The problem with lotteries is that there are too many numbers ie 49 in the UK lottery and 34 +14 in the Thunderball draw. To have the faintest chance of winning the big one(bugger the little prizes) you have to play with a greatly reduced numbers. Now the problem is what numbers to pick or leave out. I use the UK thunderball 1-34 section as a test bed. I aim to pick 15 of the winning numbers 1 out of every 4 draws. I structure the lottery in decades. A decade looks like this:
1 2 3 4 5
6 7 8 9
If a ball hits 1 to 5, I record it as 1.
If it hits from 6 to 9 I record it as 2.
If it hits in the first and second sections I record it as 3.
The format remains the same for the other decades.
A typical draw could have a pattern like 1 3 2. Most of the draws miss out one decade. So we ignore the decade 30 -34. If you get the right pattern the 15 numbers should be represented in the winning 5 numbers of the draw. Now the 3 pattern is unfriendly because it stretches across the 2 parts of the decade. Ideally you want a pattern without the 3. There are 8 possible patterns without the 3. What you need to do is record these patterns:
eg 2 3 1
3 1 2
1 1 1
etc
Say about 100. See which ones occur more frequently and try and evaluate what the next one will be. You can do the same with the 4/49 game.In some ways it is easier because you do not have so many 3 patterns. Remember in the 4/49 game one deade is missing around 70% of the time.
I use Saliu's incomparable Lotwon632 to filter out the final numbers. Get the pattern right, get the filter settings right my sons and you become a millionaire(I am not one yet but am totally confident of becoming one. Nik Barker if you are reading this how about a trip to one of London's pubs. Incidently using Saliu's software I break even on my lottery attempts in the UK and South Africa. I have tried wheeling with limited success(it's the best way if you lack Lotwon) but let's face it your chances of being struck by lightning are better or being kidnapped by aliens and taken to Alpha Centauri.
------------

Message 2 in thread
By Ion Saliu

Andrew:

Of course, I know you. But this post of yours in r.g.l. is a real surprise to me. It also coincides with surprising emails from the more reclusive user of my software, known as DEZ. There are a few people who suspect DEZ is “Deep Throat” in Watergate. DEZ is really big in the decade elimination theory. He mostly follows color-coded patterns. Here is, briefly, how I see it (not color-coded).

The decades can be used as a filter, no doubt. It's easy (in LotWon or MDIEditor and Lotto WE) to eliminate decades. If you want to eliminate the decade 20-29, simply write:
20 21 22 23 24 25
26, 27 28 29 29 29
at the very top of the LotWon data file. It would be the ‘Eliminate all numbers from the top 2 draws’ filter.
In MDIEditor and Lotto WE, write 20 21 22 23 24 25 26, 27 28 29 in the first text box of the input form (‘Eliminate numbers’).
It is hard, however to do the winning reports; that's the real problem.

You can see from the following template why it’s so hard to create winning reports with decades as filters. The number of decades varies wildly, from 30+ to 60+ (I heard it’s 90 in Italia).

You can use the design below for your games. Use QEdit, or MDIEditor, or WinWord to create and edit it. If a decade hits, write +; if a decade misses, write - in the respective column. I went a step further and wrote ++ if a decade has two hits in the draw. Then you follow the same principle as for the regular filters; that is, look for two or three consecutive + or -. In the example below, chances are good that the 0-9 decade will miss next drawing. The chance is even higher for 20-29 to miss.

Next, correlate the ‘winning report’ above with your regular WS files or statistical reports.