Informed Comment - Statistical Analysis of Iraq's resistence movement

From the comments at Juan Cole's blog:

At 12:34 AM, CW said...

Juan Cole wrote: "Indeed, given the GIs being blown up on a daily basis, the Republican phony resolution was the equivalent of trying to do a stand-up comedy routine at the funeral of someone's beloved son who had died at age 20."

I am a very expert applied mathematician. I learned from veterans of WW II who pioneered the advances in the use of mathematical and statistical models in military activities and I have applied the methods and techniques in a variety of domains. These types of models are now some of the crown jewels of our modern military as they allow for precise planning and deployment.

When US military casualty losses hit 1,000, I began analyzing them. I had good technical results and recently added another 13 months of data to re-evaluate the model and I am close to completing that project. However, here is what seems certain. The rest is technical nuance.

1. The resistance began in earnest in March 2004. The chance of having a day with no fatalities was cut in half and has remained constant ever since that time.

2. There have been four two week periods and one three week period that can be described as a "lull." Best guess is that commanders in the field just kept the troops in the garrison--if you are not on the road, you can't hit a roadside bomb. I have not tried to look at press reports for explanations.

3. There have been about 20 extraordinary days that result from rare types of events like a helicopter crash or suicide bomber or from pitched battles.

4. The statistical evidence corresponds to reports that the roadside bomb technology gradually improves.

5. There is no evidence that the daily casualty rate has leveled off, let alone decreased. It just creeps up over time.

This means that despite the best efforts of US troops to protect themselves and their bases, that is all they can do, which is quite an accomplishment. The resistance is just as persistent and also improves their weapons and practices. A deadly balance is reached although it is reasonable to expect an increase in annual fatalities with each passing year.

Perhaps that will change. I do not know. What I do know is that there is no sign of improvement. The statistical significance of the results that support the above statements is such that the odds are less than 1 in a 1000 that they reflect a chance outcome.

Informed Comment


Post a Comment

Links to this post:

Create a Link

<< Home