Psycho A Poisson distribution is a probability distribution of predictable events that happen at unpredictable intervals. Many accidents of various levels of complexity are of this kind and you can estimate the likelihood based on empirical data from the best possible sources comparable to your own, e.g. in the same sector, similar age groups and gender distribution. Programs like Excel and Minitab provide straightforward calculating routines: all you have to do is to input the key parameters. If you need help with a practical example, you're welcome to p.m. me.

In other words it is the bell shaped curve.............. If you have a body of data, of information it will be clustreerd around an average value. Thus, for example taking the height of adult males (numbers are representative - not accurate)... Number on vertical axis and height on the horizontal There will be a cluster around the average height of say 6ft (I'm an old fashioned Imperial person!) There will be fewer and fewer people stetching out towards 6ft 5 ins, then 6ft 6 ins etc Similarly there will be fewer and fewer people dropping down to 5ft 5in, etc down to 5ft 1in etc OR: You were worried about workplace transport and took readings of vehicles as they passed the 20mph sign on the site road..... The lowest figure might be 18mph The average might be 24mph The highest 32mph And if you plotted all the readings they would form a bell shaped curve Phil

Phil I wouldn't question your deftness with matters of conditions of insurance but the example you offer is of the normal (Gaussian) distribution which can be just one instance of the Poisson; refer to any editions of the 'Practical Guides to Business Statistics' by Siegel or by Berenson, Levine and Krebbiel who explain how the Poisson is a distribution of discrete variables unlike the normal distribution which is one of continuous variables. This distinction is actually very necessary: unless the right category of variable is used for statistical calculations, any inferences or conclusions become seriously erroneous. Canopener What does a distribution of 'poulets' (chickens) normally have in common with a distribution of 'poissons' (fish), other than their appearance on the IOSH Discussion show?

Argghhhhhhhhhhhh. Now I know why I was getting funny looks in that French restaurant the other day! I must brush up on my French vocab!

KeiranD. Many thanks - I stand corrected. That'll teach me to work beyond my area of expertise...!!! Phil