Talk:Quest success rate research/@comment-96.50.176.11-20110914050858

Your math is completely wrong on this btw, good job mr. mathematician.

First off, you're testing too many things at once. As you state in your opening thesis, you are wanting to see the success chances of your adventurers, and yet you close with some information about the rewards? If you want to test success chances, test success chances, halfway through you convert success into $ which are 2 completely different topics. And are you looking for the success chance by class? Or just success in general? Perhaps you're testing to see if success chance is higher if you give them the item you ask for, if you give them an item of lower quality, and if you give them an item of higher quality? Once we decide what the success chances are, then we can start tracking the actual money received. Or we could just take a simple average of reward - market value of item and come up with a rough estimate of how much, in this situation, you actually got.

At any rate, your math is wrong too. You assume that it is a normal distribution, which it is not. If you were to graph your results they would undoubtedly not look like a bell curve in any way. To correct your mistakes so that I can properly find the actual mean, actual variance, and actual std. dev, first I have to write down every number outside of your classes, because as I previously stated you're assuming class makes a difference which is an unknown and if you're worth anything as a mathematician you'd know that in a linear equation you cannot test 2 unknowns at once. After writing that down I then multiply them by the probability of the occurence, in which this is uniform at 1/59, except for 0 which occurs 33/59 times. (protip: to double check your work, add up all the probabilities! They equal 1!). Next we take the x value and multiply by p(x), and sum all of those to find our expected value. Our expected is $40,773. Now here's where its starting to get complicated so follow closely. Once we have the expected, or mean value we can move onto the next step in which we find the variation, we do this by taking each x value, subtracting the mean, squaring it, then multiplying by the probability, which gives us the formula ((x - mean)^2)*p(x). If you do this properly, your variance will come out to 5.4 billion. Yep, that's right. Now last but not least we just have to find the standard deviation, which is just a square root of this astronomically high number. Final answer is: $73,587.97.

Now as I've already said earlier, I'm beginning to seriously doubt your worth as a mathematician. But the fact is that you try testing 2 variables at once in a linear equation, that you assume that an item falls into a normal distribution pattern when it so clearly doesnt, and you don't even know the basic statistical formulas! This isn't advanced stats and calculus here, this is Stats 101! And for you to completely miss the ball on the basic equations, nevermind the other issues, nevermind the ridiculously tiny sample size, is just way too much for me, or anyone else, to ever take you seriously. Go home kid, play times over.