Maybe I should rephrase this, the data sample is only from folks reading this website. You have 2500 sample of a specific subset of much larger subset. In medical testing you have a control group and testing group from as a diversified sample as you can manage. I don't think this particular subset is reflective to entirety of the wow populace. And as Pyromelter mentions the standard deviation would be very useful in interpreting this data.

If you have 2500 people who average say 20k gold but in there only 10 people have 5 million gold you suddenly have a huge statistical skew. The standard deviation lets you know that those 10 people are way outside the bell curve. You can then eliminate say those more than 2 deviations out and get a really good idea what most players have gold wise.

Presenting the data differently with more information might show that the largest cross section is "middle classed" or at least what most folks have for gold reserves. Then you look at it go hmm I only have 10k gold, but so does almost everyone else as opposed to holy crap these guys have millions of gold what the hell I only have 10k!!!

Reply