Well, I knew that those more or less semi-automatic predictions only once in a while hit, the only thing about them, is that sometimes they hit straight as they did on the first post, but when they do hit, they might more often hit box, so the predictions are Straight-Any, as most of the numbers that it predicts are singles, it might not be good or very good for doubles, the reason that I am making these tests of it, is because:
The system is more or less semi-automatic so I myself don't have to make the prediction.
It sometimes hits straight.
And the numbers, as I am using it, I am only using about 20 or so combinations for these tests per each draw, I can use fewer or very many more, but I think that 20 more or less are about right, later if that does not work OK, I will then see how it does with very many more combos.
I want to see how it does in the more or less long run, there is a good chance that it might fail the tests.
-----------
For Tonight Eve and for Tomorrow Midday:
493 143 362
149 319 027
134 394 907
----
124 293 953
643 951 431
146 513 441
-------
3 Extra combos:
326
965
795
--------
It is very late predicting the 124 boxed, but I will leave it there anyway.
I didn't develop this prediction system myself, I don't really know how it predicts, I am just testing it as best as I can.
As I am using this right now, each set of 3 numbers is a separate prediction from the others, but is flexible so it can be made to give more or fewer combos for each prediction, on this post then there are 7 separate predictions.
I chose 3 combos for each prediction as on the few private tests 1 and 2 combos for each prediction were not enough, if in the long run that does not work, then I might see if more combos for each prediction might work better.
Also I can use fewer or more predictions.
As you see, some combos repeat on the predictions, that might not mean much of anything, I don't know.