# USA TODAY can't do arithmetic (rate of National Guard fatalities in Iraq) **URL:** \ **Category:** The BBQ Pit\ **Created:** [December 13, 2004, 3:32pm UTC](https://boards.straightdope.com/t/usa-today-cant-do-arithmetic-rate-of-national-guard-fatalities-in-iraq/279610 "2004-12-13T15:32:10Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![RTFirefly](https://avatars.discourse-cdn.com/v4/letter/r/c77e96/32.png) [@RTFirefly](https://boards.straightdope.com/u/RTFirefly)\ **Post date:** [December 13, 2004, 3:32pm UTC](https://boards.straightdope.com/t/usa-today-cant-do-arithmetic-rate-of-national-guard-fatalities-in-iraq/279610/1 "2004-12-13T15:32:10Z")
[From the McPaper](http://www.usatoday.com/news/world/2004-12-13-guard-deaths_x.htm): > [@](#): > > In a reversal of trends from past wars, part-time soldiers in the Army National Guard are about one-third more likely to be killed in Iraq than full-time active-duty soldiers serving there, a USA TODAY analysis of Pentagon statistics shows. > > According to figures furnished by the military branches, the active Army has sent about 250,000 soldiers to Iraq, and 622 have been killed. That works out to one death for every 402 soldiers who have deployed. About 37,000 Army Guard soldiers have been sent to Iraq since the war began and 140 have died there — one fatality for every 264 soldiers who have served, or about a 35% higher death rate. That’s not how you do the arithmetic, dudes. For the active-duty Army, 622/250000 = .002488, or 2.488 deaths per 1000 troops. For the Army NG, 140/37000 = .003784, or 3.784 deaths per 1000 troops. 3.784/2.488 = 1.52. IOW, the death rate amongst Army NG in Iraq is 1.52 times the death rate of regular Army in Iraq, or 52% higher. Not 35% higher. One more example of innumeracy for [John Allen Paulos](http://www.math.temple.edu/~paulos/)’ file. :rolleyes: ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [December 13, 2004, 3:41pm UTC](https://boards.straightdope.com/t/usa-today-cant-do-arithmetic-rate-of-national-guard-fatalities-in-iraq/279610/2 "2004-12-13T15:41:20Z")
> [@RTFirefly](#): > > [From the McPaper](http://www.usatoday.com/news/world/2004-12-13-guard-deaths_x.htm): > > That’s not how you do the arithmetic, dudes. > > For the active-duty Army, 622/250000 = .002488, or 2.488 deaths per 1000 troops. > For the Army NG, 140/37000 = .003784, or 3.784 deaths per 1000 troops. > > 3.784/2.488 = 1.52. IOW, the death rate amongst Army NG in Iraq is 1.52 times the death rate of regular Army in Iraq, or 52% higher. Not 35% higher. > > One more example of innumeracy for [John Allen Paulos](http://www.math.temple.edu/~paulos/)’ file. :rolleyes: Ugh! Newspapers can _never_ get their percentages right, it seems. It’s clear where the error came from: (3.784-2.488)/2.488 = 52% (3.784-2.488)/ **3.784** = 34% In other words, the death rate of the regular army is 34% lower than that of the reserve army, but the death rate of the reserve army is 52% higher than the regular army. This divisor error happens all the time. pulykamell, who used to be a copy editor in college and saw this mistake more times than he can count ---
**Author:** ![RTFirefly](https://avatars.discourse-cdn.com/v4/letter/r/c77e96/32.png) [@RTFirefly](https://boards.straightdope.com/u/RTFirefly)\ **Post date:** [December 13, 2004, 4:16pm UTC](https://boards.straightdope.com/t/usa-today-cant-do-arithmetic-rate-of-national-guard-fatalities-in-iraq/279610/3 "2004-12-13T16:16:04Z")
> [@pulykamell](#): > > It’s clear where the error came from: > > (3.784-2.488)/2.488 = 52% > (3.784-2.488)/ **3.784** = 34% > > In other words, the death rate of the regular army is 34% lower than that of the reserve army, but the death rate of the reserve army is 52% higher than the regular army. > > This divisor error happens all the time. That’s it, alright.