Really interesting, and I'm not surprised. I've worked in the simulation industry for more than 25 years and became very interested in the Florida lottery several years ago, specifically their fantasy 5 game, a 5/36. A quick review of their games made me realize at the time that the 1-in-376,992 odds of winning were the best of all their offerings. A win would not make you rich but would provide a nice payout. So I set up a simulation and analysis system to help me decide if such would be a useful playing aid. Florida had 20 years of winning numbers accessible online plus posted daily winning numbers. I wrote some software that did daily analyses of the numbers for all the trends that I could imagine might be useful - histograms of number frequencies, standard deviations, pairing frequencies. These were placed in daily updated databases. I then wrote what I called "pickers." I had a lot of fun with this part but basically these were number selection algorithms that used the statistics collected to pick winning numbers. I went through every lottery expert book that I could access Plus came up with some ideas of my own. The basic daily sequence of operation concluded by going back to the beginning of the 20-year period and letting my collection of algorithms play the lottery each day and I would plot their success. I concluded that randomness is a bitch. No picking algorithm attained a higher degree of success than did my pure random pickers. The very difficult thing to believe and understand is the line that demarcates past and present. You can look backwards and see trends and patterns emerge, plainly. It is very hard to accept that they break completely when they cross the line of present into future, that they have zero bearing on what will happen. I knew this to be true but had to show it to myself in pictures.