Many years ago I wanted to learn how to predict lottery winning numbers and saw that it was a lot of trouble for me to keep long lists of past winning numbers, also that I had to try to do it with the pick 3 game as it had fewer possible combinations than those of other lottery games.

Also, that I should try boxed and singles as there were fewer combinations.

But even like that, I saw that if I wanted a chance to predict right more often that the prediction numbers had to be very many and not just a few, but as the payouts are not very high for boxed singles that I should not try to buy the predicted numbers and that I should do it just for the fun of it.

I needed to make some sort of prediction method that had some kind of logic to it, some kind of statistical logic, I needed some kind of more or less logic rules to go by.

I saw some posts here at the lottery post about filter patterns or about something like that and also about one or more lottery softwares.

Using search engines I went looking for lottery softwares and found a few, I downloaded, looked at them and gave them a try.

They used filters and their patterns, I started to learn about filtration as a prediction technique.

I also found the Sums and Roots chart for the boxed pick 3 game at Bob's P internet site.

I began to study the past winning numbers of the Texas pick 3 game, I had to learn by myself how to use the numbers and patterns that I saw there in order to make predictions by filtration.

Sometime maybe very much later I came to call it the "Negative Prediction Technique".

First I would study about 40 to 20 past draws more or less so I could kind of learn the general way in which the numbers and their patterns moved about, this had to be done only once in a while for each of the pick3 games that I wanted to make predictions for.

There-after I only needed to look a the last 4 to 10 past draws and sometimes just at the last 1 to 4 past draws, in order to make predictions for the very next winning draw.

But for boxed singles more often than not the predicted numbers were from about 20 up to about 45 and maybe most of the time were about 45 of them, I don't remember that well about that.

And even with from 20 to 45 predicted numbers the predictions were not very accurate, but I was trying to do that which they said that was not possible, to predict that which is not predictable.

I also started to make many posts about pick 3 filters and their patterns.

I put and end to all of that little by little.