The last time we commented on Trying to Grok’s statistical analysis, the results weren’t pretty. We thought there might be hope however when we found out that TTG’s management was enrolled in a stats class. We were wrong:
I feel like I should say something about the 1000 servicemembers who have died in Iraq […] Should I point out that, statistically speaking, one is more likely to die in a car accident — as my boss’ son unfortunately did over the weekend — than in Iraq?
Gentlemen (and ladies) — start your calculators!
Let’s start by looking at casualties resulting from transport accidents in 2001. There were 47,288 such deaths, approximately 129/day.
Since the start of Operation Chalabi Freedom, 1,002 US soldiers have died, a rate of 1.86/day (or 678/year.)
Let’s say that, to get numbers we can compare, we divide the respective population for both groups by the number of annual casualties.
The US population totaled 285,093,813 in 2001, which gives us a chance of dying from a transport accident for that year of 1 in 6,029. (See also here.)
US forces in Iraq have fluctuated, but let’s take 150,000 as our estimate. This would give us a chance of dying in Iraq for a year of 1 in 221. [If there were as many victims of transport accidents in the United States as US casualties in Iraq, (relative to population size of course,) 1,290,017 Americans would die every year on the roads.]
We’ll grant you that we too have struggled with our stats classes, but we’re pretty confident that 1 in 221 is not better than 1 in 6,029. Given our choice of odds of dying, put us down for 1 in 6,029.
In the comments over at TTG, Mike offers his support:
All you need to do is add up all the highway deaths in the US, and divide that by the population. Then compare it to 1000/200,000 (or the total of all who have served in Iraq).
So let’s see: Highway deaths (2003🙂 43,000. US population: 285,000,000. Rate: 0.00015.
US casualties in Iraq: 1,000. US troops in Iraq (per Mike:) 200,000. Rate: 0.005.
Which is the bigger number? And why are we comparing something other than yearly figures? (It really should be 0.00015 v. 0.0034 using Mike’s 200,000 for US troops.) TTG adds her efforts:
Mike, I was getting ready to do all the math, but you beat me to it and I’m way too grumpy after a long day of work to do more probabilities. […] The actual figures are hard to find, both for car accidents and for troops downrange.
Yeah, those car accident figures are very hard to find. Very.
We should also add this: if you can’t find the numbers, then why the fuck are you opining on which probability is higher?

