Where to start on GAT Graphs

[20mm]

Hi LA
I'm still in training mode with Both GAT Engine and WG. and have read several topics to help me better understand some areas where I was a bit unsure. My question here is, where on the GAT Hits Chart do I read to predict the next outcome. Looking at the example on page 27 of the Version 2.3 Manual, the table shows 16 occurrence of 3hits, 23 occurrence of 2hits etc. over the 100 tested draws. Do I also interpret this as future occurrences, to read for example, that in the 3rd and 4th future draws, 3hts can be expected or do I start at the right hand end of the chart where 2 hits can be expected.
Form my own runs I've exported GAT's to excel to try and find a correlation with hits across the chart, but could find none. How is this table useful for selecting GAT's or for performing any analysis.

[lottoarchitect]

Hi 20mm, the graph illustrates the hits production performance over the last X tested draws of the real history, in this case we have 100 tested draws. What we look for in a given graph, is repeatability, or as we say it around here "regularity of hits". This particular graph has quite good regularity for 3 hits production, as it demonstrates ability to come up with a 3-hit every so often. Now, given its hits production performance with the "always-changing" dynamics that are applicable over the tested draws, and given that the same exact methodology continues in the future unknown draws too, we can have a reasonable expectation that the particular GAT table will continue to provide a similar hits performance. We can't say that the 3-hit will come at exactly i.e. 3 draws ahead but given this graph, we can say it is about to produce that 3-hit again. The normal usage of GAT is to decide on a GAT ID based on its overall performance on the desired hit category (the whole tested draws range) - judged mostly by the regularity of hits production - and keep using that particular GAT ID in the future draws, using the RF (Run Factor) functionality.
Overall, the panorama under each column displays the most well performing GAT IDs, judged and sorted primarily by the most hits production performance for the desired hit category. Each of those GATs can actually deliver the very good hit at the very next draw. It is really impossible however to say, this particular GAT will do that at the very next draw. This is why we try to establish which GAT has the more apparent smooth hits behavior (the regularity of the desired hits) and stick with it in the next few future draws (always playing the new proposed numbers it carries for every future draw).

[20mm]

LA thanks for that clarification.
Given I decide on using a GAT and wants to view the expected /calculated delay on hits, I know I now have to view the Hits Delay Chart for this. As in the case of the chart on page-28, is it, for example, 1 draw to give 39 occurence of 1hits. Even having said that it doesn't sound right in my head. Please help me understand as graphs can be bit challenging.

[lottoarchitect]

You refer to the height of the red line with the number 1 below? The delay graph shows the delay in draws we have to produce a given hit and upwards. The red bar is for 1+ hit production, thus it shows the delay in draws we had to produce again an 1/2/3/4/5 etc hit.
The horizontal axis is the delay in draws. The vertical axis is the total occurrences of a given delay for the defined hit category.
Each bar is placed above an indicated delay, if that particular delay has occurred of course. For example, in that delay graph, the red bar represents 1+ hits production. You see the red bar on top of number 1 delay and a height of about 39. That means, over the tested draws period, we had 39 times the case "a given draw produced a 1+ hit, given its previous draw (delay = 1 here) was again an 1+ hit". The best delay graphs are those that concentrate the bars of a given hit category over a small delay range and we aim to be at the beginning of this range as indicated by the current delay.