Thursday, January 6, 2011

Lying about lying with statistics

I read this comment recently by an anti-gun type:

Pretty soon we'll get the 2010 statistics from which the pro-gun crowd can pick and choose those which support their irresponsible policies.  Maybe I can help.  Let's take all the "accidental gun deaths" and divide them by all the stars in the galaxy.  This will clearly show how insignificant the number is.

Statistics don't lie.  They are what they are.  If you were to literally divide the number of accidental gun deaths by the number of stars in the galaxy, you would get a rate (which is not a statistic) that has no meaningful interpretation.  This blogger's intention is to make the naive reader think that dividing accidental shootings by a large number is as dishonest as dividing them by the number of stars.  It's a false equivalence, a rhetorical trick. 

To get a meaningful rate, you start with a meaningful question.  No one asks "how many people per star in the galaxy get shot every year."  But it is meaningful to ask "what is the probability of the average person being accidentally shot in a given year?"  A question like this puts a rare, but serious, event into perspective.  And there's an easy answer. 

To calculate the probability of the average person getting shot accidentally, we just divide the total number of events by the population at risk.  According to this site, there were 642 accidental shootings in the US in 2006, with an at-risk population of 300 million.  Therefore the probability of getting shot accidentally is 642/300,000,000.  In percent, 0.000214%. 

Yes, this is a low number, but it means something very important that discredits a major argument for gun control: there's virtually a zero probability that the average person will be accidentally shot and killed in a given year.  It's an honest and clear calculation - no calculus, no regression analysis, just a meaningful ratio.  We didn't have to divide by the number of stars. 


  1. There is a percentage formula that I like so much more than the stars in the sky and the virtually zero probability formulas:

    0% guns x 0% gun related deaths = 0%

    And try justifing your way Mr. Virtually Zero to the parent, bother, sister or grandparent that just lost one of those .000214% children. The only thing that this proves is that you have very little, if any, value in our children. 1 child is 100% of the unneccessary deaths related to guns every year, of any year!

  2. Anonymous:

    Is a gun-related death any more tragic than any other negligent act that kills a child?

    Here's another formula:
    0 cars = 0 children killed by drunk drivers.

    Where does your banning of objects stop, if your goal is to prevent every untimely death?

  3. Cars are not intended nor built with the sole purpose of killing something. Cars have become a neccessity in our society (and I don't like that fact much either) where guns have become less of one. "Is a gun-related death any more tragic than any other negligent act that kills a child?" You asked. Yes,in my opinion it is. Because it's the only time someone feels the neccessity to explain away the death with stupid statistics that minimizes the importance of the child, like you just did. No death of a child is okay, but the controllability of guns is far more likely and reasonable then the control of cars, or even alcohol. But, you folks just don't get it....

  4. #8 on the 10 Leading Causes of Unintentional Injury Deaths, United States is: Unintentional Firearm Deaths of Children Ages 1-17 years (1999-2007 - The most up to date stats given by National Center for Injury Prevention and Control, a division of the CDC)=1,091 children
    Now lets use my percetage formula on that count: 0% guns x 1,091 children = 0% unintentional firearm deaths of children. You see how that can change things? If nothing more that moving those kids along to better stats like graduating from high school, going to their senior prom, being the first in there family to go to college, or just getting to their 18th birthday for God's sake. But like I said, you folks just don't get it....

  5. I have just a general question for you, answer if you choose, it doesn't matter:
    The Banner of your blog shows what appears to be 8 notched copper jacketed hollow point 45cal acp rounds, correct? In what competitive shooting compitition do they use a strictly anti-personel type round? One that's sole intent is to enter a body with a large mass that is designed to become larger as moves through the body to a point where the exit wound can become sometimes as much 300% larger than the entry wound?
    Lets say one of those rounds being accidently discharged from your firearm was to strike your 9 year old child (the mean age of 1-17 years given by the CDC) lets just say in the leg. In your expert opinion what would happen to your child? I know what would happen to your child! The leg would almost certainly would be severed which means the femoral artery would be severed as well. Before you could get a turniquet in place and certainly before you could get the correct medical help your child would bleed out. Your child would be dead! Then I would like to see you justify away his or her death by the zero probability rating you gave in that happening in the first place. It wouldn't work then and it doesn't work now.

  6. "Cars are not intended nor built with the sole purpose of killing something."

    Nor are guns. I've fired 25,000 rounds of pistol ammunition during the past two years in matches and practice. Not a single bullet has hit a body. If guns intended solely for killing, and my gun hasn't killed anyone after all those rounds, can I successfully sue Smith & Wesson for a defective product? Know any lawyers who would take my case? If not, then why?

    "#8 on the 10 Leading Causes of Unintentional Injury Deaths, United States is: Unintentional Firearm Deaths of Children Ages 1-17 years"

    So we shouldn't ban #1-7? If your goal was to save lives, you should focus on the most frequent reasons.

  7. Thank you for noticing my banner. It's a tribute to the greatest industrial design of the 20th century - John Browning's M1911, which holds - you guessed it - 8 rounds of 45ACP. It is the gun I compete with (USPSA single stack division) and to which I trust my life every day.

    You are right that the purpose of the hollow point bullet is to mushroom and stay inside of the first body it hits. The alternative would be ball ammo, which tends to pass right through and hit people behind your target.

    Hollow points also have greater stopping power, which means you can stop the person who's trying to kill you with fewer shots. Fewer holes = fewer bleeding sites, fewer major organs hit, safer for the one who's shot.

    This is why nearly all American police departments carry them, and it's why I carry them on the streets of Boston. Because they're safer for everyone involved.

    The probability of me shooting someone accidentally is conditional on how I handle my gun. If I follow the NRA's 3 Rules of Gun Safety religiously, then yes, my probability of accidentally shooting someone is zero, so I won't entertain your question.

    Pistol bullets only sever limbs in the movies.