# Using Poisson Distribution Models

Poisson Distribution
In the perfect world, all the information you sample will be typically distributed so you could apply classic statistical investigation into your data. In reality, you’ll frequently collect data samples that don’t seem to be typically distributed. In most instance, the information may not appear to be generally distributed, but really is. This guide will show you the 7 most frequent and many causes of standard distribution data appearing to be completely normal. When all your data samples are generally distributed, you can normally use those well known parametric statistical tests such as ANOVA, or perhaps T-test, and use regression to apply to your resultant data sets.

So what are the options if your data doesn’t seem to be normally distributed? Well first of all you can look to apply some non-parametric tests to the information. These non-parametric tests do not rely on those inherent data to possess any particular distributions You should be able to assess if a non normal information was typically dispersed before it was influenced by one of the standard seven correctable causes well known to statisticians. If you are unfamiliar with these you can find more information on most statistical sites. Search for the 7 Correctable Reasons For Non Normality in Data Samples.

It doesn’t actually take much to create such situations even with small data sets. All you need is a few outliers and they can easily skew normally dispersed data. Depending on the studies you could seek to disallow these numbers but you have to be extremely careful. Ignoring valid and returned data due to expectations can lead to all sorts of issues and most statisticians would seek to avoid this.

There’s a great example of where you cannot ignore outliers in your data sets and that’s when you analyse games of chance.   Many individuals seek to make their fortune when gambling in casinos both online and real ones.  They figure that they can manipulate their chances by looking at raw datasets and looking for biases in these games.  Obviously computers are perfect for this sort of work and it’s much easier to analyse online games.   They feed the data into all sorts of software and roulette sim games and try and identify the bets to place.  However in games like roulette the outliers are crucial especially when trying to identify sequences.

If you ignore them you’re basically skewing all the data and it could lead to some expensive mistakes.

However if you do, remember to be very careful when you identify and eliminate outliers. Valid reasons for disregarding include those that are brought on by error in measurement or perhaps data information entry. In these situations you would be justified in removing them from your results. However sometimes you could be able to obtain typically distributed data in the skewed data set. Outliers should only be removed if a particular cause of their extreme value is identified.

James Collins:

Author of Are Online Casinos Bonuses Worth it

This entry was posted in casino maths, Probability theory, statistics. Bookmark the permalink.