G.A.T. Options

[iWin]

Hi LA,


Thank you for all of your previous responses.


If I want G.A.T. to analyze the entire history of a game each time I run a new scan, how should I set the settings in the options?

For example, if my game has a history of 1000 draws, and with every scan I want my statistical data to begin with the very first draw (draw number 1) and include all the rest of the available data (up to draw number 1000) to predict for the next game (draw number 1001), then how should I set the settings?

According to you;

Total History Required = Tested Draws + (2 x Statistical Data) + Run Factor Overhead

So in this case, if I want to test 100 draws, should I go like this:

Total History: 1000
Tested Draws: 100
Data: 450

and then let the run factor overhead just go up for every upcoming draw, or should I manually reset the run factor to 0 and add 1 to data;

"Next Game" :

Total History: 1001
Tested Draws: 100
Data: 450
Run Factor Overhead: 1

or:

Total History: 1001
Tested Draws: 100
Data: 451
Run Factor Overhead: 0


or maybe neither of these and it has to be done in a different way?


I hope you understand the question. The point is to analyze the entire history for each scan and use the entire history as statistical data.


Thank you in advance.

[lottoarchitect]

Hi iWin, you should use the first approach, only let RF go up in the case you want to keep using GATs that were picked with RF 0. If you want to perform completely new scans (RF=0) and want all the history to be used (as much as stat.data allows according to the equation), then you'll always have RF=0 and adjust stat.data accordingly.

if you do this

Total History: 1001
Tested Draws: 100
Data: 451
Run Factor Overhead: 0

you break the continuity of GATs since you change the stat.data.

[iWin]

So if I just calculate the number of stat data that would take G.A.T. to the very first game and then just let RF go up, would it then start analyzing from that very draw with each new scan? Because my understanding is that if I do that, then G.A.T. would ignore draw #1 and start with draw #2 to analyze the data for the new scan?

In this example (having 1000 draws in the history), if using stat data of 450 makes G.A.T. to start with draw #1, then having the same stat data (450) and testing 100 draws should then make G.A.T. to start with draw #2 to analyze its data to predict draw #1002, isn't that right?

Could you please write an equation as an example, if you have a total history of 1100 draws, and you want to use all of the 1100 draws as stat data to be analyzed, and then the next game you want all the 1101 draws to be analyzed, and then the next game all the 1002 and so on, then how would you approach that?

-------------------------------------------------

A second question; if a G.A.T. has always had a 50% success rate in the past, does that mean that it will continue to do so in the future, or is it possible that such a successful G.A.T. would completely stop predicting correctly all together at some point in the future?

[lottoarchitect]

So if I just calculate the number of stat data that would take G.A.T. to the very first game and then just let RF go up, would it then start analyzing from that very draw with each new scan? Because my understanding is that if I do that, then G.A.T. would ignore draw #1 and start with draw #2 to analyze the data for the new scan?

Run Factor (RF) takes care of this. If your initial run with RF=0 resulted in GAT using the very first history draw, after adding a new history draw and having RF=1 to accommodate for that added draw, will again result in GAT using the very first history draw in its analysis. After all, if you want to keep using GATs with RF>0 (so not to break the continuity in future draws) the engine must always start its analysis at the same draw where we had RF=0. The whole point of RF is to ensure the engine starts from the same initial history draw when new draws are added to the history.

Could you please write an equation as an example, if you have a total history of 1100 draws, and you want to use all of the 1100 draws as stat data to be analyzed, and then the next game you want all the 1101 draws to be analyzed, and then the next game all the 1002 and so on, then how would you approach that?

[Start draw in history]=[total available history] - [RF] - 2 * [stat.data] - [tested draws]

All that needs to change is RF. Since RF is controlled automatically by the start point at the options, all you have to do is keep all the parameters (tested draws, stat.data, start point) the same as in the very first run with RF=0.

A second question; if a G.A.T. has always had a 50% success rate in the past, does that mean that it will continue to do so in the future, or is it possible that such a successful G.A.T. would completely stop predicting correctly all together at some point in the future?

There isn't any contract between a GAT and a lottery that results in a given hit percentage to always come true forever and ever in the future. It is actually very possible for that GAT to stop delivering or deliver even more. What that percentage tell us is the overall ability of that GAT table to produce hits over the always changing dynamics of the lottery game. Treat any GAT as a short-living entity to be useful only for the next few draws (the 100/X rule). If it fails to deliver, it doesn't owe to come back however for various reasons so insisting on that GAT is probably bad practice. Whatever the engine does to predict the test draws, the same methodology continues unaltered in the future unknown draws too. So, whatever observed over the test period, something similar should come up too however this can't be guaranteed; the nature of the experiment forbids guarantees and a definite answer to this question.

[iWin]

Thank you!