Hi stephen5984
Welcome to the bulletin board.
I can partially answer your question about the basis of the 20 number selection. The numbers are picked on many factors with weighted preference. A major factor is what you have mentioned ... hot and cold/due numbers. Much more preference is given to hot as due numbers can remain due for a long time .. it is best to bet a due number just after it starts to "come alive" again. This due aspect does not apply to #31. When #31 is due always play it and again even after it hits .. (and for those asking, yes as I've said in GillesD's topic, I'll still be playing 31 on Sat even though it didn't hit last Wed.) Other factors that are taken into account are the number's skip hit performance and to a minor extent 10's grouping placements. There are several other factors as well. The overall blend is a secret
It is important to note that your basis of a determination for a "hot" or a "cold" number is "over the history of 6/49". The range that I use for that determination is extremely smaller as I'm seeking only short term trends for the Top 20.
Let us understand what we are really looking at here, which is determined by the total possible number of 649 combinations .. just under 14 million. So far the entire history of Lotto 649 is just reaching 1700 draws (and it is mature lottery compared to many others). If we took your basis of the entire draw to determine hot or cold or any other factor we would also be looking at a "short term trend" when compared with the total potential of the lottery. To truly see the behaviour of the Canadian 649 we would have to live for trillions of years or there would have to be 14 milllion draws a day for several years
And what we would fine would be what all the statisticians say .... all numbers and combinations would eventuall be equal. Yes even the combination of 1,2,3,4,5,6 would have shown up several times
But we won't live to be a trillion and there won't be 14 million draws a day (except for keno draws in Las Vegas) so an extremely short term analysis of an overall short term (<1700 draws) can offer a slight advantage with number predictions. And with the odds we are up against even a slight advantage is really good.
Regarding the stats of "(if you include the bonus) = 2.87" . Yes of course your math is correct but we are talking apples and oranges here in regards to the Top 20. The number predictions do not factor for the bounus number and are only targeted at regular numbers. There has been lots of debate in the lotto industry of the question should the bonus number be kept or ignored? I ignore it. What good is it? unless of course the "all important" 5 regular numbers are captured in the number pool with the bonus. When the Top 20 contains the bonus number just think of it as "bonus"
I also ignore factors like "which ball sets" are used and in what machines and "was the draw on a Wednesday or a Saturday?" Re: "So what's so impressive about picking 4 or 5 out of 20" ... even if we went with the inclusion of the bonus number in the pool ... I'd be really impressed with consistent predictions of 4 or 5 (there would always need to be 4 or 5 to guarantee cashing a ticket ... 2 regulars plus a bonus is just a tease)as this would be well above the "statistically" expected norm of 2.87.
Again the Top 20 go for short term trends to beat the "normaly" expected occurrance of 2.4 "Regular" numbers. The most important thing is not that the Top 20 always get 2.4 numbers for each draw prediction. Even if they always captured 3 for a winning ticket potential that would be no fun. It would be much better for the Top 20 to produce 0 winners for a few draws and then produce sets that contain 5 + for the next few draws. That is what the Top 20 is targeted at ... short term trends which will produce a few pools per year that offer site visitors a realistic, significant chance at a top tier prize win and to have some playing fun! as opposed to just the routine purchase of a quick pick
For those who wish to look at look at the prediction histories you can see that the Top 20 produces "clumps" of top prise potential pools four time a year.
1997 - Jan, April, June, Oct. 1998 - Jan, Feb, April, Sept. 1999 - May, June, Sept, Nov.
It is "due" for another now that we are in year 2000!
This has been somewhat long winded but the issues that you have raised are important ones.