If we want to bring "reason" in here, it is reasonable to suppose that with some given failure rate, and assuming those failures caused ordnance to go off target, that failures and associated casualties would have been distributed evenly through the bombing campaign, on random areas, and not fallen on two identical types of highly-populated targets within a short time of each other long after the campaign had begun.

Statistically, there should have been many more such incidents during the initial "Shock and Awe" campaign of the first night.

Interestingly, since the rate of antiaircraft fire is roughly consistent with the rate of bombing, this should also have been true with off-target Iraqi ordnance.

Additionally, the physical evidence at the first site is inconsistent with a bomb of the type we have been using that had suffered guidance failure. I don't know if it's inconsistent with a missed antiaircraft missile.

I've drawn my own conclusions...but am waiting for more "incidents."