Originally Posted by cipher
That study is pretty ridiculous. Statistics that are formed to quantify something that cannot be done is done so with an agenda. That is fine. I know goes both ways. The study cannot calculate how many lives are saved correctly and how many crimes still would have happened had guns not been available. It only knows how to quantify the bad. It is a ridiculous attempt to quantify such a comparision, and the idea that it is not even close is just the same. Sorry, it just can't be known. I still disagree.
DISCLAIMER: Didn't see @cipher's second post before clicking post on this one...
You're entitled to disagree obviously. But just recognize that your belief is driven by your priors (subjective belief, ideology, faith, whatever you want to call it). How strong would the evidence have to be for you to believe having a gun in the home is more danger to the family than otherwise? If it were 100:1? How about 500:1? If there were 300 suicides, 100 accidental deaths, and 100 wife murders for every documentable use of a gun in home defense, would you still believe more lives are saved by having guns in the home? I say this just to make you consider. Even with the real 22:1 ratio of measurable incidents, you have to belief there is an absolutely massive and unmeasurable number of lives saved to discount the statistics.
Consider a different case. How about smoking? Like with guns in the US, it is impossible to truly know the potential outcome in an alternative universe, and you can't do an experiment. Like your argument about guns is factually and logically possible, it is factually possible that there are genetic or other factors that both make people more likely to smoke and more likely to get lung cancer (i.e., confounders). However, the evidence is so overwhelming in the other direction, and so many different analyses with different types of statistical control agree, that pretty much everyone, except maybe tobacco company execs, agrees smoking causes lung cancer even though we can't experimentally prove it.
As an (I think) instructive and interesting aside, there's a case in statistics and experimentation that sort of reverses the typical roles, where more typically liberal and science minded folks don't like to believe the strong evidence:
Longish Paragraph on Statistics of Psychic Abilities (Click to show)
There's been a long history of experimental study of essentially ESP, or psychic abilities of some sort. In fact the CIA had a long running research program looking into whether they could use psychics in intelligence work in the cold war! If you look at all the studies that have been done, restricting on stringent experimental controls, the overall evidence that there are in fact people with slight but real psychic abilities is as strong as the evidence that daily aspirin helps reduce heart attacks. But most scientists will accept that the collection of aspirin studies convincingly shows that it helps while questioning the existence of psychic abilities. There have been a series of interesting statistical papers exploring why this is using Bayesian statistics. The basic conclusion is that you can show mathematically how strong someone's prior belief that psychic abilities don't exist must be in order for them not to believe the evidence. I don't remember the exact numbers, but it was something like, if you're agnostic and think there's a 50/50 chance that psychic abilities are real, then the evidence will convince you that there is a very low probability that the psychics are getting lucky. In other words, the evidence will convince you that we can be quite sure that some psychics have small but real abilities. On the other hand, if you start with 95% or 99% certainty that psychics are all frauds, then the current evidence will only make you, say, ~90% certain that psychics are fraud. In other words, despite the evidence you'll still be quite sure that psychic abilities are not real. For the record, I saw a presentation by one of the leading statisticians in this sub-sub-field and came away convinced that at least some psychics probably do have relatively small psychic abilities.