Prediction

One of the reasons that statistics is listed after "damn lies" is that it's not very good a predicting specific outcomes.

The Monty Hall problem is famous.

Given three choices, you pick one.  One of the choices you did not select will be revealed to not contain the prize.  Should you change your selection to the remaining box or stick with your initial choice?

Statistics says change your selection.

Mythbusters even did an experiment that proved you win more often if you change.

The thing is, this doesn't predict if any given iteration of the game is a winner or not.

Example.  Prize is in box A.  Contestant choses box A.  Box C is shown to not have a prize in it.  Changing your selection in that case is a loss!

This is also why the plural of anecdote is not data.

Even though I have a better chance of winning by changing my choice that doesn't mean I am guaranteed to win by changing my choice.

Another example is from crime statistics.

A disproportionate number of crimes are committed by black people.  That statistic does not provide an accurate predictor that a specific black person is about to commit a crime.

Most murders in the USA are committed with a firearm.  That statistic is not an accurate predictor that any specific firearm will be used to murder someone.  The records that are kept  only generally predict what kind of firearm will be used.

This part of statistics is why I get the RCB when I pay my car insurance.  I am a man, I have a Corvette, I am under 65.  Statistically speaking I speed, drive recklessly and have accidents.

I've gotten speeding out of my system.  Track days do that.  I've gotten reckless out of my system.  Being DQ'd for being reckless on the track and losing the rather spendy entry fee really mellows you, besides, when you slow the heck down you have less NEED to be reckless.

I have not had an accident where I was at fault since REFORGER 1988.  In my defense, running over the parked car was preferable to running over the child who ran out in front of the tank.  Darwin was sleeping that day.

Yet my insurance rates are based on the behavior of the worst of others not on what I do.

I am so very sick of being judged by the behavior of others.