gat engine working more then a day

[kman]

Hi,
I am new here,
The gat engine is working more the a day and still no results, is this normal?

[lottoarchitect]

Hi kman, what do you mean "no results"? The results are readily available under the tab "predicted numbers" when you display a given GAT table.

[kman]

Hi,
thanks for the quick replay.
I don't get numbers under predicted numbers. I guess because it's a demo?
why do I get only one 6/6 result? is it normal?
Why When I am adding more numbers at 14 numbers the software still find 1 6/6?
another question, hope you might help.
I am looking for a software that can tell me which 14 numbers give the most 6/6 in the 100 past results,
and the 14 numbers that gave the least winning results.
also I want to be able to pick 1- 4 of those numbers and let the software find me the rest. can your software do it?

If your program does those things I'll be happy to purchase it.

[lottoarchitect]

I don't get numbers under predicted numbers. I guess because it's a demo?

When the program runs in demo mode, you see No1, No2 etc as predicted numbers, not the real picked numbers. You can evaluate the system using the compare numbers feature and the graphs/statistics shown.

why do I get only one 6/6 result? is it normal?

For the bare minimum of numbers requested, yes it is normal.

Why When I am adding more numbers at 14 numbers the software still find 1 6/6?

How do you add more numbers? The way to ask from the system to pick 14 numbers is via the +- buttons of top of the panorama. You should get more than just 1 success for a 6 hit with that amount of requested numbers.

I am looking for a software that can tell me which 14 numbers give the most 6/6 in the 100 past results,

That is the default behavior of the panorama display of GAT. Whatever GAT table is shown, it has managed the highest amount of hits for the desired hit category it belongs to (the column it resides in). However, I understand that you look for a static set of 14 numbers. GAT does not produce static predictions, this is sort of pointless approach because it rely on statistical evaluation and statistics do not really work on lotteries; this is old fashioned methodology. GAT proposes dynamic predictions, thus for every future draw you'll get a new set of 14 proposed numbers specifically prepared for that draw to come. If you look for static predictions, you'll not find this in GAT.

and the 14 numbers that gave the least winning results.

Not supported.

also I want to be able to pick 1- 4 of those numbers and let the software find me the rest. can your software do it?

Why don't let the software pick all the numbers for you? After all, that's the whole point of GAT, to pick the most correct numbers it can find. There are various ways to use GAT, i.e. pick all the numbers at once, form groups from various GATs and mix them as matrix or even you can find more ways to use it.

[kman]

Hi,
Here is the predication comare to the actual numbers I got for 14 numbers
http://i897.photobucket.com/albums/ac17 ... 6071db.jpg

and here is the resualts when I run it with no actual numbers
http://i897.photobucket.com/albums/ac17 ... be3e10.jpg

In the second picture how can I tell which is the best strategy to take? and which field in the table to choose in order to get the best results?
should I focus on the 6/6 ? 5/6 ?

Thanks

[lottoarchitect]

This is a good question kman, I'll try to briefly describe the whole idea. First the panorama has done most of the work for you. What the panorama shows is the most promising predictions it can detect based on an overall performance each GAT ID delivers over the course of the tested draws. Each column contains the GAT ID tables that performed best at the hit category represented by that column and they are shown from best hitters to lesser good hitters (the X, X-1, X-2 feature). That means, the top GAT ID of a given column produced the most hits for the hit category represented by that column and as we move downwards, the GATs produce fewer hits than the top performing GAT ID in the same column. From this point of view, it sounds reasonable to always pick the top-most GAT from a column. This does make sense although it does not always deliver the best hits performance. I'll explain below basic principles to follow:

The easiest approach is to pick any GAT ID from those in the panorama and keep using it in the future draws, using the Run Factor mechanism. That means, as soon as you decide on a GAT ID for use with RF=0 and play the predicted numbers, you keep using this GAT ID table in the future draws too (RF>0) by picking the new prediction it proposes for each subsequent future draw.

The recommended approach is however to:
Concentrate on columns that offer regularity of hits. That means, we need to have regular hit performance from a GAT table at the desired hit category and this hits production should be experienced every so often. For example, consider the column 4-33 in your screenshots. This 4-33 means, the top GAT ID 3104309 has managed to produce 33 times a 4+ (includes 5 & 6 hits since you compute the augmentative mode) over the 100 tested draws. This is about 2.53 times better performance (the imp.ratio) for a 4+ hit than picking randomly 14 numbers to achieve a 4+ hit. However, regularity of hits means, we must experience this 4+ hit production over the whole range of those 100 tested draws, thus you have to inspect the hits graph of that GAT ID table and observe how it delivers those 4+ hits. It is much better to pick a GAT ID that does produce that 4+ hit e.g. every 2-7 draws (to make a total of 33 4+ hits) rather than having a huge 4+ hits sequence e.g. at the initial tested draws to make an equal 33 4+ hits and then produce nothing worthy for a long period. This is known as hot/cold cycles in GAT. Also, good GAT IDs to pick are those that offer the least variation of hits delay over the requested hit category (the delays graph).
Thus, although it might be obvious to pick the top-most GAT ID from a column, this GAT may experience a cold cycle and in that case another GAT below the top-most one might be preferable given it offers a much more steady hits performance judged by the regularity we look for.
Finally, I would concentrate mostly on the middle columns where they are more likely steady performers like the 4-hits column here, where 33 4-hits production already suggests a very good regularity (but still it is a good advice to inspect the graph for hits production over the whole test range) and among those with good regularity, to also prefer those that produce more 5 & 6 hits as well. Again, as soon as you decide on a GAT ID for use (RF=0), you keep using this GAT ID in the future draws too (RF>0) and every time you pick the new prediction it proposes.

[Rustamshah]

HI LA

Please explain what you mean with regular hits, I am attaching a pic with this post, could you tell me what this says ? the number at the bottom tells us how many times does it hit or its the draw number ?

http://s27.postimg.org/db7kb8eyr/GAT28.jpg

[lottoarchitect]

Hi Rustamshah, regular hits means to experience the desired hit category every so often, without waiting at any point during the test period a long time to get that desired hit. For example, assume a hits graph of 100 tested draws and a GAT table that delivers over the initial 50 tested draws a 4+ hit and 0-3 hits at the later 50 tested draws. Although here overall you have 50 4+ hits, this graph lacks that form of regularity because right now we we'll be at 50 draws without any 4+ hit! In fact, picking that particular Gat in this example is a very bad idea because this clearly demonstrates a GAT that got quite cold without any clue when it is about to start hiting. A very consistent GAT ID table with 50 4+ hits would be the one that delivers this 4+ hit at exactly every 2 draws. This consistency over the whole test range is what gives value to a very good hits statistics. So in fact you may come across two GATs that deliver the same hits at a given category but they vastly differ in the way they deliver those same hits. Regularity refers mostly to the case of not having to wait much for any specific hit to occur.
The delay graph shows this hits delay distribution. A good delay graph should show a steady decrease of height as we move to the right of the graph and a small value at the bottom-right of the hit category we are interested in. Each color of a bar indicates a particular hit category and the values below those bars are the waiting time (in draws) to experience that hit category again. The minimum delay possible is 1, which means the same hit category occurred at the previous draw. In your graph e.g. you see the red bar (which refers to a 1+ hit) right above the value 1 to have a height of 77 or so. That means, over the test period, we had 77 times the case a 1+ hit to occur right after another 1+ hit (delay is 1). Similarly, the blue line is for 2+ hits and if you check the delay 2 where the blue line has a height of 16 or 17, this means we had 16-17 times the case, a 2+ occurred after waiting 2 draws from a 2+ hit. That means, 16-17 times over the test period we had somewhere in the tested draws a 2+ hit, then something lower than a 2 hit and then again a 2+ hit (delay 2).
Also make a search for the delay graph, it is explained even more at an older post.

lottoarchitect