Does GAT in any way produce random results?

Absolutely not! The process involved is very complex (see the concept) and tries to harvest the even way higher complexity of randomness in draws but there isn't any single element of random behaviour involved in the process. You'll always get the same results for the same parameters in GAT. Besides, if a prediction to the same data with the same parameters change, it is a random prediction therefore useless; better to opt for quick-picks!


What sort of lotteries does GAT predict for?

Any lotto style game supported, including bonus balls or Powerballs. Pick 3-4 and KENO style games are not supported however. A specific version for KENO may be possible in the future. The universal logic of the prediction allows this possibility.


Will GAT require me to play with most of the ball numbers and so cause me to buy so many tickets I can barely afford to?

One great aspect of GAT is it can provide any amount of numbers, backed up with their relevant statistics, you may care to request. It is up to the user to decide on his strategy and what results he wants from the engine, he can get exactly what he is looking for. Besides, GAT has an endeavor to be successful even when very few numbers are requested.


You say GAT can predict for a Powerball number. How does it do that?

The prediction logic is universal. It is not uncommon to have a boost of about 2-4 times better than luck when predicting a Powerball. For example, a 1/20 field has a chance of 5%, GAT can make that 10-20%. That's way better than just picking randomly a Powerball to use! With the introduction of GAT, Powerballs should only be played using GAT from now on!


How many past draws of the lottery should I have in my database?

Contrary to common belief, if we attempt to analyze a lot of history draws to make predictions, we just average the overall outcome, which barely leads to a good prediction. This is just another example why randomness is so hard to tackle. GAT needs only a very small amount of past draws to make effectively a prediction, i.e. 3-20 history draws is usually enough. However, in order to determine which GAT table is more optimal for use, we have to let the engine make comparison tests against real history draws, so to provide statistics for us to decide. So, if we request statistics for the latest 100 history draws and have a set of 20 draws analyzed every time for each prediction test, we need at least 100 + 2 * 20 history draws available.


Do I need a degree in mathematics to understand and get the best results out of GAT?

If that was the case, then GAT wouldn't be user friendly at all! Well, there are two elements involved: deciding on a GAT table to use based on its overall performance via the statistics generated (the system displays the best tables detected) and then inspecting its performance graphically to decide if this overall performance will continue. There is always a chance that performance achieved e.g. in the last 100 draws may not keep up in the future but regularity of hits is one important element to observe. For example, it is possible to have one 6-hit win, when picking 6 numbers only, in a test of 100 draws but this occurrence is rather infrequent so to have confidence it will happen again in the next 100 draws; it may do, it may not.


How is it that one GAT prediction can be used for several succeeding draws?

GAT exploits the possibility a set of consecutive draws to have common characteristics, called a "signature" as explained in the concept. This signature remains more or less valid for a few draws ahead. This is identifiable in the graph produced for each GAT table. When you observe a sequence of good successive hits, that GAT table detected a good signature to use. Therefore, our task is to identify which table is more capable in identifying good signatures to use.


Can your Wheel Generator be used with GAT number predictions?

In fact, the best suited tool for proper and full utilization of the numbers obtained by GAT, is to use them in Wheel Generator! This is why we also offer a bundle of GAT & WG. These programs complement each other to make a complete suite for regular lotto play. As my favorite quote say "What to do with these numbers from GAT? Use them in Wheel Generator of course!".


Is it complex to set up the various operating parameters?

You wouldn't have to even inspect the parameters to run a prediction. However, we recommend running the engine with as many GAT tables as possible. One important parameter needs attention however to get correct results when we continue to use a prediction of a GAT table for subsequent future draws. Overall, the parameters are self explanatory but you should get familiar with the operations affected by these parameters.


I note you have had several releases of Wheel Generator. Is this version 'it' for GAT? Or do you have more still to come?

As in any application, new features are added or improved over the time. GAT is no different to that and there are already scheduled a few more important additions. More in due time.


Do you guarantee I will win a prize?

You should be aware that in any forecasting model there is no guarantee of performance. We try to do our best in making a prediction, so to be way better if possible than pure luck and GAT is on the good side of this, however as in all aspects of life which involve uncertainty, performance of the past doesn't necessary guarantee performance of the future. Besides, even if a prediction is accurate enough, a prize solely depends on how the user utilize that result in his playing strategy. So, guarantees cannot be given.

If you want a better and deeper understanding of the various prediction methodologies available today and how GAT Engine deals with the prediction subject and what makes it novel and unique, a detailed review called "The Art of Lottery Numbers Prediction", written by the author of the GAT software Anastasios Tampakis, is available as a PDF file. The reader will appreciate the difficulty in making a lottery prediction and why G.A.T. Engine clearly stands out from the other numerous approaches available today.