|Posted: April 10, 2006, 3:12 am - IP Logged|
I still don't understand what information your 5X56 chart with 68211 in every position is trying to show. What law of physic did you use to define the perfect world you described? I would think in any world perfect or not, if five balls were randomly drawn out a pool of 56, the results would be just as unpredictable each time. In the last 56 MegaMillions drawings 74% of the numbers were from a pool of 33 which might suggest a bias that can be exploited.
I understood it perfectly. It says that knowing absolutely everything there is to know about previous drawings doesn't change the odds or offer any advantage. His perfect world comes from the laws of math, but he's wrong about how it works. The laws of probability, whether in a perfect world or an imperfect one, don't say that each of 56 numbers will be drawn an equal number of times in 56 drawings. Probability merely says that if every ball is identical then every ball has an equal chance of being drawn. Probability also says that there's nothing unexpected about some balls being drawn more than expected and some balls being drawn less than expected. In fact, too uniform a distribution would strongly suggest a problem. Graphing the frequencies with which numbers are drawn will always result in a bell curve, where most numbers are drawn a similar number of times, a few are drawn more often and a few are drawn less often. The shape of the curve will remain uniform over time, but the position of each number will shift as it is drawn more or less than expected.
Lotteries have multiple sets of balls and multiple machines, and to keep drawings as random as possible they start by selecting which balls and which machines to use by a random process. That means that the number 7 ball that was drawn three weeks ago may not be the same number 7 ball that was drawn two months ago. As such, if there are four sets of balls, instead of analyzing a set of balls from 1 to 56 you'd really be analyzing a set that goes from 1 to 224 and you wouldn't know which ball is which. Even if balls really were hot or cold, what would the history of the number 89 ball tell you about the 36 ball in the next drawing?
I'm not a complete skeptic. In fact, I'm absolutley positive that the drawings aren't 100% random. I'm also positive that they're close enough that at best a proper analysis might result in a 1% improvement in the odds, but there are still two problems. The first is that there isn't any practical difference in odds of 1:175,000,000 and 1: 174,240,000. The second is that you won't know when they retire set of balls, and your odds shift to 1:177,760,000 until your analysis catches up. As you said, the results would be just as unpredictable each time.