Your comment reminds me of trading arguments about t/a. What if the point of tracking numbers if the results of random. is someone going to say the balls are not selected at random? Do the balls exhibit flocking behavior. Should we put a bollinger band around the results and look for a breakout or a reversion to the mean?
Technical analysis works forever and works like crazy, based on law of large numbers distribution. If a certain price pattern works profitably 65 out of 100 times, you can take all 100 similar trades and arrive at predictable results when average losses are smaller than average wins. Very simple, universal law in effect. But Friday's drawing is a one-time dice roll: there are no 100 or 1,000 draws using the same number sequence (chart pattern) repeatedly every time for that same jackpot prize (trade profit). Just like any lone trade on its own is a random event, so is tomorrow's Mega drawing. Why flailing traders never come to grips with the reality of T/A: it ain't about every time or any one time. It is about all the many times blended together into one cumulative sum result
This is a better way to play. Basically the same return for less effort. http://cockeyed.com/citizen/poker/lottery_simulator100.php The thing that I find rather ironic is that if one could design a system with the expected return of this lottery, one wouldn't need to play the lottery!
Also remember...the more tickets you buy, the better your chance. $1 ticket = 1 in 176 million chance of winning 10 $1 tickets = 1 in 17.6 million chance of winning 50 tickets and my odds are only 1 in 3.52 million. 1,000,000 tickets = 1 in 176 chance of winning.
I made money with charts for years... but it was not because of the law of large number distributions. Its because some price behavior was indicative of larger players getting involved. and in my case... by looking at charts I could tell which nyse specialist ran the show the way I liked the show to be run. if you are buying into a true definition of random... there is no way to predict or beat or win in the long run. If you make a profit it was luck and you better quit.
Each of the 6 position numbers is a time series which oscillates within its own historical 90% range and a mean. Question is what method can you apply to project an even smaller target range in the next period...spectral analysis, neural network, perhaps markov chain(independent draw)? Any suggestions how to open a 6 combination lock?
This from ET's resident Right Wing science guy. Please tell us again how evolution and climate change are fairy tales. It never gets old.
Interesting... Ok... now this is where I might learn something. 1. I am dubious that past history of the 5th position... has any value in predicting the next ball coming if the next ball is truly a random event. Are we really looking for non random things... such as some balls being lighter or too large?
The actual reason for bias is not as important as 'inference' drawn from the residuals/errors of predictions from your own analysis, whatever that might be. Only when you discover the condition where your discrepencies are smallest from the natural frequencies of all 6 oscillations, you will resonate to get the jackpot. Question is what are the conditions?