This is the bell curve.

This is the bell curve summary

Remember one thing about statistics, if nothing else; they are drawn up from a sample that the creator of the statistics chooses. Some people (GASP!) choose to use misleading info when creating statistics. It doesn't mean that the use of statistics is a cheat, it's just that you need to be aware of WHO is giving you stats, and WHY they are doing so.

I'm not going into a summary of statistics, much of which I don't recall off hand anyway, but I will give a short description of three concepts: the bell curve, the norm, and outlliers.

• The bell curve is a graph used to show normal distribution. The freaky thing about statistics is how often the normal distribution can represent measurements...it is fairly common to have the normal distribution match your set of measurements. I guess that's why they call it normal.
  The graph is read by looking at the value of the measure going crossways, and the amount of measures with that measure going up. In the middle, you see there there are the most things with the normal measurement.
  In any case, most of the info falls into the middle of the graph, with extremes of either higher than normal or lower than normal becoming fewer and fewer as you get away from the middle.

• That area in the middle is called the norm. Most measures of data end up here. Normal does not mean good, or right, just because it's normal. It just means that the most of the thing being measured falls into the same range of value.

• Outliers lie outside the norm; these are the exceptions, the oddballs, the one in a million shot. Outliers are where shit happens or where somebody has hit a home run.

Quick example. IQ tests. Most people have an IQ of 100, or pretty close to it. But as you get farther away from the norm, the fewer people that are dumb or are smart there are. The second part of that is that there are pretty equal measures of really dumb and really smart people.

Average people are normal. Really dumb people and really smart people are outliers

This is the where decisions are made based on bell curves

Most of the time, rules, strategies, regulations, policies, and all the similar malarkey are made using statistics to make judgements or predictions and to measure the results of whatever it is you're doing.

That is where the first problem occurs.

A lot of decision making is based on what happens in the norm. Maybe you are trying to increase the amount of the norm, or make things better for the people already in the norm. What happens to the outliers doesn't really matter in these decisions.

But from my observations, it is in the outliers that problems are created. In fact, I harp all the time about how there is usually more than one cause to a problem.

I say that there are intersections of outliers, and this are where things hit the fan the messiest. Consider the guy born with an over-aggressive biology, abusive parents (or lack of parents), and a hostile environment:

At the simplest level, when these folks use these methods to help them make decisions, they are looking at one cause and seeing if it has one effect.

This is easy to do in a laboratory when you are mixing two chemicals together in a sterile test tube; it is not too easy when you have social difference, multiple possible causes, political interference with the proposed changes, or any other thing that may effect whatever it is you are trying to to, especially when you don't know what some of those things may be (unknown unkowns)

The other side of this is that if you target your decision to "fix" outliers, you are likely to have negative effects on the norm.

Let's look at DUI checkpoints. How many Saturday night drivers are shitfaced to the point of danger? Is is worth it to shut down everybody to get El Drunko off the road?

Just some things to keep in mind when somebody gives you statistics to justify whatever scheme they are trying to sell you ;>