I am totally confused by this discussion and the use of the term 'statistics'.
There are descriptive statistics and inferential statistics. To describe and predict. People do act as intuitive scientists and make predictions.
One may not explicitly worry if one's wife will dismember them but one does usually have health insurance to pay for the cost of injuries. One may not engage in adulterous behavior or spousal abuse because one predicts that this usually doesn't end well. That's making a prediction.
Most of us don't wear body armor all the time as compared to officers. Most of us don't walk around with M4s as compared to soldiers in Iraq or Afghanistan. That's because we have a feel for the probabilities of a civilian in our environments needing either.
Formal statistics just quantify and test such informal views.
On fora like this, the statistics arguments usually center about:
1. How many mags do you carry?
2. Do you need a BUG?
3. Don't practice beyond 3 inches as gun fights are close up.
or something like that.
Then folks carry on about extreme cases - well, statistic methods are based on predicting the likelihood of such an outcome and the likelihood of an error if you don't take such an action. However, anyone with stat training knows that the cutoff and decision points for an action based on distributions is a subjective decision.
So when you say that you will carry one extra mag, even though it is rarely used, you are making a choice based on your view of the risk of not being ready for an extreme case. That is not outside our statistical methodologies. We evaluate the cutoffs based on such outcomes.
I see little here that does negate statistical methodolgies if correctly applied and understood. It is usually that someone wants to make a decision based solely on central tendency where these arguments go awry.
Well, a gun fight only has 3 shots on the average, thus I NEVER need more that a J frame with 5 - it's the implicit never which makes those kind of comments baloney.
I think we are smart enough to get beyond that. You plan for extremes if the risk of being in one is so horrendous that you take the inconvenience of planning for it. But these analyses are second nature if you really use statistics as they were designed.