fact the coalition numbers seem to demonstrate a dip in overall civilian deaths. This clearly doesn't jive with what has been broadly reported. The Washington Post, for instance, reported on the deadly month that was March as thus.

A total of 923 civilians were killed in March, up 31 percent from February and the deadliest month since August 2007, according to data compiled by Iraq's interior, defense and health ministries and obtained by Reuters.

Petraeus' graph clearly doesn't show anything close to a 31 percent spike in March from February, no where near the 923 deaths in the Post report, instead showing a virtual plateau. It also clearly doesn't put civilian deaths in March 2008 on par with August 2007. Of course, the important question for Petreus is: why the discrepency?