People have been asking so here it goes,
1. Take a sample any sample, preferably 100 of the past draws to determine if the next will be singles or doubles. I usually take 10, 100, 200, and maybe 300.
2. Determine how many doubles are in that sample. There are 270 doubles from 000 - 1000. In theory doubles will occur in a sample 26.97% of the time. I don't remember the exact percentage, but divide 270 by the 1001 possible numbers. It has to be exact.
3. Now, for a 100 number sample the theoritical mean or average for doubles would be 26.97 or whatever. if it's 200 double that number.
4. Next, find the actual number of times doubles have happened. Now for the theorem.
x = the actual number of doubles in the sample.
x = the theortical mean
N = the size of the sample
5. The name of the theorem is "The infferential standard deviation of a sample"
a. First you subtract x from x and square the result
b. Then you divide that number by N - 1
c. Then take the square root of that number
6. That number gives you the standard deviation. It will be close to X - X very close but don't let that fool you. You also have to multiply by 100 to get it up to the sample size
7. Now for the complicated part. go to the last number in the sample #100 or 200. Say, there haven't been enough doubles. How do you know if singles will hit? If there isn't a double from #100 of your sample to the number that is #100 plus the standard deviation rounded then singles will hit.
Example, If there have been only 24 doubles in the sample is is under it's average by 2.973 samples. Square that number, which gives you 8 .83. Then divide that number by N-1 or 100 - 1 (99). That number is a percentage multiply by 100 to get it correct for a sample of 100. Then take the square root which will give you 2.98. If there isn't a double with in the range of #100-#97. Including #100 and #97. Then there won't be doubles. If it is within that range then doubles can happen, but not always. If #100 is a double and if leaving the range would cause the number of doubles to go down on the next range, then the next number will be a double.
8. If the number of doubles is 27. Take a smaller sample or a larger sample. Do this formula on all of them and you can deduce if doubles will hit or not.
9. If doubles have hit too many times just backward engineer the proccess.
10. Find the difference in the # of doubles from the theortical mean and do the theorem again.
