GAT prediction engine has been designed under the principle of being easy to use and straight forward. No cluttered screens, intuitive design and easy access to prediction results. Below is the main application window.

The user opens his favorite lottery game, adjusts, if needed, any parameters and commences to prediction. When the prediction is over, he inspects the results to pick the most favorable prediction found.



The creation and population of a lottery game is straight forward. GAT can support any style of lottery game, including bonus balls and powerballs. Population of your lottery game is supported with the new AI import draws mechanism which is capable to understand the structure of the data on its own! Just download history draws from your commission web site, copy the relevant column data from the excel in a text file and let the AI understand what is what! No need for special formatting or any complicated adjustments.





Options define how the prediction engine will behave. A brief explanation of the options :

- Tested draws : How many draws will be used as test subjects by the engine to produce prediction accuracy statistics for the user.

- Statistical data : How many draws are used to make a prediction for each of the tested draws. GAT usually needs very few draws to make a good prediction.

- GAT tables : How many different ways to evaluate the statistical data to make a prediction. More tables mean more chances to find even better predictions.

- Run factor : In order to keep using a prediction of a particular GAT table in future draws, utilization of this parameter is necessary. More in the help file.



Panorama is the display of the best GAT tables found so far (providing the most hits). Below we see GAT provide information on the 2 GAT tables that returned the most hits on the powerball field (1/20) of the Greek lottery game Joker. We can see that GAT has managed to predict 18 times (18%) the powerball over the last 100 real powerball draws! This is 3,6 times better (or once every 5,5 draws compared to once every 20 draws of natural chance)! Double-clicking on a GAT number, will take us to the details of that GAT table.

We have the ability to request any amount of predicted numbers, all backed up with their statistics.



Follows another example when predicting the main balls of the Greek game Joker (5/45), requesting 10 numbers. The results are the hits achievement of the predictions over the last 100 tested draws. The Imp. ratio shows how many times better we are in performance compared to natural probability. The "absolute position hits" is needed when we want an exact amount of hits produced i.e. exactly two hits. This is necessary when planning to play matrix coverings.



Two more examples of hits produced at the Belgian lotto (6/45 + 1 Bonus ball), when the bonus ball is included at the prediction process.




Follows the display of a particular GAT table. We can inspect its overall hits performance, we can see the prediction for the next draw and a graph that illustrates at a glance how this GAT table performed in prediction over the whole tested draws! The red line represents hits achieved for each tested draw and the blue line represents an average (hits per draw) of the following X tested draws (5 in this example). This can give instantly an overview of the performance.

Chance to hit 4 numbers correct in a 5/45 game (when picking 10 numbers) is around 0.57% or once every 176 draws. GAT has managed to hit 7 times within 100 tested draws! This is 11,25 times better performance than luck! Such hit improvements are common with the GAT engine and the reason we believe this program will become an invaluable partner to your lotto interests.