GAT found another 4-hit with 13 predicted

G.A.T. Engine general discussion
Post Reply
Sooz
Proficient
Proficient
Posts: 246
Joined: Wed Feb 09, 2011 12:24 am
Location: Canada

GAT found another 4-hit with 13 predicted

Post by Sooz » Sun Jun 03, 2012 4:30 pm

But again, I wheeled those 13 numbers with WG, asking for 100% 4-if-5 and ended up with 22 tickets which contained 4 separate 3-hits, but none of my tickets had the 4 correct predictions all on one ticket.
It's not so unusual anymore for GAT to find 13 predicted numbers with 4 correct. This itself is awesome! But I'm asking GAT to find tables with 5-hits. Which brings me to the bigger question -

How long does a hot table stay hot? And how long is it cold for? Once it turns cold should we forget about it? Or does it eventually return as hot as before?

Is it a function of the lottery itself? If so, then maybe one can observe several hot tables to get an idea. Or take a good look at the blue line, comparing various moving averages to see which moving average appears most cyclic.

Is it a function of how the engine works? Somehere on this blog I read that when a table turns cold, it takes a few draws for it to sort of recalibrate itself, then it becomes hot again. So how many is "a few draws"? Again, of course one can observe this by looking at the blue line. But if the answer is also a function of the engine itself, then maybe one can predict how long a cycle is.

There's no problem to observe a blue line over many draws. But for using the unofficial shorter faster method, the GAT finds hot tables, but there's no way to observe this GAT through several cycles. So I want to know if those several cycles can be predicted. Argh! This message has become rambling and I know I'm asking for the moon!

.....Sooz

draughtsman
Site Admin
Posts: 112
Joined: Thu Jan 17, 2008 4:22 am
Location: Melbourne, Australia

Re: GAT found another 4-hit with 13 predicted

Post by draughtsman » Sun Jun 03, 2012 11:58 pm

Hi Sooz,

I like your posts - always interesting that mostly come out out of a well studied process.

I am starting to come at using numbers developed another way. Instead of seeking the best GATs with the highest hits within your number spread, I am building a set of numbers from using a very low spread ie 2 or 3 numbers and then finding the best performing GATs - and then combing 4 or 5 or 6 of these to build up the number set. Usually you get several common numbers so the overall count is not excessive to use, but with these individual GATs of three numbers the success rate of prediction can be of the order 90% plus (of at least 1 number) so my 'theory' is that combined they may all offer a good chance to give a result. Yes I am aware that the statistics will say .9 x .9 x.9 x.9 etc will give the likelihood of getting all 6 but we are working with a lottery with millions of combinations and GAT is purely a predictive tool so a bit of luck always required as well.

regards

draughtsman

Sooz
Proficient
Proficient
Posts: 246
Joined: Wed Feb 09, 2011 12:24 am
Location: Canada

Re: GAT found another 4-hit with 13 predicted

Post by Sooz » Mon Jun 04, 2012 12:46 am

Thank you Draughtsman :-))

Now let me make sure I understand you.
First at the GAT Analysis Options panel, at "minimum/maximum requested numbers" you enter 2-3, or maybe 2-4, is that right?
Then for Hits category range to compute you would also enter 2-3 or 4. If I'm understanding, it wouldn't make sense to have the Hits category range larger than the number of predicted numbers.
Then you simply lump some tables (of 2 or 3 or 4 numbers each) together to create your group of numbers to wheel?
Hmmm, I'll have to think about this for awhile. I hope you won't mind answering more questions.
How many "total real draws to be tested" do you use in the options panel? How many GATS has the engine analyzed before finding the best performing GATs?

I had considered lumping a couple of tables of 12 predicted numbers together, but even with the common numbers, the total to be wheeled was too many for my budget. It never occurred to me to lump together some very small tables. What sort of success have you had?

.....Sooz

draughtsman
Site Admin
Posts: 112
Joined: Thu Jan 17, 2008 4:22 am
Location: Melbourne, Australia

Re: GAT found another 4-hit with 13 predicted

Post by draughtsman » Mon Jun 04, 2012 1:47 am

First at the GAT Analysis Options panel, at "minimum/maximum requested numbers" you enter 2-3, or maybe 2-4, is that right?
Then for Hits category range to compute you would also enter 2-3 or 4. If I'm understanding, it wouldn't make sense to have the Hits category range larger than the number of predicted numbers.
I have been putting it at 6 max - just like to see what is out there. It doesn't matter how many you have - is quicker to process with fewer
Then you simply lump some tables (of 2 or 3 or 4 numbers each) together to create your group of numbers to wheel?
Correct.
How many "total real draws to be tested" do you use in the options panel?
Just using 20. Less averaging with a fewer draws ie the predictive process doesn't have to fit across too many draws.
How many GATS has the engine analyzed before finding the best performing GATs?
I presume you mean total statistical data here? I use 20.

I have only done this twice - 3 numbers and 3 numbers.

I believe Anastasios has an approach he calls Synthetic GAT in his black book of things to do where several GATs are combined to produce a good outcome. This is similar to that.

regards

draughtsman

Sooz
Proficient
Proficient
Posts: 246
Joined: Wed Feb 09, 2011 12:24 am
Location: Canada

Re: GAT found another 4-hit with 13 predicted

Post by Sooz » Mon Jun 04, 2012 4:18 am

Draughtsman thank you for your speedy comprehensive reply. I still need to think about this method.
First I ask myself - why combine 2 or more sets of selected numbers by hand when that is precisely what the engine does in the first place - ie it finds the best numbers to combine.
But then I answer myself :-) When GAT finds more numbers to combine, it does so in the same table. That is, it adds more numbers to the same table. We still have to decide which table to use. Your suggested method eliminates that decision by combining tables. Sounds like a good idea.
Nevertheless, I can only do one thing at a time. For now, I'll simply increase the number of draws tested.
.....Sooz

draughtsman
Site Admin
Posts: 112
Joined: Thu Jan 17, 2008 4:22 am
Location: Melbourne, Australia

Re: GAT found another 4-hit with 13 predicted

Post by draughtsman » Thu Jun 07, 2012 11:55 am

Hi Sooz,

I am beginning to like my approach - in powerball (5/45 with PB 1/45) tonight GAT gave me 4 in 5 within a 14 number spread.

It actually gave me 5 in 5 but as this would have given me 15 numbers to work with for wheel making so I deleted a number (26) and sure enough out it came. (I should add I have never had anything like this before.) :D

GAT missed the power ball unfortunately.

Whilst this result is very desirable - Is it any better than just going for a 14 number spread out of GAT? I don't know as I didn't keep a record of a 14 number grouping. Such a grouping may well of course have done exactly the same.

But needless to say I shall be continuing with this method and see where it leads.

Actually, a bit frustrated :( as I skimped on the wheel size and so only ended up with 2 hits of 3 numbers. Ahh.... playing to one's budget.

Never the less GAT and WG - what excellent tools they are for playing one's lottery. Thanks Mr LA.

draughtsman

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: GAT found another 4-hit with 13 predicted

Post by lottoarchitect » Thu Jun 07, 2012 1:15 pm

Nice GAT seeking I sense over here! Good result draughtsman, although unlucky with the actual wheel within budget. Nevertheless, keep it up! :)

draughtsman
Site Admin
Posts: 112
Joined: Thu Jan 17, 2008 4:22 am
Location: Melbourne, Australia

Re: GAT found another 4-hit with 13 predicted

Post by draughtsman » Thu Jun 14, 2012 11:39 am

Hi Sooz,

Another minor prize result tonight - 3 numbers in 14 (for Powerball 5/45 and 1/45) all nicely wheeled by WG with more lines to work with in the wheel this time.

I have been using 20 draws and 20 stat data draws but could not get a clean set of GAT tables as I did last week (with 5 tables of 3 giving me a prediction of 19 in the 20) - so I changed the stat data draws to 21 but still could not get a straight set of GATs to work with. Thus my number set ended up being a bit synthetic derived from the best GATs I could determine. The powerball was missed by GAT unfortunately.

I let the 'puter run over night developing out to about 14m GAT tables where the GATs I used ranged from about 6m to 11m so it appears worth it to let it run to seek the better GATs.

Good luck

draughtsman

Post Reply

Who is online

Users browsing this forum: No registered users and 28 guests