A question for statisticians

**Kenyth** · 10-06-2009, 01:41 PM

Originally Posted by Shelby07

That is the reason that a coin toss can't be used as an example here

As I mentioned before, I'm not sure I agree with this. It appears that "5 chances out of 1000" is not the same as "1 chance in 200" which is how the contest was originally presented.

Odds presented as this are a simple proportional expression, which means that 5 chances in 1000 is equal to 1 in 200 just as it is equal to 10 in 2000. This is a simple method of trying to predict chances of an outcome with known information at a given point in time. The instant one of the chances is used, more information becomes available, and everything changes.

The confusion here is occurring because folks are trying to break it down into it's singular events and then average it back together. It doesn't work that way. In order to even be eligible for the second, third, fourth, or fifth ball you would already have lost the previous draws making those draws irrelavent to your current odds. More information is available and the odds keep changing for the worse since the numbers aren't being changed proportionally on both sides of the equation.

The fact that you had a previous chance or chances still factors in when calculating the odds for before the contest started. The odds for each individual chance don't simply average out to the original odds, because each subsequent chance assumes you lost the preceding one(s).