| What you are trying to do here, will be impossible.
If you think at lottery numbers for the statistics which is involved, then yes you can say theres a higher chance for one number to be next in line, since statisticaly it hasn't been chose as much as the one that's been chosen previusly.
The real calculation method involved here is completely missed, once the new drawing is taking place the following day, the previus selected numbers are once again placed within the bowl, which once again rewards them with the same significance statisticaly speaking as all the other numbers, thereby effectively eliminating any advantage any of the other numbers would have gained by the fact that a few numbers were removed.
The way this statistic is involved is by the pure luck factor, if you flip a coin 1.000 times 50% would be heads and 50% would be tail, the same form of elimination, as described above, has it's influence here, yet we know for a fact if you keep on trying you'll end up with a 50/50 chance of getting what you hope for, because your starting point only has two result values to present you with.
In lottery it's a different ball game, in your example you have 1.000 numbers to chose from, so in order to have a number of drawings which statisticaly would favor one from the other is so minimal, even with 1.000.000 drawings to calculate your expected next number, would improve the statistics for precisely that number so insignificant that it wouldn't matter, so your best bet would be to go for the hit-and-miss approach, by simply selecting a random number for each time, hoping the random generator for the lottery will be influenced the same way your random generator is.
(think along the lines of "even in caos theres structure" here) | 
Linear Mode
