This is the graph I normally color in blue, so I'm sure to cause unlimited confusion if anyone goes back comparing this graph with the older ones. Blame the malicious attacker.
In assessing the numbers from February I projected lower numbers for March based on an extrapolation of the early number from March. That's obviously a tenuous piece of data on which to build a projection--March deaths ended up higher, helped in part by violence from the Mahdi Army. The conflict with the Mahdi Army has persisted through the present day, though the coalition effort against the sectarian militia has been very effective. Though the early numbers for May project to higher totals than for April (54 deaths), I'll predict a lower number based on the fading resistance from the Mahdi Army and the increasing capability of the Iraqi security forces. Despite the somewhat higher death toll on U.S. troops, the overall news from Iraq has been good.