Reply to thread

Message: [QUOTE="azbo, post: 909111, member: 44090"] this thread is really long, in the beginning there were people asking about data sample size but I just didn't have the the motivation to read 11 pages of posts. The thing with sample size is increased sample size just means a higher degree of confidence. The sample mean is rarely the same as the population mean but you can usually get to a point where you have enough data to suggest a certain confidence interval. It works out to be that a 95% confidence interval is a common target in statistical analysis. You also have a range of error that comes with any analysis. This can also be calculated. A lot of times you'll see a political poll with +/- x% attached to it, which means the actual value can be expected to be within that margin of error from the result 95% of the time. The required sample size to achieve 95% confidence interval is dependent on the total population of the data. It turns out that 95% confidence is equivalent to 3 std deviations from the mean under a bell curve. In laymans terms that means if you see a normal bell curve with a tail drifting off to each end .025 of the data points will be greater than 3 std deviations away from the mean at each end which adds up to .05 which is where the 95% or .95 comes from. I haven't done much statistics in a while but you need to know the expected outcome of an event % wise, the total number of events, your error tolerance (it takes a lot more data to achieve a 1% margin of error than a 5% for instance)and your confidence interval. We're not going top have the total population number because the poker sites aren't going to give you a db including all the hands. You don't need that number to do this it just makes it easier and more accurate. [/QUOTE]

Verification

Search

Reply to thread

About CardsChat

Our Community

Our Site

Follow us