GAT looks good

[relowe]

Hi Anastasios,

Been reading about GAT over at your download site. It does indeed look interesting and hopefully quite capable of aiding one to achieve rather much better than standard grade odds when selecting our numbers. Been some time coming this GAT tool but is very welcome of course. Thanks for this Christmas present. I shall look forward to exploring its features and perhaps one day I shall be able to insert something in your testimonials page.

relowe

[lottoarchitect]

Someone had to make a start in this forum relowe GAT is indeed innovative, beating the odds in the range 2x-a few thousand better and I have to say it is a true predictor by the means, when a draw is predicted during the generation of the statistics, that draw (and all following ones) are completely unknown to the system. This real back-testing performance gives confidence on what a user can expect and get from this engine. The calibration process of the engine tries to get an understanding of the randomness of the lotto being processed (obviously without using any data from the draws to be tested - they should and are unknown to the engine at that point) so to be ready for the first prediction of the tested draws.
You may also notice that GAT can give a 6 numbers prediction, even when picking only 6 numbers. This is true prediction and is amazing of course, however since it is not a regular occurrence, cannot be trusted by the means we can't be sure if it will repeat again in the next tested draws. Thus my recommendation of observing and utilizing regular hits. In the future, I hope to improve further the prediction accuracy if possible, so to have even such hits occurring more frequently. If I ever manage that, so a GAT can give the correct combination more than once in e.g. 100 tested draws, then just playing a single ticket we possibly win once the big prize in the next 100 draws just by playing the 6 numbers prediction of that GAT table! But if this is not actually possible, an even more focused engine will give more regular hits than now.
Anyway, I really hope users will find GAT very useful and performing favorable to their lotto game. In the various lotto games around the world I have tested, the results I get are promising and quite better than pure luck, so I'm confident it will do good for any lotto game too.
GAT''s release is a dream come true for me. It was a project formulated back in 1999. We are now 12 years ahead of that first engine!

cheers
lottoarchitect

[foreverbob]

Going for a six number prediction would be the best of the best. Indeed, it happens rarely, still it does.
I have verified this for the Euromillions tracking the 5 number prediction throughout some tables.
I was actualy looking for 2 occurances of a 5 number prediction, to more or less see if they repeat and how long it takes for them to repeat....unfortunately, none yet has repeated.
I took 500 total real draws to be tested, assuming less would be of little use.
The Category of 5 numbers to be predicted produced GATS that hit "only" once.
I will continue to let it run, but if a GAT is able to score a 5 number hit once in evey 500 draws so far, then it might repeat again. Suppose it did get a direct hit early in the beginning, why not play that GAT for a long period of time? Using the % would be of some guidance.
Just assuming....

[lottoarchitect]

Somewhere here I have an old post discussing theoretically the drawback of picking the bare minimum of numbers from the highest hit category, which produces just one maximum hit over the tested period. Although it is actually a real hit it turns out, at least in theory, the expected waiting time will be a few thousands of draws, even if GAT can e.g. deliver 1000x better odds. The additional drawback is that whilst waiting for that outstanding hit to occur, most hits produced are in the range of 1-2 hits, thus we really don't win anything in the meantime. The advice is always to look for regularity of hits when deciding on a GAT table for use.

[foreverbob]

Thanks for letting me know this, LA.
I guess playing an extra single ticket along to wait for a 5 hit thats been due, just takes too much time and would not be worth the effort of playing it.
There is however a thin line between the regularity of hits + their amount of predicted numbers and the total pool of numbers to wheel within one's budget;
I am still fiddling around with most settings to get the best balance between all these.

Bob