Quote:
|
This is an outlier example even for LE shootings.
|
This is an interesting discussion, but all the information provided seems to be outliers. As a retired financial executive that was responsible for financial forecasting (for large corporations) one of the things you find out very fast is that you have to be better than the local weather man in making your forecasts for the year ahead.
So you use empirical data (history) but as many have mentioned you throw out the outliers whether it is 10% above and below or 20% above or below your bell curve. The other criteria is that the data sets have to be for the same information, apples to apples or oranges to oranges, comparing apples to oranges just does not give good results. The third criteria is that the data sets have to be large enough to provide a statistical probability (80%) of correctness. A data set of 100 points is definitely not usable sometimes a data set of 1,000 points is also too small depending on the population (history) of the graph like when forecasting $150,000,000 in sales.
So of the 350,000,000 people or so here in the USA, at what distance and how many will have to defend themselves tomorrow I really can't say from the information provided, but I can tell you we will have 5 inches of snow tomorrow (LOL).
Jim