random vs "real world" actual numbers

I am having trouble interpreting the results of Chebyshev's theorem as applied to random and actual data. In Excel I have two columns of data. The first column are random numbers and the second column are actual numbers from a "real world" set of data. Each column is 63400 rows in length. I apply the following Excel equations to the random number data in column A:

=(AVERAGE(A63098:A63297)-($B$63307*STDEV(A63098:A63297)))

=IF(A63297<B63297,IF(A63298>A63297,"yes","no"),"")

The absolute cell $B$63307 in the first equation is a variable for the number of standard deviations in Chebyshev's theorem.

The cell B63297 in the second equation is the result of the first equation. What I am trying to do is predict that the next random number (A63298) will be greater than the previous random number (A63297). I have set the random number generator at a low of 1.0000 and a high of 1.0001

Using the above limitations I can successfully predict the next random number will be greater than the previous random number 99.2% of the time with a standard deviation variable of 1.8

Here's the problem: When I apply the same equations to the "real world" data set it only predicts the next number to be greater than the previous number 48.94% of the time. Why is that? What is so different about the "real world" data vs. the random number data?

jmj said:

Using the above limitations I can successfully predict the next random number will be greater than the previous random number 99.2% of the time with a standard deviation variable of 1.8

Click to expand...

Ummm... If you can predict it, it isn't random. Your success rate for RANDOM data should be very close to 50%. My first effort would be to take a good hard look at those that are not so random as thought.

Further: 99.2% vs. 1.8? Is that 1.8% or 180%?